Original Research

Limb loss is a significant and growing concern in the United States. Nearly 2 million Americans are living with limb loss, and up to 185,000 people undergo amputations annually.^1-4 Of these patients, about 35% are women.⁵ The Veterans Health Administration (VHA) provides about 10% of US amputations.^6-8 Between 2015 and 2019, the number of prosthetic devices provided to female veterans increased from 3.3 million to 4.6 million.^5,9,10

Previous research identified disparities in prosthetic care between men and women, both within and outside the VHA. These disparities include slower prosthesis prescription and receipt among women, in addition to differences in self-reported mobility, satisfaction, rates of prosthesis rejection, and challenges related to prosthesis appearance and fit.^5,10,11 Recent studies suggest women tend to have worse outcomes following amputation, and are underrepresented in amputation research.^12,13 However, these disparities are poorly described in a large, national sample. Because women represent a growing portion of patients with limb loss in the VHA, understanding their needs is critical.¹⁴

The Johnny Isakson and David P. Roe, MD Veterans Health Care and Benefits Improvement Act of 2020 was enacted, in part, to improve the care provided to women veterans.¹⁵ The law required the VHA to conduct a survey of ≥ 50,000 veterans to assess the satisfaction of women veterans with prostheses provided by the VHA. To comply with this legislation and understand how women veterans rate their prostheses and related care in the VHA, the US Department of Veterans Affairs (VA) Center for Collaborative Evaluation (VACE) conducted a large national survey of veterans with limb loss that oversampled women veterans. This article describes the survey results, including characteristics of female veterans with limb loss receiving care from the VHA, assesses their satisfaction with prostheses and prosthetic care, and highlights where their responses differ from those of male veterans.

Methods

We conducted a cross-sectional, mixedmode survey of eligible amputees in the VHA Support Service Capital Assets Amputee Data Cube. We identified a cohort of veterans with any major amputation (above the ankle or wrist) or partial hand or foot amputation who received VHA care between October 1, 2019, and September 30, 2020. The final cohort yielded 46,646 potentially eligible veterans. Thirty-three had invalid contact information, leaving 46,613 veterans who were asked to participate, including 1356 women.

Survey

We created a survey instrument de novo that included questions from validated instruments, including the Trinity Amputation Prosthesis and Experience Scales to assess prosthetic device satisfaction, the Prosthesis Evaluation Questionnaire to assess quality of life (QOL) satisfaction, and the Orthotics Prosthetics Users Survey to assess prosthesis-related care satisfaction. ^16-18 Additional questions were incorporated from a survey of veterans with upper limb amputation to assess the importance of cosmetic considerations related to the prosthesis and comfort with prosthesis use in intimate relationships.¹⁹ Questions were also included to assess amputation type, year of amputation, if a prosthesis was currently used, reasons for ceasing use of a prosthesis, reasons for never using a prosthesis, the types of prostheses used, intensity of prosthesis use, satisfaction with time required to receive a prosthetic limb, and if the prosthesis reflected the veteran’s selfidentified gender. Veterans were asked to answer questions based on their most recent amputation.

We tested the survey using cognitive interviews with 6 veterans to refine the survey and better understand how veterans interpreted the questions. Pilot testers completed the survey and participated in individual interviews with experienced interviewers (CL and RRK) to describe how they selected their responses.²⁰ This feedback was used to refine the survey. The online survey was programmed using Qualtrics Software and manually translated into Spanish.

Given the multimodal design, surveys were distributed by email, text message, and US Postal Service (USPS). Surveys were emailed to all veterans for whom a valid email address was available. If emails were undeliverable, veterans were contacted via text message or the USPS. Surveys were distributed by text message to all veterans without an email address but with a cellphone number. We were unable to consistently identify invalid numbers among all text message recipients. Invitations with a survey URL and QR code were sent via USPS to veterans who had no valid email address or cellphone number. Targeted efforts were made to increase the response rate for women. A random sample of 200 women who had not completed the survey 2 weeks prior to the closing date (15% of women in sample) was selected to receive personal phone calls. Another random sample of 400 women was selected to receive personalized outreach emails. The survey data were confidential, and responses could not be traced to identifying information.

Data Analyses

We conducted a descriptive analysis, including percentages and means for responses to variables focused on describing amputation characteristics, prosthesis characteristics, and QOL. All data, including missing values, were used to document the percentage of respondents for each question. Removing missing data from the denominator when calculating percentages could introduce bias to the analysis because we cannot be certain data are missing at random. Missing variables were removed to avoid underinflation of mean scores.

We compared responses across 2 groups: individuals who self-identified as men and individuals who self-identified as women. For each question, we assessed whether each of these groups differed significantly from the remaining sample. For example, we examined whether the percentage of men who answered affirmatively to a question was significantly higher or lower than that of individuals not identifying as male, and whether the percentage of women who answered affirmatively was significantly higher or lower than that of individuals not identifying as female. We utilized x² tests to determine significant differences for percentage calculations and t tests to determine significant differences in means across gender.

Since conducting multiple comparisons within a dataset may result in inflating statistical significance (type 1 errors), we used a more conservative estimate of statistical significance (α = 0.01) and high significance (α = 0.001). This study was deemed quality improvement by the VHA Rehabilitation and Prosthetic Services (12RPS) and acknowledged by the VA Research Office at Eastern Colorado Health Care System and was not subject to institutional review board review.

Results

Surveys were distributed to 46,613 veterans and were completed by 4981 respondents for a 10.7% overall response rate. Survey respondents were generally similar to the eligible population invited to participate, but the proportion of women who completed the survey was higher than the proportion of women eligible to participate (2.0% of eligible population vs 16.7% of respondents), likely due to specific efforts to target women. Survey respondents were slightly younger than the general population (67.3 years vs 68.7 years), less likely to be male (97.1% vs 83.3%), showed similar representation of Operation Enduring Freedom/Operation Iraqi Freedom (OEF/OIF) veterans (4.4% vs 4.1%), and were less likely to have diabetes (58.0% vs 52.7% had diabetes) (Table 1).

The mean age of male respondents was 67.3 years, while the mean age of female respondents was 58.3 years. The majority of respondents were male (83.3%) and White (77.2%). Female respondents were less likely to have diabetes (35.4% of women vs 53.5% of men) and less likely to report that their most recent amputation resulted from diabetes (10.1% of women vs 22.2% of men). Women respondents were more likely to report an amputation due to other causes, such as adverse results of surgery, neurologic disease, suicide attempt, blood clots, tumors, rheumatoid arthritis, and revisions of previous amputations. Most women respondents did not serve during the OEF or OIF eras. The most common amputation site for women respondents was lower limb, either below the knee and above the ankle or above the knee.

Most participants use an everyday prosthesis, but women were more likely to report using a sports-specific prosthesis (Table 2). Overall, most respondents report using a prosthesis (87.7%); however, women were more likely to report not using a prosthesis (19.4% of women vs 11.1% of men; P ≤ .01). Additionally, a lower proportion of women report using a prosthesis for < 12 hours per day (30.6% of women vs 46.4% of men; P ≤ .01) or using a prosthesis every day (54.8% of women vs 74.6% of men; P ≤ .001).

In the overall sample, the mean satisfaction score with a prosthesis was 2.7 on a 5-point scale, and women had slightly lower overall satisfaction scores (2.6 for women vs 2.7 for men; P ≤ .001) (Table 3). Women also had lower satisfaction scores related to appearance, usefulness, reliability, and comfort. Women were more likely to indicate that it was very important to be able to wear jewelry and accessories (20.2% of women vs 11.6% of men; P ≤ .01), while men were less likely to indicate that it was somewhat or very important that the prosthesis not restrict clothing or shoes (95.2% of women vs 82.9% of men; P ≤ .001). Men were more likely than women to report being comfortable or very comfortable using their prosthesis in intimate contact: 40.5% vs 29.0%, respectively (P ≤ .001).

Overall, participants reported high satisfaction with appointment times, wait times, courteous treatment, opportunities to express concerns, and staff responsiveness. Men were slightly more likely than women to be satisfied with training (P ≤ 0.001) and problem discussion (P ≤ 0.01) (Table 4). There were no statistically significant differences in satisfaction or QOL ratings between women and men. The overall sample rated both QOL and satisfaction with QOL 6.7 on a 10-point scale.

Discussion

The goal of this study was to characterize the experience of veterans with limb loss receiving care in the VHA and assess their satisfaction with prostheses and prosthetic care. We received responses from nearly 5000 veterans, 158 of whom were women. Women veteran respondents were slightly younger and less likely to have an amputation due to diabetes. We did not observe significant differences in amputation level between men and women but women were less likely to use a prosthesis, reported lower intensity of prosthesis use, and were less satisfied with certain aspects of their prostheses. Women may also be less satisfied with prosthesis training and problem discussion. However, we found no differences in QOL ratings between men and women.

Findings indicating women were more likely to report not using a prosthesis and that a lower proportion of women report using a prosthesis for > 12 hours a day or every day are consistent with previous research. 21,22 Interestingly, women were more likely to report using a sports-specific prosthesis. This is notable because prior research suggests that individuals with amputations may avoid participating in sports and exercise, and a lack of access to sports-specific prostheses may inhibit physical activity.^23,24 Women in this sample were slightly less satisfied with their prostheses overall and reported lower satisfaction scores regarding appearance, usefulness, reliability, and comfort, consistent with previous findings.²⁵

A lower percentage of women in this sample reported being comfortable or very comfortable using their prosthesis during intimate contact. Previous research on prosthesis satisfaction suggests individuals who rate prosthesis satisfaction lower also report lower body image across genders. ²⁶ While women in this sample did not rate their prosthesis satisfaction lower than men, they did report lower intensity of prosthesis use, suggesting potential issues with their prostheses this survey did not evaluate. Women indicated the importance of prostheses not restricting jewelry, accessories, clothing, or shoes. These results have significant clinical and social implications. A recent qualitative study emphasizes that women veterans feel prostheses are primarily designed for men and may not work well with their physiological needs.⁹ Research focused on limbs better suited to women’s bodies could result in better fitting sockets, lightweight limbs, or less bulky designs. Additional research has also explored the difficulties in accommodating a range of footwear for patients with lower limb amputation. One study found that varying footwear heights affect the function of adjustable prosthetic feet in ways that may not be optimal.²⁷

Ratings of satisfaction with prosthesisrelated services between men and women in this sample are consistent with a recent study showing that women veterans do not have significant differences in satisfaction with prosthesis-related services.²⁸ However, this study focused specifically on lower limb amputations, while the respondents of this study include those with both upper and lower limb amputations. Importantly, our findings that women are less likely to be satisfied with prosthesis training and problem discussions support recent qualitative findings in which women expressed a desire to work with prosthetists who listen to them, take their concerns seriously, and seek solutions that fit their needs. We did not observe a difference in QOL ratings between men and women in the sample despite lower satisfaction among women with some elements of prosthesis-related services. Previous research suggests many factors impact QOL after amputation, most notably time since amputation.^16,29

Limitations

This survey was deployed in a short timeline that did not allow for careful sample selection or implementing strategies to increase response rate. Additionally, the study was conducted among veterans receiving care in the VHA, and findings may not be generalizable to limb loss in other settings. Finally, the discrepancy in number of respondents who identified as men vs women made it difficult to compare differences between the 2 groups.

Conclusions

This is the largest sample of survey respondents of veterans with limb loss to date. While the findings suggest veterans are generally satisfied with prosthetic-related services overall, they also highlight several areas for improvement with services or prostheses. Given that most veterans with limb loss are men, there is a significant discrepancy between the number of women and men respondents. Additional studies with more comparable numbers of men and women have found similar ratings of satisfaction with prostheses and services.²⁸ Further research specifically focused on improving the experiences of women should focus on better characterizing their experiences and identifying how they differ from those of male veterans. For example, understanding how to engage female veterans with limb loss in prosthesis training and problem discussions may improve their experience with their care teams and improve their use of prostheses. Understanding experiences and needs that are specific to women could lead to the development of processes, resources, or devices that are tailored to the unique requirements of women with limb loss.

Limb loss is a significant and growing concern in the United States. Nearly 2 million Americans are living with limb loss, and up to 185,000 people undergo amputations annually.^1-4 Of these patients, about 35% are women.⁵ The Veterans Health Administration (VHA) provides about 10% of US amputations.^6-8 Between 2015 and 2019, the number of prosthetic devices provided to female veterans increased from 3.3 million to 4.6 million.^5,9,10

Previous research identified disparities in prosthetic care between men and women, both within and outside the VHA. These disparities include slower prosthesis prescription and receipt among women, in addition to differences in self-reported mobility, satisfaction, rates of prosthesis rejection, and challenges related to prosthesis appearance and fit.^5,10,11 Recent studies suggest women tend to have worse outcomes following amputation, and are underrepresented in amputation research.^12,13 However, these disparities are poorly described in a large, national sample. Because women represent a growing portion of patients with limb loss in the VHA, understanding their needs is critical.¹⁴

The Johnny Isakson and David P. Roe, MD Veterans Health Care and Benefits Improvement Act of 2020 was enacted, in part, to improve the care provided to women veterans.¹⁵ The law required the VHA to conduct a survey of ≥ 50,000 veterans to assess the satisfaction of women veterans with prostheses provided by the VHA. To comply with this legislation and understand how women veterans rate their prostheses and related care in the VHA, the US Department of Veterans Affairs (VA) Center for Collaborative Evaluation (VACE) conducted a large national survey of veterans with limb loss that oversampled women veterans. This article describes the survey results, including characteristics of female veterans with limb loss receiving care from the VHA, assesses their satisfaction with prostheses and prosthetic care, and highlights where their responses differ from those of male veterans.

Methods

We conducted a cross-sectional, mixedmode survey of eligible amputees in the VHA Support Service Capital Assets Amputee Data Cube. We identified a cohort of veterans with any major amputation (above the ankle or wrist) or partial hand or foot amputation who received VHA care between October 1, 2019, and September 30, 2020. The final cohort yielded 46,646 potentially eligible veterans. Thirty-three had invalid contact information, leaving 46,613 veterans who were asked to participate, including 1356 women.

Survey

We created a survey instrument de novo that included questions from validated instruments, including the Trinity Amputation Prosthesis and Experience Scales to assess prosthetic device satisfaction, the Prosthesis Evaluation Questionnaire to assess quality of life (QOL) satisfaction, and the Orthotics Prosthetics Users Survey to assess prosthesis-related care satisfaction. ^16-18 Additional questions were incorporated from a survey of veterans with upper limb amputation to assess the importance of cosmetic considerations related to the prosthesis and comfort with prosthesis use in intimate relationships.¹⁹ Questions were also included to assess amputation type, year of amputation, if a prosthesis was currently used, reasons for ceasing use of a prosthesis, reasons for never using a prosthesis, the types of prostheses used, intensity of prosthesis use, satisfaction with time required to receive a prosthetic limb, and if the prosthesis reflected the veteran’s selfidentified gender. Veterans were asked to answer questions based on their most recent amputation.

We tested the survey using cognitive interviews with 6 veterans to refine the survey and better understand how veterans interpreted the questions. Pilot testers completed the survey and participated in individual interviews with experienced interviewers (CL and RRK) to describe how they selected their responses.²⁰ This feedback was used to refine the survey. The online survey was programmed using Qualtrics Software and manually translated into Spanish.

Given the multimodal design, surveys were distributed by email, text message, and US Postal Service (USPS). Surveys were emailed to all veterans for whom a valid email address was available. If emails were undeliverable, veterans were contacted via text message or the USPS. Surveys were distributed by text message to all veterans without an email address but with a cellphone number. We were unable to consistently identify invalid numbers among all text message recipients. Invitations with a survey URL and QR code were sent via USPS to veterans who had no valid email address or cellphone number. Targeted efforts were made to increase the response rate for women. A random sample of 200 women who had not completed the survey 2 weeks prior to the closing date (15% of women in sample) was selected to receive personal phone calls. Another random sample of 400 women was selected to receive personalized outreach emails. The survey data were confidential, and responses could not be traced to identifying information.

Data Analyses

We conducted a descriptive analysis, including percentages and means for responses to variables focused on describing amputation characteristics, prosthesis characteristics, and QOL. All data, including missing values, were used to document the percentage of respondents for each question. Removing missing data from the denominator when calculating percentages could introduce bias to the analysis because we cannot be certain data are missing at random. Missing variables were removed to avoid underinflation of mean scores.

We compared responses across 2 groups: individuals who self-identified as men and individuals who self-identified as women. For each question, we assessed whether each of these groups differed significantly from the remaining sample. For example, we examined whether the percentage of men who answered affirmatively to a question was significantly higher or lower than that of individuals not identifying as male, and whether the percentage of women who answered affirmatively was significantly higher or lower than that of individuals not identifying as female. We utilized x² tests to determine significant differences for percentage calculations and t tests to determine significant differences in means across gender.

Since conducting multiple comparisons within a dataset may result in inflating statistical significance (type 1 errors), we used a more conservative estimate of statistical significance (α = 0.01) and high significance (α = 0.001). This study was deemed quality improvement by the VHA Rehabilitation and Prosthetic Services (12RPS) and acknowledged by the VA Research Office at Eastern Colorado Health Care System and was not subject to institutional review board review.

Results

Surveys were distributed to 46,613 veterans and were completed by 4981 respondents for a 10.7% overall response rate. Survey respondents were generally similar to the eligible population invited to participate, but the proportion of women who completed the survey was higher than the proportion of women eligible to participate (2.0% of eligible population vs 16.7% of respondents), likely due to specific efforts to target women. Survey respondents were slightly younger than the general population (67.3 years vs 68.7 years), less likely to be male (97.1% vs 83.3%), showed similar representation of Operation Enduring Freedom/Operation Iraqi Freedom (OEF/OIF) veterans (4.4% vs 4.1%), and were less likely to have diabetes (58.0% vs 52.7% had diabetes) (Table 1).

The mean age of male respondents was 67.3 years, while the mean age of female respondents was 58.3 years. The majority of respondents were male (83.3%) and White (77.2%). Female respondents were less likely to have diabetes (35.4% of women vs 53.5% of men) and less likely to report that their most recent amputation resulted from diabetes (10.1% of women vs 22.2% of men). Women respondents were more likely to report an amputation due to other causes, such as adverse results of surgery, neurologic disease, suicide attempt, blood clots, tumors, rheumatoid arthritis, and revisions of previous amputations. Most women respondents did not serve during the OEF or OIF eras. The most common amputation site for women respondents was lower limb, either below the knee and above the ankle or above the knee.

Most participants use an everyday prosthesis, but women were more likely to report using a sports-specific prosthesis (Table 2). Overall, most respondents report using a prosthesis (87.7%); however, women were more likely to report not using a prosthesis (19.4% of women vs 11.1% of men; P ≤ .01). Additionally, a lower proportion of women report using a prosthesis for < 12 hours per day (30.6% of women vs 46.4% of men; P ≤ .01) or using a prosthesis every day (54.8% of women vs 74.6% of men; P ≤ .001).

In the overall sample, the mean satisfaction score with a prosthesis was 2.7 on a 5-point scale, and women had slightly lower overall satisfaction scores (2.6 for women vs 2.7 for men; P ≤ .001) (Table 3). Women also had lower satisfaction scores related to appearance, usefulness, reliability, and comfort. Women were more likely to indicate that it was very important to be able to wear jewelry and accessories (20.2% of women vs 11.6% of men; P ≤ .01), while men were less likely to indicate that it was somewhat or very important that the prosthesis not restrict clothing or shoes (95.2% of women vs 82.9% of men; P ≤ .001). Men were more likely than women to report being comfortable or very comfortable using their prosthesis in intimate contact: 40.5% vs 29.0%, respectively (P ≤ .001).

Overall, participants reported high satisfaction with appointment times, wait times, courteous treatment, opportunities to express concerns, and staff responsiveness. Men were slightly more likely than women to be satisfied with training (P ≤ 0.001) and problem discussion (P ≤ 0.01) (Table 4). There were no statistically significant differences in satisfaction or QOL ratings between women and men. The overall sample rated both QOL and satisfaction with QOL 6.7 on a 10-point scale.

Discussion

The goal of this study was to characterize the experience of veterans with limb loss receiving care in the VHA and assess their satisfaction with prostheses and prosthetic care. We received responses from nearly 5000 veterans, 158 of whom were women. Women veteran respondents were slightly younger and less likely to have an amputation due to diabetes. We did not observe significant differences in amputation level between men and women but women were less likely to use a prosthesis, reported lower intensity of prosthesis use, and were less satisfied with certain aspects of their prostheses. Women may also be less satisfied with prosthesis training and problem discussion. However, we found no differences in QOL ratings between men and women.

Findings indicating women were more likely to report not using a prosthesis and that a lower proportion of women report using a prosthesis for > 12 hours a day or every day are consistent with previous research. 21,22 Interestingly, women were more likely to report using a sports-specific prosthesis. This is notable because prior research suggests that individuals with amputations may avoid participating in sports and exercise, and a lack of access to sports-specific prostheses may inhibit physical activity.^23,24 Women in this sample were slightly less satisfied with their prostheses overall and reported lower satisfaction scores regarding appearance, usefulness, reliability, and comfort, consistent with previous findings.²⁵

A lower percentage of women in this sample reported being comfortable or very comfortable using their prosthesis during intimate contact. Previous research on prosthesis satisfaction suggests individuals who rate prosthesis satisfaction lower also report lower body image across genders. ²⁶ While women in this sample did not rate their prosthesis satisfaction lower than men, they did report lower intensity of prosthesis use, suggesting potential issues with their prostheses this survey did not evaluate. Women indicated the importance of prostheses not restricting jewelry, accessories, clothing, or shoes. These results have significant clinical and social implications. A recent qualitative study emphasizes that women veterans feel prostheses are primarily designed for men and may not work well with their physiological needs.⁹ Research focused on limbs better suited to women’s bodies could result in better fitting sockets, lightweight limbs, or less bulky designs. Additional research has also explored the difficulties in accommodating a range of footwear for patients with lower limb amputation. One study found that varying footwear heights affect the function of adjustable prosthetic feet in ways that may not be optimal.²⁷

Ratings of satisfaction with prosthesisrelated services between men and women in this sample are consistent with a recent study showing that women veterans do not have significant differences in satisfaction with prosthesis-related services.²⁸ However, this study focused specifically on lower limb amputations, while the respondents of this study include those with both upper and lower limb amputations. Importantly, our findings that women are less likely to be satisfied with prosthesis training and problem discussions support recent qualitative findings in which women expressed a desire to work with prosthetists who listen to them, take their concerns seriously, and seek solutions that fit their needs. We did not observe a difference in QOL ratings between men and women in the sample despite lower satisfaction among women with some elements of prosthesis-related services. Previous research suggests many factors impact QOL after amputation, most notably time since amputation.^16,29

Limitations

This survey was deployed in a short timeline that did not allow for careful sample selection or implementing strategies to increase response rate. Additionally, the study was conducted among veterans receiving care in the VHA, and findings may not be generalizable to limb loss in other settings. Finally, the discrepancy in number of respondents who identified as men vs women made it difficult to compare differences between the 2 groups.

Conclusions

This is the largest sample of survey respondents of veterans with limb loss to date. While the findings suggest veterans are generally satisfied with prosthetic-related services overall, they also highlight several areas for improvement with services or prostheses. Given that most veterans with limb loss are men, there is a significant discrepancy between the number of women and men respondents. Additional studies with more comparable numbers of men and women have found similar ratings of satisfaction with prostheses and services.²⁸ Further research specifically focused on improving the experiences of women should focus on better characterizing their experiences and identifying how they differ from those of male veterans. For example, understanding how to engage female veterans with limb loss in prosthesis training and problem discussions may improve their experience with their care teams and improve their use of prostheses. Understanding experiences and needs that are specific to women could lead to the development of processes, resources, or devices that are tailored to the unique requirements of women with limb loss.

Limb loss is a significant and growing concern in the United States. Nearly 2 million Americans are living with limb loss, and up to 185,000 people undergo amputations annually.^1-4 Of these patients, about 35% are women.⁵ The Veterans Health Administration (VHA) provides about 10% of US amputations.^6-8 Between 2015 and 2019, the number of prosthetic devices provided to female veterans increased from 3.3 million to 4.6 million.^5,9,10

Previous research identified disparities in prosthetic care between men and women, both within and outside the VHA. These disparities include slower prosthesis prescription and receipt among women, in addition to differences in self-reported mobility, satisfaction, rates of prosthesis rejection, and challenges related to prosthesis appearance and fit.^5,10,11 Recent studies suggest women tend to have worse outcomes following amputation, and are underrepresented in amputation research.^12,13 However, these disparities are poorly described in a large, national sample. Because women represent a growing portion of patients with limb loss in the VHA, understanding their needs is critical.¹⁴

The Johnny Isakson and David P. Roe, MD Veterans Health Care and Benefits Improvement Act of 2020 was enacted, in part, to improve the care provided to women veterans.¹⁵ The law required the VHA to conduct a survey of ≥ 50,000 veterans to assess the satisfaction of women veterans with prostheses provided by the VHA. To comply with this legislation and understand how women veterans rate their prostheses and related care in the VHA, the US Department of Veterans Affairs (VA) Center for Collaborative Evaluation (VACE) conducted a large national survey of veterans with limb loss that oversampled women veterans. This article describes the survey results, including characteristics of female veterans with limb loss receiving care from the VHA, assesses their satisfaction with prostheses and prosthetic care, and highlights where their responses differ from those of male veterans.

Methods

We conducted a cross-sectional, mixedmode survey of eligible amputees in the VHA Support Service Capital Assets Amputee Data Cube. We identified a cohort of veterans with any major amputation (above the ankle or wrist) or partial hand or foot amputation who received VHA care between October 1, 2019, and September 30, 2020. The final cohort yielded 46,646 potentially eligible veterans. Thirty-three had invalid contact information, leaving 46,613 veterans who were asked to participate, including 1356 women.

Survey

We created a survey instrument de novo that included questions from validated instruments, including the Trinity Amputation Prosthesis and Experience Scales to assess prosthetic device satisfaction, the Prosthesis Evaluation Questionnaire to assess quality of life (QOL) satisfaction, and the Orthotics Prosthetics Users Survey to assess prosthesis-related care satisfaction. ^16-18 Additional questions were incorporated from a survey of veterans with upper limb amputation to assess the importance of cosmetic considerations related to the prosthesis and comfort with prosthesis use in intimate relationships.¹⁹ Questions were also included to assess amputation type, year of amputation, if a prosthesis was currently used, reasons for ceasing use of a prosthesis, reasons for never using a prosthesis, the types of prostheses used, intensity of prosthesis use, satisfaction with time required to receive a prosthetic limb, and if the prosthesis reflected the veteran’s selfidentified gender. Veterans were asked to answer questions based on their most recent amputation.

We tested the survey using cognitive interviews with 6 veterans to refine the survey and better understand how veterans interpreted the questions. Pilot testers completed the survey and participated in individual interviews with experienced interviewers (CL and RRK) to describe how they selected their responses.²⁰ This feedback was used to refine the survey. The online survey was programmed using Qualtrics Software and manually translated into Spanish.

Given the multimodal design, surveys were distributed by email, text message, and US Postal Service (USPS). Surveys were emailed to all veterans for whom a valid email address was available. If emails were undeliverable, veterans were contacted via text message or the USPS. Surveys were distributed by text message to all veterans without an email address but with a cellphone number. We were unable to consistently identify invalid numbers among all text message recipients. Invitations with a survey URL and QR code were sent via USPS to veterans who had no valid email address or cellphone number. Targeted efforts were made to increase the response rate for women. A random sample of 200 women who had not completed the survey 2 weeks prior to the closing date (15% of women in sample) was selected to receive personal phone calls. Another random sample of 400 women was selected to receive personalized outreach emails. The survey data were confidential, and responses could not be traced to identifying information.

Data Analyses

We conducted a descriptive analysis, including percentages and means for responses to variables focused on describing amputation characteristics, prosthesis characteristics, and QOL. All data, including missing values, were used to document the percentage of respondents for each question. Removing missing data from the denominator when calculating percentages could introduce bias to the analysis because we cannot be certain data are missing at random. Missing variables were removed to avoid underinflation of mean scores.

We compared responses across 2 groups: individuals who self-identified as men and individuals who self-identified as women. For each question, we assessed whether each of these groups differed significantly from the remaining sample. For example, we examined whether the percentage of men who answered affirmatively to a question was significantly higher or lower than that of individuals not identifying as male, and whether the percentage of women who answered affirmatively was significantly higher or lower than that of individuals not identifying as female. We utilized x² tests to determine significant differences for percentage calculations and t tests to determine significant differences in means across gender.

Since conducting multiple comparisons within a dataset may result in inflating statistical significance (type 1 errors), we used a more conservative estimate of statistical significance (α = 0.01) and high significance (α = 0.001). This study was deemed quality improvement by the VHA Rehabilitation and Prosthetic Services (12RPS) and acknowledged by the VA Research Office at Eastern Colorado Health Care System and was not subject to institutional review board review.

Results

Surveys were distributed to 46,613 veterans and were completed by 4981 respondents for a 10.7% overall response rate. Survey respondents were generally similar to the eligible population invited to participate, but the proportion of women who completed the survey was higher than the proportion of women eligible to participate (2.0% of eligible population vs 16.7% of respondents), likely due to specific efforts to target women. Survey respondents were slightly younger than the general population (67.3 years vs 68.7 years), less likely to be male (97.1% vs 83.3%), showed similar representation of Operation Enduring Freedom/Operation Iraqi Freedom (OEF/OIF) veterans (4.4% vs 4.1%), and were less likely to have diabetes (58.0% vs 52.7% had diabetes) (Table 1).

The mean age of male respondents was 67.3 years, while the mean age of female respondents was 58.3 years. The majority of respondents were male (83.3%) and White (77.2%). Female respondents were less likely to have diabetes (35.4% of women vs 53.5% of men) and less likely to report that their most recent amputation resulted from diabetes (10.1% of women vs 22.2% of men). Women respondents were more likely to report an amputation due to other causes, such as adverse results of surgery, neurologic disease, suicide attempt, blood clots, tumors, rheumatoid arthritis, and revisions of previous amputations. Most women respondents did not serve during the OEF or OIF eras. The most common amputation site for women respondents was lower limb, either below the knee and above the ankle or above the knee.

Most participants use an everyday prosthesis, but women were more likely to report using a sports-specific prosthesis (Table 2). Overall, most respondents report using a prosthesis (87.7%); however, women were more likely to report not using a prosthesis (19.4% of women vs 11.1% of men; P ≤ .01). Additionally, a lower proportion of women report using a prosthesis for < 12 hours per day (30.6% of women vs 46.4% of men; P ≤ .01) or using a prosthesis every day (54.8% of women vs 74.6% of men; P ≤ .001).

In the overall sample, the mean satisfaction score with a prosthesis was 2.7 on a 5-point scale, and women had slightly lower overall satisfaction scores (2.6 for women vs 2.7 for men; P ≤ .001) (Table 3). Women also had lower satisfaction scores related to appearance, usefulness, reliability, and comfort. Women were more likely to indicate that it was very important to be able to wear jewelry and accessories (20.2% of women vs 11.6% of men; P ≤ .01), while men were less likely to indicate that it was somewhat or very important that the prosthesis not restrict clothing or shoes (95.2% of women vs 82.9% of men; P ≤ .001). Men were more likely than women to report being comfortable or very comfortable using their prosthesis in intimate contact: 40.5% vs 29.0%, respectively (P ≤ .001).

Overall, participants reported high satisfaction with appointment times, wait times, courteous treatment, opportunities to express concerns, and staff responsiveness. Men were slightly more likely than women to be satisfied with training (P ≤ 0.001) and problem discussion (P ≤ 0.01) (Table 4). There were no statistically significant differences in satisfaction or QOL ratings between women and men. The overall sample rated both QOL and satisfaction with QOL 6.7 on a 10-point scale.

Discussion

The goal of this study was to characterize the experience of veterans with limb loss receiving care in the VHA and assess their satisfaction with prostheses and prosthetic care. We received responses from nearly 5000 veterans, 158 of whom were women. Women veteran respondents were slightly younger and less likely to have an amputation due to diabetes. We did not observe significant differences in amputation level between men and women but women were less likely to use a prosthesis, reported lower intensity of prosthesis use, and were less satisfied with certain aspects of their prostheses. Women may also be less satisfied with prosthesis training and problem discussion. However, we found no differences in QOL ratings between men and women.

Findings indicating women were more likely to report not using a prosthesis and that a lower proportion of women report using a prosthesis for > 12 hours a day or every day are consistent with previous research. 21,22 Interestingly, women were more likely to report using a sports-specific prosthesis. This is notable because prior research suggests that individuals with amputations may avoid participating in sports and exercise, and a lack of access to sports-specific prostheses may inhibit physical activity.^23,24 Women in this sample were slightly less satisfied with their prostheses overall and reported lower satisfaction scores regarding appearance, usefulness, reliability, and comfort, consistent with previous findings.²⁵

A lower percentage of women in this sample reported being comfortable or very comfortable using their prosthesis during intimate contact. Previous research on prosthesis satisfaction suggests individuals who rate prosthesis satisfaction lower also report lower body image across genders. ²⁶ While women in this sample did not rate their prosthesis satisfaction lower than men, they did report lower intensity of prosthesis use, suggesting potential issues with their prostheses this survey did not evaluate. Women indicated the importance of prostheses not restricting jewelry, accessories, clothing, or shoes. These results have significant clinical and social implications. A recent qualitative study emphasizes that women veterans feel prostheses are primarily designed for men and may not work well with their physiological needs.⁹ Research focused on limbs better suited to women’s bodies could result in better fitting sockets, lightweight limbs, or less bulky designs. Additional research has also explored the difficulties in accommodating a range of footwear for patients with lower limb amputation. One study found that varying footwear heights affect the function of adjustable prosthetic feet in ways that may not be optimal.²⁷

Ratings of satisfaction with prosthesisrelated services between men and women in this sample are consistent with a recent study showing that women veterans do not have significant differences in satisfaction with prosthesis-related services.²⁸ However, this study focused specifically on lower limb amputations, while the respondents of this study include those with both upper and lower limb amputations. Importantly, our findings that women are less likely to be satisfied with prosthesis training and problem discussions support recent qualitative findings in which women expressed a desire to work with prosthetists who listen to them, take their concerns seriously, and seek solutions that fit their needs. We did not observe a difference in QOL ratings between men and women in the sample despite lower satisfaction among women with some elements of prosthesis-related services. Previous research suggests many factors impact QOL after amputation, most notably time since amputation.^16,29

Limitations

This survey was deployed in a short timeline that did not allow for careful sample selection or implementing strategies to increase response rate. Additionally, the study was conducted among veterans receiving care in the VHA, and findings may not be generalizable to limb loss in other settings. Finally, the discrepancy in number of respondents who identified as men vs women made it difficult to compare differences between the 2 groups.

Conclusions

This is the largest sample of survey respondents of veterans with limb loss to date. While the findings suggest veterans are generally satisfied with prosthetic-related services overall, they also highlight several areas for improvement with services or prostheses. Given that most veterans with limb loss are men, there is a significant discrepancy between the number of women and men respondents. Additional studies with more comparable numbers of men and women have found similar ratings of satisfaction with prostheses and services.²⁸ Further research specifically focused on improving the experiences of women should focus on better characterizing their experiences and identifying how they differ from those of male veterans. For example, understanding how to engage female veterans with limb loss in prosthesis training and problem discussions may improve their experience with their care teams and improve their use of prostheses. Understanding experiences and needs that are specific to women could lead to the development of processes, resources, or devices that are tailored to the unique requirements of women with limb loss.

Atherosclerotic cardiovascular disease (ASCVD) is a significant cause of morbidity and mortality in the United States. ASCVD involves the buildup of cholesterol plaque in arteries and includes acute coronary syndrome, peripheral arterial disease, and events such as myocardial infarction and stroke.¹ Cardiovascular disease (CVD) risk factors include high cholesterol levels, elevated blood pressure, insulin resistance, elevated blood glucose levels, smoking, poor dietary habits, and a sedentary lifestyle.²

According to the Centers for Disease Control and Prevention, about 86 million adults aged ≥ 20 years have total cholesterol levels > 200 mg/dL. More than half (54.5%) who could benefit are currently taking cholesterol-lowering medications.³ Controlling high cholesterol in American adults, especially veterans, is essential for reducing CVD morbidity and mortality.

The 2018 American College of Cardiology/American Heart Association (ACC/AHA) guideline recommends a low-density lipoprotein cholesterol (LDL-C) target goal of < 70 mg/dL for patients at high risk for ASCVD. Very high-risk ASCVD includes a history of multiple major ASCVD events or 1 major ASCVD event and multiple high-risk conditions (eg, age ≥ 65 years, smoking, or diabetes).⁴ Major ASCVD events include recent acute coronary syndrome (within the past 12 months), a history of myocardial infarction or ischemic stroke, and symptomatic peripheral artery disease. 

The ACC/AHA guideline suggests that if the LDL-C level remains ≥ 70 mg/dL, adding ezetimibe (a dietary cholesterol absorption inhibitor) to maximally tolerated statin therapy is reasonable. If LDL-C levels remain ≥ 70 mg/dL, adding a proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitor, such as alirocumab, is reasonable.⁴ The US Departments of Veterans Affairs/US Department of Defense guidelines recommend using maximally tolerated statins and ezetimibe before PCSK9 inhibitors due to established long-term safety and reduction in CVD events. 

Generic statins and ezetimibe are administered orally and widely available. In contrast, PCSK9 inhibitors have unknown long-term safety profiles, require subcutaneous injection once or twice monthly, and are significantly more expensive. They also require patient education on proper use while providing comparable or lesser relative risk reductions.²

These 3 classes of medication vary in their mechanisms of action to reduce LDL.^5,6 Ezetimibe and several statin medications are included on the Veterans Affairs Sioux Falls Health Care System (VASFHCS) formulary and do not require review prior to prescribing. Alirocumab is available at VASFHCS but is restricted to patients with a history of ASCVD or a diagnosis of familial hypercholesterolemia, and who are receiving maximally tolerated statin and ezetimibe therapy but require further LDL-C lowering to reduce their ASCVD risk. 

Studies have found ezetimibe monotherapy reduces LDL-C in patients with dyslipidemia by 18% after 12 weeks.⁷ One found that the percentage reduction in LDL-C was significantly greater (P < .001) with all doses of ezetimibe plus simvastatin (46% to 59%) compared with either atorvastatin 10 mg (37%) or simvastatin 20 mg (38%) monotherapy after 6 weeks.⁸

Although alirocumab can be added to other lipid therapies, most VASFHCS patients are prescribed alirocumab monotherapy. In the ODYSSEY CHOICE II study, patients were randomly assigned to receive either a placebo or alirocumab 150 mg every 4 weeks or alirocumab 75 mg every 2 weeks. The primary efficacy endpoint was LDL-C percentage change from baseline to week 24. In the alirocumab 150 mg every 4 weeks and 75 mg every 2 weeks groups, the least-squares mean LDL-C changes from baseline to week 24 were 51.7% and 53.5%, respectively, compared to a 4.7% increase in the placebo group (both groups P < .001 vs placebo). The authors also reported that alirocumab 150 mg every 4 weeks as monotherapy demonstrated a 47.4% reduction in LDL-C levels from baseline in a phase 1 study.⁹Although alirocumab monotherapy and ezetimibe plus statin therapy have been shown to effectively decrease LDL-C independently, a direct comparison of alirocumab monotherapy vs ezetimibe plus statin therapy has not been assessed, to our knowledge. Understanding the differences in effectiveness and safety between these 2 regimens will be valuable for clinicians when selecting a medication regimen for veterans with a history of ASCVD.

METHODS

This retrospective, single-center chart review used VASFHCS Computerized Patient Record System (CPRS) and Joint Longitudinal Viewer (JLV) records to compare patients with a history of ASCVD events who were treated with alirocumab monotherapy or ezetimibe plus statin. The 2 groups were randomized in a 1:3 ratio. The primary endpoint was achieving LDL-C < 70 mg/dL after 4 to 12 weeks, 13 to 24 weeks, and 25 to 52 weeks. Secondary endpoints included the mean percentage change from baseline in total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), LDL-C, and triglycerides (TG) over 52 weeks. The incidence of ASCVD events during this period was also assessed. If LDL-C < 70 mg/dL was achieved > 1 time during each time frame, only 1 incident was counted for analysis. Safety was assessed based on the incidence of any adverse event (AE) that led to treatment discontinuation.

Patients were identified by screening the prescription fill history between October 1, 2019, and December 31, 2022. The 52-week data collection period was counted from the first available fill date. Additionally, the prior authorization drug request file from January 1, 2017, to December 31, 2022, was used to obtain a list of patients prescribed alirocumab. Patients were included if they were veterans aged ≥ 18 years and had a history of an ASCVD event, had a alirocumab monotherapy or ezetimibe plus statin prescription between October 1, 2019, and December 31, 2022, or had an approved prior authorization drug request for alirocumab between January 1, 2017, and December 31, 2022. Patients missing a baseline or follow-up lipid panel and those with concurrent use of alirocumab and ezetimibe and/or statin were excluded. 

Baseline characteristics collected for patients included age, sex, race, weight, body mass index, lipid parameters (LDL-C, TC, HDL-C, and TG), dosing of each type of statin before adding ezetimibe, and use of any other antihyperlipidemic medication. We also collected histories of hypertension, hyperlipidemia, diabetes, chronic kidney disease, congestive heart failure, and smoking or tobacco use status. The baseline lipid panel was the most recent lipid panel documented before starting alirocumab or ezetimibe plus statin therapy. Follow-up lipid panel values were gathered at 4 to 12 weeks, 13 to 24 weeks, and 25 to 52 weeks following initiation of either therapy.

High-, moderate-, and low-intensity dosing of statin therapy and alirocumab dosing (75 mg every 2 weeks, 150 mg every 2 weeks, or 300 mg every 4 weeks) were recorded at the specified intervals. However, no patients in this study received the latter dosing regimen. ASCVD events and safety endpoints were recorded based on a review of clinical notes over the 52 weeks following the first available start date.

Statistical Analysis

The primary endpoint of achieving the LDL-C < 70 mg/dL goal from baseline to 4 to 12 weeks, 13 to 24 weeks, and 25 to 52 weeks after initiation was compared between alirocumab monotherapy and ezetimibe plus statin therapy using the χ² test. Mean percentage change from baseline in TC, HDL-C, LDL-C, and TG were compared using the independent t test. P < .05 was considered statistically significant. Incidence of ASCVD events and the safety endpoint (incidence of AEs leading to treatment discontinuation) were also compared using the χ² test. Continuous baseline characteristics were reported mean (SD) and nominal baseline characteristics were reported as a percentage.

RESULTS

There were 80 participants in this study: 20 in the alirocumab monotherapy group and 60 in the ezetimibe plus statin therapy group. More than 100 patients did not meet the prespecified inclusion criteria and were excluded. Mean (SD) age was 75 (8) years in the alirocumab group and 74 (8) years in the ezetimibe plus statin group. There was no significant differences in mean (SD) weight or mean (SD) body mass index. All study participants identified as White and male except for 2 patients in the ezetimibe plus statin therapy group whose race was not documented. Differences in lipid parameters were observed between groups, with mean baseline LDL-C, HDL-C, and TC higher in the alirocumab monotherapy group than in the ezetimibe plus statin therapy group, with significant differences in LDL-C and TC (Table 1).

Fourteen patients (70%) in the alirocumab monotherapy group had hypertension, compared with 31 (52%) in the ezetimibe plus statin therapy group. In both groups, most patients had previously been diagnosed with hyperlipidemia. More patients (60%) in the alirocumab group had diabetes than in the ezetimibe plus statin therapy group (37%). The alirocumab monotherapy group also had a higher percentage of patients with diagnoses of congestive heart failure and used other antihyperlipidemic medications than in the ezetimibe plus statin therapy group. Five patients (25%) in the alirocumab monotherapy group and 12 patients (20%) in the ezetimibe plus statin therapy group took fish oil. In the ezetimibe plus statin therapy group, 2 patients (3%) took gemfibrozil, and 2 patients (3%) took fenofibrate. Six (30%) patients in the alirocumab monotherapy group and 12 (20%) patients in the ezetimibe plus statin therapy group had chronic kidney disease. Although the majority of patients in each group did not use tobacco products, there were more tobacco users in the ezetimibe plus statin therapy group.

In the alirocumab monotherapy group, 15 patients (75%) were prescribed 75 mg every 2 weeks and 5 patients (25%) were prescribed 150 mg every 2 weeks. In the ezetimibe plus statin therapy group, 59 patients (98%) were prescribed ezetimibe 10 mg/d (Table 2). Forty-three patients (72%) were prescribed a high-intensity statin 10 received moderate-intensity (17%) and 7 received low-intensity statin (12%). Most patients were prescribed rosuvastatin (45%), followed by atorvastatin (42%), pravastatin (10%), and simvastatin (3%).

Primary Endpoint

During the 52-week study, more patients met the LDL-C goal of < 70 mg/dL in the alirocumab monotherapy group (70%) than in the ezetimibe plus statin therapy group (57%); however, the difference was not significant (P = .29). Of the patients prescribed alirocumab monotherapy who achieved LDL-C < 70 mg/dL, 15% achieved this goal in 4 to 12 weeks, 40% in 13 to 24 weeks, and 45% in 25 to 52 weeks. In the ezetimibe plus statin therapy group, 28% of patients achieved LDL-C < 70 mg/dL in 4 to 12 weeks, 31% in 13 to 24 weeks, and 41% in 25 to 52 weeks (Table 3).

Secondary Endpoints

During weeks 4 to 52 of treatment, the mean percentage change decreased in LDL-C (37.7% vs 21.4%; P = .01), TC (24.7% vs 12.5%; P = .01), and TG (0.9% vs 7.0%; P = .28) in the alirocumab monotherapy group and the ezetimibe plus statin therapy group, respectively (Table 4). The mean percentage change increased in HDL-C by 3.6% in the alirocumab monotherapy group and 1.8% in the ezetimibe plus statin therapy group (P = .36). During the study, ASCVD events occurred in 1 patient (5%) in the alirocumab monotherapy group and 3 patients (5%) in the ezetimibe plus statin therapy group (P = .99). The patient in the alirocumab monotherapy group had unstable angina 1 month after taking alirocumab. One patient in the ezetimibe plus statin therapy group had coronary artery disease and 2 patients had coronary heart disease that required stents during the 52-week period. There was 1 patient in each group who reported an AE that led to treatment discontinuation (P = .41). One patient stopped alirocumab after a trial of 2 months due to intolerance, but no specific AE was reported in the CPRS. In the ezetimibe plus statin therapy group, 1 patient requested to discontinue ezetimibe after a trial of 3 months without a specific reason noted in the medical record.

DISCUSSION

This study found no statistically significant difference in the incidence of reaching an LDL-C goal of < 70 mg/dL after alirocumab monotherapy initiation compared with ezetimibe plus statin therapy. This occurred despite baseline LDL-C being lower in the ezetimibe plus statin therapy group, which required a smaller reduction in LDL-C to reach the primary goal. Most patients on alirocumab monotherapy were prescribed a lower initial dose of 75 mg every 2 weeks. Of those patients, 30% did not achieve the LDL-C goal < 70 mg/dL. Thus, a higher dose may have led to more patients achieving the LDL-C goal.

Secondary endpoints, including mean percentage change in HDL-C and TG and incidence of ASCVD events during 52 weeks of treatment, were not statistically significant. The mean percentage increase in HDL-C was negligible in both groups, while the mean percentage reduction in TG favored the ezetimibe plus statin therapy group. In the ezetimibe plus statin therapy group, patients who also took fenofibrate experienced a significant reduction in TG while none of the patients in the alirocumab group were prescribed fenofibrate. Although the alirocumab monotherapy group had a statistically significant greater reduction in LDL-C and TC compared with those prescribed ezetimibe plus statin, the mean baseline LDL-C and TC were significantly greater in the alirocumab monotherapy group, which could contribute to higher reductions in LDL-C and TC after alirocumab monotherapy.Based on the available literature, we expected greater reductions in LDL-C in both study groups compared with statin therapy alone.^8,9 However, it was unclear whether the LDL-C and TC reductions were clinically significant.

Limitations

The study design did not permit randomization prior to the treatments, restricting our ability to account for some confounding factors, such as diet, exercise, other antihyperlipidemic medication, and medication adherence, which may have affected LDL-C, HDL-C, TG, and TC levels. Differences in baseline characteristics—particularly major risk factors, such as hypertension, diabetes, and tobacco use—also could have confounding affect on lipid levels and ASCVD events. Additionally, patients prescribed alirocumab monotherapy may have switched from statin or ezetimibe therapy, and the washout period was not reviewed or recorded, which could have affected the lipid panel results.

The small sample size of this study also may have limited the ability to detect significant differences between groups. A direct comparison of alirocumab monotherapy vs ezetimibe plus statin therapy has not been performed, making it difficult to prospectively evaluate what sample size would be needed to power this study. A posthoc analysis was used to calculate power, which was found to be only 17%. Many patients were excluded due to a lack of laboratory results within the study period, contributing to the small sample size. 

Another limitation was the reliance on documentation in CPRS and JLV. For example, having documentation of the specific AEs for the 2 patients who discontinued alirocumab or ezetimibe could have helped determine the severity of the AEs. Several patients were followed by non-VA clinicians, which could have contributed to limited documentation in the CPRS and JLV. It is difficult to draw any conclusions regarding ASCVD events and AEs that led to treatment discontinuation between alirocumab monotherapy and ezetimibe plus statin therapy based on the results of this retrospective study due to the limited number of events within the 52-week period.

CONCLUSIONS

This study found that there was no statistically significant difference in LDL-C reduction to < 70 mg/dL between alirocumab monotherapy and ezetimibe plus statin therapy in a small population of veterans with ASCVD, with a higher percentage of participants in both groups achieving that goal in 25 to 52 weeks. There also was no significant difference in percentage change in HDL-C or TG or in incidence of ASCVD events and AEs leading to treatment discontinuation. However, there was a statistically significant difference in percentage reduction for LDL-C and TC during 52 weeks of alirocumab monotherapy vs ezetimibe plus statin therapy.

Although there was no significant difference in LDL-C reduction to < 70 mg/dL, targeting this goal in patients with ASCVD is still clinically warranted. This study does not support a change in current VA criteria for use of alirocumab or a change in current guidelines for secondary prevention of ASCVD. Still, this study does indicate that the efficacy of alirocumab monotherapy is similar to that of ezetimibe plus statin therapy in patients with a history of ASCVD and may be useful in clinical settings when an alternative to ezetimibe plus statin therapy is needed. Alirocumab also may be more effective in lowering LDL-C and TC than ezetimibe plus statin therapy in veterans with ASCVD and could be added to statin therapy or ezetimibe when additional LDL-C or TC reduction is needed.

Atherosclerotic cardiovascular disease (ASCVD) is a significant cause of morbidity and mortality in the United States. ASCVD involves the buildup of cholesterol plaque in arteries and includes acute coronary syndrome, peripheral arterial disease, and events such as myocardial infarction and stroke.¹ Cardiovascular disease (CVD) risk factors include high cholesterol levels, elevated blood pressure, insulin resistance, elevated blood glucose levels, smoking, poor dietary habits, and a sedentary lifestyle.²

According to the Centers for Disease Control and Prevention, about 86 million adults aged ≥ 20 years have total cholesterol levels > 200 mg/dL. More than half (54.5%) who could benefit are currently taking cholesterol-lowering medications.³ Controlling high cholesterol in American adults, especially veterans, is essential for reducing CVD morbidity and mortality.

The 2018 American College of Cardiology/American Heart Association (ACC/AHA) guideline recommends a low-density lipoprotein cholesterol (LDL-C) target goal of < 70 mg/dL for patients at high risk for ASCVD. Very high-risk ASCVD includes a history of multiple major ASCVD events or 1 major ASCVD event and multiple high-risk conditions (eg, age ≥ 65 years, smoking, or diabetes).⁴ Major ASCVD events include recent acute coronary syndrome (within the past 12 months), a history of myocardial infarction or ischemic stroke, and symptomatic peripheral artery disease. 

The ACC/AHA guideline suggests that if the LDL-C level remains ≥ 70 mg/dL, adding ezetimibe (a dietary cholesterol absorption inhibitor) to maximally tolerated statin therapy is reasonable. If LDL-C levels remain ≥ 70 mg/dL, adding a proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitor, such as alirocumab, is reasonable.⁴ The US Departments of Veterans Affairs/US Department of Defense guidelines recommend using maximally tolerated statins and ezetimibe before PCSK9 inhibitors due to established long-term safety and reduction in CVD events. 

Generic statins and ezetimibe are administered orally and widely available. In contrast, PCSK9 inhibitors have unknown long-term safety profiles, require subcutaneous injection once or twice monthly, and are significantly more expensive. They also require patient education on proper use while providing comparable or lesser relative risk reductions.²

These 3 classes of medication vary in their mechanisms of action to reduce LDL.^5,6 Ezetimibe and several statin medications are included on the Veterans Affairs Sioux Falls Health Care System (VASFHCS) formulary and do not require review prior to prescribing. Alirocumab is available at VASFHCS but is restricted to patients with a history of ASCVD or a diagnosis of familial hypercholesterolemia, and who are receiving maximally tolerated statin and ezetimibe therapy but require further LDL-C lowering to reduce their ASCVD risk. 

Studies have found ezetimibe monotherapy reduces LDL-C in patients with dyslipidemia by 18% after 12 weeks.⁷ One found that the percentage reduction in LDL-C was significantly greater (P < .001) with all doses of ezetimibe plus simvastatin (46% to 59%) compared with either atorvastatin 10 mg (37%) or simvastatin 20 mg (38%) monotherapy after 6 weeks.⁸

Although alirocumab can be added to other lipid therapies, most VASFHCS patients are prescribed alirocumab monotherapy. In the ODYSSEY CHOICE II study, patients were randomly assigned to receive either a placebo or alirocumab 150 mg every 4 weeks or alirocumab 75 mg every 2 weeks. The primary efficacy endpoint was LDL-C percentage change from baseline to week 24. In the alirocumab 150 mg every 4 weeks and 75 mg every 2 weeks groups, the least-squares mean LDL-C changes from baseline to week 24 were 51.7% and 53.5%, respectively, compared to a 4.7% increase in the placebo group (both groups P < .001 vs placebo). The authors also reported that alirocumab 150 mg every 4 weeks as monotherapy demonstrated a 47.4% reduction in LDL-C levels from baseline in a phase 1 study.⁹Although alirocumab monotherapy and ezetimibe plus statin therapy have been shown to effectively decrease LDL-C independently, a direct comparison of alirocumab monotherapy vs ezetimibe plus statin therapy has not been assessed, to our knowledge. Understanding the differences in effectiveness and safety between these 2 regimens will be valuable for clinicians when selecting a medication regimen for veterans with a history of ASCVD.

METHODS

This retrospective, single-center chart review used VASFHCS Computerized Patient Record System (CPRS) and Joint Longitudinal Viewer (JLV) records to compare patients with a history of ASCVD events who were treated with alirocumab monotherapy or ezetimibe plus statin. The 2 groups were randomized in a 1:3 ratio. The primary endpoint was achieving LDL-C < 70 mg/dL after 4 to 12 weeks, 13 to 24 weeks, and 25 to 52 weeks. Secondary endpoints included the mean percentage change from baseline in total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), LDL-C, and triglycerides (TG) over 52 weeks. The incidence of ASCVD events during this period was also assessed. If LDL-C < 70 mg/dL was achieved > 1 time during each time frame, only 1 incident was counted for analysis. Safety was assessed based on the incidence of any adverse event (AE) that led to treatment discontinuation.

Patients were identified by screening the prescription fill history between October 1, 2019, and December 31, 2022. The 52-week data collection period was counted from the first available fill date. Additionally, the prior authorization drug request file from January 1, 2017, to December 31, 2022, was used to obtain a list of patients prescribed alirocumab. Patients were included if they were veterans aged ≥ 18 years and had a history of an ASCVD event, had a alirocumab monotherapy or ezetimibe plus statin prescription between October 1, 2019, and December 31, 2022, or had an approved prior authorization drug request for alirocumab between January 1, 2017, and December 31, 2022. Patients missing a baseline or follow-up lipid panel and those with concurrent use of alirocumab and ezetimibe and/or statin were excluded. 

Baseline characteristics collected for patients included age, sex, race, weight, body mass index, lipid parameters (LDL-C, TC, HDL-C, and TG), dosing of each type of statin before adding ezetimibe, and use of any other antihyperlipidemic medication. We also collected histories of hypertension, hyperlipidemia, diabetes, chronic kidney disease, congestive heart failure, and smoking or tobacco use status. The baseline lipid panel was the most recent lipid panel documented before starting alirocumab or ezetimibe plus statin therapy. Follow-up lipid panel values were gathered at 4 to 12 weeks, 13 to 24 weeks, and 25 to 52 weeks following initiation of either therapy.

High-, moderate-, and low-intensity dosing of statin therapy and alirocumab dosing (75 mg every 2 weeks, 150 mg every 2 weeks, or 300 mg every 4 weeks) were recorded at the specified intervals. However, no patients in this study received the latter dosing regimen. ASCVD events and safety endpoints were recorded based on a review of clinical notes over the 52 weeks following the first available start date.

Statistical Analysis

The primary endpoint of achieving the LDL-C < 70 mg/dL goal from baseline to 4 to 12 weeks, 13 to 24 weeks, and 25 to 52 weeks after initiation was compared between alirocumab monotherapy and ezetimibe plus statin therapy using the χ² test. Mean percentage change from baseline in TC, HDL-C, LDL-C, and TG were compared using the independent t test. P < .05 was considered statistically significant. Incidence of ASCVD events and the safety endpoint (incidence of AEs leading to treatment discontinuation) were also compared using the χ² test. Continuous baseline characteristics were reported mean (SD) and nominal baseline characteristics were reported as a percentage.

RESULTS

There were 80 participants in this study: 20 in the alirocumab monotherapy group and 60 in the ezetimibe plus statin therapy group. More than 100 patients did not meet the prespecified inclusion criteria and were excluded. Mean (SD) age was 75 (8) years in the alirocumab group and 74 (8) years in the ezetimibe plus statin group. There was no significant differences in mean (SD) weight or mean (SD) body mass index. All study participants identified as White and male except for 2 patients in the ezetimibe plus statin therapy group whose race was not documented. Differences in lipid parameters were observed between groups, with mean baseline LDL-C, HDL-C, and TC higher in the alirocumab monotherapy group than in the ezetimibe plus statin therapy group, with significant differences in LDL-C and TC (Table 1).

Fourteen patients (70%) in the alirocumab monotherapy group had hypertension, compared with 31 (52%) in the ezetimibe plus statin therapy group. In both groups, most patients had previously been diagnosed with hyperlipidemia. More patients (60%) in the alirocumab group had diabetes than in the ezetimibe plus statin therapy group (37%). The alirocumab monotherapy group also had a higher percentage of patients with diagnoses of congestive heart failure and used other antihyperlipidemic medications than in the ezetimibe plus statin therapy group. Five patients (25%) in the alirocumab monotherapy group and 12 patients (20%) in the ezetimibe plus statin therapy group took fish oil. In the ezetimibe plus statin therapy group, 2 patients (3%) took gemfibrozil, and 2 patients (3%) took fenofibrate. Six (30%) patients in the alirocumab monotherapy group and 12 (20%) patients in the ezetimibe plus statin therapy group had chronic kidney disease. Although the majority of patients in each group did not use tobacco products, there were more tobacco users in the ezetimibe plus statin therapy group.

In the alirocumab monotherapy group, 15 patients (75%) were prescribed 75 mg every 2 weeks and 5 patients (25%) were prescribed 150 mg every 2 weeks. In the ezetimibe plus statin therapy group, 59 patients (98%) were prescribed ezetimibe 10 mg/d (Table 2). Forty-three patients (72%) were prescribed a high-intensity statin 10 received moderate-intensity (17%) and 7 received low-intensity statin (12%). Most patients were prescribed rosuvastatin (45%), followed by atorvastatin (42%), pravastatin (10%), and simvastatin (3%).

Primary Endpoint

During the 52-week study, more patients met the LDL-C goal of < 70 mg/dL in the alirocumab monotherapy group (70%) than in the ezetimibe plus statin therapy group (57%); however, the difference was not significant (P = .29). Of the patients prescribed alirocumab monotherapy who achieved LDL-C < 70 mg/dL, 15% achieved this goal in 4 to 12 weeks, 40% in 13 to 24 weeks, and 45% in 25 to 52 weeks. In the ezetimibe plus statin therapy group, 28% of patients achieved LDL-C < 70 mg/dL in 4 to 12 weeks, 31% in 13 to 24 weeks, and 41% in 25 to 52 weeks (Table 3).

Secondary Endpoints

During weeks 4 to 52 of treatment, the mean percentage change decreased in LDL-C (37.7% vs 21.4%; P = .01), TC (24.7% vs 12.5%; P = .01), and TG (0.9% vs 7.0%; P = .28) in the alirocumab monotherapy group and the ezetimibe plus statin therapy group, respectively (Table 4). The mean percentage change increased in HDL-C by 3.6% in the alirocumab monotherapy group and 1.8% in the ezetimibe plus statin therapy group (P = .36). During the study, ASCVD events occurred in 1 patient (5%) in the alirocumab monotherapy group and 3 patients (5%) in the ezetimibe plus statin therapy group (P = .99). The patient in the alirocumab monotherapy group had unstable angina 1 month after taking alirocumab. One patient in the ezetimibe plus statin therapy group had coronary artery disease and 2 patients had coronary heart disease that required stents during the 52-week period. There was 1 patient in each group who reported an AE that led to treatment discontinuation (P = .41). One patient stopped alirocumab after a trial of 2 months due to intolerance, but no specific AE was reported in the CPRS. In the ezetimibe plus statin therapy group, 1 patient requested to discontinue ezetimibe after a trial of 3 months without a specific reason noted in the medical record.

DISCUSSION

This study found no statistically significant difference in the incidence of reaching an LDL-C goal of < 70 mg/dL after alirocumab monotherapy initiation compared with ezetimibe plus statin therapy. This occurred despite baseline LDL-C being lower in the ezetimibe plus statin therapy group, which required a smaller reduction in LDL-C to reach the primary goal. Most patients on alirocumab monotherapy were prescribed a lower initial dose of 75 mg every 2 weeks. Of those patients, 30% did not achieve the LDL-C goal < 70 mg/dL. Thus, a higher dose may have led to more patients achieving the LDL-C goal.

Secondary endpoints, including mean percentage change in HDL-C and TG and incidence of ASCVD events during 52 weeks of treatment, were not statistically significant. The mean percentage increase in HDL-C was negligible in both groups, while the mean percentage reduction in TG favored the ezetimibe plus statin therapy group. In the ezetimibe plus statin therapy group, patients who also took fenofibrate experienced a significant reduction in TG while none of the patients in the alirocumab group were prescribed fenofibrate. Although the alirocumab monotherapy group had a statistically significant greater reduction in LDL-C and TC compared with those prescribed ezetimibe plus statin, the mean baseline LDL-C and TC were significantly greater in the alirocumab monotherapy group, which could contribute to higher reductions in LDL-C and TC after alirocumab monotherapy.Based on the available literature, we expected greater reductions in LDL-C in both study groups compared with statin therapy alone.^8,9 However, it was unclear whether the LDL-C and TC reductions were clinically significant.

Limitations

The study design did not permit randomization prior to the treatments, restricting our ability to account for some confounding factors, such as diet, exercise, other antihyperlipidemic medication, and medication adherence, which may have affected LDL-C, HDL-C, TG, and TC levels. Differences in baseline characteristics—particularly major risk factors, such as hypertension, diabetes, and tobacco use—also could have confounding affect on lipid levels and ASCVD events. Additionally, patients prescribed alirocumab monotherapy may have switched from statin or ezetimibe therapy, and the washout period was not reviewed or recorded, which could have affected the lipid panel results.

The small sample size of this study also may have limited the ability to detect significant differences between groups. A direct comparison of alirocumab monotherapy vs ezetimibe plus statin therapy has not been performed, making it difficult to prospectively evaluate what sample size would be needed to power this study. A posthoc analysis was used to calculate power, which was found to be only 17%. Many patients were excluded due to a lack of laboratory results within the study period, contributing to the small sample size. 

Another limitation was the reliance on documentation in CPRS and JLV. For example, having documentation of the specific AEs for the 2 patients who discontinued alirocumab or ezetimibe could have helped determine the severity of the AEs. Several patients were followed by non-VA clinicians, which could have contributed to limited documentation in the CPRS and JLV. It is difficult to draw any conclusions regarding ASCVD events and AEs that led to treatment discontinuation between alirocumab monotherapy and ezetimibe plus statin therapy based on the results of this retrospective study due to the limited number of events within the 52-week period.

CONCLUSIONS

This study found that there was no statistically significant difference in LDL-C reduction to < 70 mg/dL between alirocumab monotherapy and ezetimibe plus statin therapy in a small population of veterans with ASCVD, with a higher percentage of participants in both groups achieving that goal in 25 to 52 weeks. There also was no significant difference in percentage change in HDL-C or TG or in incidence of ASCVD events and AEs leading to treatment discontinuation. However, there was a statistically significant difference in percentage reduction for LDL-C and TC during 52 weeks of alirocumab monotherapy vs ezetimibe plus statin therapy.

Although there was no significant difference in LDL-C reduction to < 70 mg/dL, targeting this goal in patients with ASCVD is still clinically warranted. This study does not support a change in current VA criteria for use of alirocumab or a change in current guidelines for secondary prevention of ASCVD. Still, this study does indicate that the efficacy of alirocumab monotherapy is similar to that of ezetimibe plus statin therapy in patients with a history of ASCVD and may be useful in clinical settings when an alternative to ezetimibe plus statin therapy is needed. Alirocumab also may be more effective in lowering LDL-C and TC than ezetimibe plus statin therapy in veterans with ASCVD and could be added to statin therapy or ezetimibe when additional LDL-C or TC reduction is needed.

Atherosclerotic cardiovascular disease (ASCVD) is a significant cause of morbidity and mortality in the United States. ASCVD involves the buildup of cholesterol plaque in arteries and includes acute coronary syndrome, peripheral arterial disease, and events such as myocardial infarction and stroke.¹ Cardiovascular disease (CVD) risk factors include high cholesterol levels, elevated blood pressure, insulin resistance, elevated blood glucose levels, smoking, poor dietary habits, and a sedentary lifestyle.²

According to the Centers for Disease Control and Prevention, about 86 million adults aged ≥ 20 years have total cholesterol levels > 200 mg/dL. More than half (54.5%) who could benefit are currently taking cholesterol-lowering medications.³ Controlling high cholesterol in American adults, especially veterans, is essential for reducing CVD morbidity and mortality.

The 2018 American College of Cardiology/American Heart Association (ACC/AHA) guideline recommends a low-density lipoprotein cholesterol (LDL-C) target goal of < 70 mg/dL for patients at high risk for ASCVD. Very high-risk ASCVD includes a history of multiple major ASCVD events or 1 major ASCVD event and multiple high-risk conditions (eg, age ≥ 65 years, smoking, or diabetes).⁴ Major ASCVD events include recent acute coronary syndrome (within the past 12 months), a history of myocardial infarction or ischemic stroke, and symptomatic peripheral artery disease. 

The ACC/AHA guideline suggests that if the LDL-C level remains ≥ 70 mg/dL, adding ezetimibe (a dietary cholesterol absorption inhibitor) to maximally tolerated statin therapy is reasonable. If LDL-C levels remain ≥ 70 mg/dL, adding a proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitor, such as alirocumab, is reasonable.⁴ The US Departments of Veterans Affairs/US Department of Defense guidelines recommend using maximally tolerated statins and ezetimibe before PCSK9 inhibitors due to established long-term safety and reduction in CVD events. 

Generic statins and ezetimibe are administered orally and widely available. In contrast, PCSK9 inhibitors have unknown long-term safety profiles, require subcutaneous injection once or twice monthly, and are significantly more expensive. They also require patient education on proper use while providing comparable or lesser relative risk reductions.²

These 3 classes of medication vary in their mechanisms of action to reduce LDL.^5,6 Ezetimibe and several statin medications are included on the Veterans Affairs Sioux Falls Health Care System (VASFHCS) formulary and do not require review prior to prescribing. Alirocumab is available at VASFHCS but is restricted to patients with a history of ASCVD or a diagnosis of familial hypercholesterolemia, and who are receiving maximally tolerated statin and ezetimibe therapy but require further LDL-C lowering to reduce their ASCVD risk. 

Studies have found ezetimibe monotherapy reduces LDL-C in patients with dyslipidemia by 18% after 12 weeks.⁷ One found that the percentage reduction in LDL-C was significantly greater (P < .001) with all doses of ezetimibe plus simvastatin (46% to 59%) compared with either atorvastatin 10 mg (37%) or simvastatin 20 mg (38%) monotherapy after 6 weeks.⁸

Although alirocumab can be added to other lipid therapies, most VASFHCS patients are prescribed alirocumab monotherapy. In the ODYSSEY CHOICE II study, patients were randomly assigned to receive either a placebo or alirocumab 150 mg every 4 weeks or alirocumab 75 mg every 2 weeks. The primary efficacy endpoint was LDL-C percentage change from baseline to week 24. In the alirocumab 150 mg every 4 weeks and 75 mg every 2 weeks groups, the least-squares mean LDL-C changes from baseline to week 24 were 51.7% and 53.5%, respectively, compared to a 4.7% increase in the placebo group (both groups P < .001 vs placebo). The authors also reported that alirocumab 150 mg every 4 weeks as monotherapy demonstrated a 47.4% reduction in LDL-C levels from baseline in a phase 1 study.⁹Although alirocumab monotherapy and ezetimibe plus statin therapy have been shown to effectively decrease LDL-C independently, a direct comparison of alirocumab monotherapy vs ezetimibe plus statin therapy has not been assessed, to our knowledge. Understanding the differences in effectiveness and safety between these 2 regimens will be valuable for clinicians when selecting a medication regimen for veterans with a history of ASCVD.

METHODS

This retrospective, single-center chart review used VASFHCS Computerized Patient Record System (CPRS) and Joint Longitudinal Viewer (JLV) records to compare patients with a history of ASCVD events who were treated with alirocumab monotherapy or ezetimibe plus statin. The 2 groups were randomized in a 1:3 ratio. The primary endpoint was achieving LDL-C < 70 mg/dL after 4 to 12 weeks, 13 to 24 weeks, and 25 to 52 weeks. Secondary endpoints included the mean percentage change from baseline in total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), LDL-C, and triglycerides (TG) over 52 weeks. The incidence of ASCVD events during this period was also assessed. If LDL-C < 70 mg/dL was achieved > 1 time during each time frame, only 1 incident was counted for analysis. Safety was assessed based on the incidence of any adverse event (AE) that led to treatment discontinuation.

Patients were identified by screening the prescription fill history between October 1, 2019, and December 31, 2022. The 52-week data collection period was counted from the first available fill date. Additionally, the prior authorization drug request file from January 1, 2017, to December 31, 2022, was used to obtain a list of patients prescribed alirocumab. Patients were included if they were veterans aged ≥ 18 years and had a history of an ASCVD event, had a alirocumab monotherapy or ezetimibe plus statin prescription between October 1, 2019, and December 31, 2022, or had an approved prior authorization drug request for alirocumab between January 1, 2017, and December 31, 2022. Patients missing a baseline or follow-up lipid panel and those with concurrent use of alirocumab and ezetimibe and/or statin were excluded. 

Baseline characteristics collected for patients included age, sex, race, weight, body mass index, lipid parameters (LDL-C, TC, HDL-C, and TG), dosing of each type of statin before adding ezetimibe, and use of any other antihyperlipidemic medication. We also collected histories of hypertension, hyperlipidemia, diabetes, chronic kidney disease, congestive heart failure, and smoking or tobacco use status. The baseline lipid panel was the most recent lipid panel documented before starting alirocumab or ezetimibe plus statin therapy. Follow-up lipid panel values were gathered at 4 to 12 weeks, 13 to 24 weeks, and 25 to 52 weeks following initiation of either therapy.

High-, moderate-, and low-intensity dosing of statin therapy and alirocumab dosing (75 mg every 2 weeks, 150 mg every 2 weeks, or 300 mg every 4 weeks) were recorded at the specified intervals. However, no patients in this study received the latter dosing regimen. ASCVD events and safety endpoints were recorded based on a review of clinical notes over the 52 weeks following the first available start date.

Statistical Analysis

The primary endpoint of achieving the LDL-C < 70 mg/dL goal from baseline to 4 to 12 weeks, 13 to 24 weeks, and 25 to 52 weeks after initiation was compared between alirocumab monotherapy and ezetimibe plus statin therapy using the χ² test. Mean percentage change from baseline in TC, HDL-C, LDL-C, and TG were compared using the independent t test. P < .05 was considered statistically significant. Incidence of ASCVD events and the safety endpoint (incidence of AEs leading to treatment discontinuation) were also compared using the χ² test. Continuous baseline characteristics were reported mean (SD) and nominal baseline characteristics were reported as a percentage.

RESULTS

There were 80 participants in this study: 20 in the alirocumab monotherapy group and 60 in the ezetimibe plus statin therapy group. More than 100 patients did not meet the prespecified inclusion criteria and were excluded. Mean (SD) age was 75 (8) years in the alirocumab group and 74 (8) years in the ezetimibe plus statin group. There was no significant differences in mean (SD) weight or mean (SD) body mass index. All study participants identified as White and male except for 2 patients in the ezetimibe plus statin therapy group whose race was not documented. Differences in lipid parameters were observed between groups, with mean baseline LDL-C, HDL-C, and TC higher in the alirocumab monotherapy group than in the ezetimibe plus statin therapy group, with significant differences in LDL-C and TC (Table 1).

Fourteen patients (70%) in the alirocumab monotherapy group had hypertension, compared with 31 (52%) in the ezetimibe plus statin therapy group. In both groups, most patients had previously been diagnosed with hyperlipidemia. More patients (60%) in the alirocumab group had diabetes than in the ezetimibe plus statin therapy group (37%). The alirocumab monotherapy group also had a higher percentage of patients with diagnoses of congestive heart failure and used other antihyperlipidemic medications than in the ezetimibe plus statin therapy group. Five patients (25%) in the alirocumab monotherapy group and 12 patients (20%) in the ezetimibe plus statin therapy group took fish oil. In the ezetimibe plus statin therapy group, 2 patients (3%) took gemfibrozil, and 2 patients (3%) took fenofibrate. Six (30%) patients in the alirocumab monotherapy group and 12 (20%) patients in the ezetimibe plus statin therapy group had chronic kidney disease. Although the majority of patients in each group did not use tobacco products, there were more tobacco users in the ezetimibe plus statin therapy group.

In the alirocumab monotherapy group, 15 patients (75%) were prescribed 75 mg every 2 weeks and 5 patients (25%) were prescribed 150 mg every 2 weeks. In the ezetimibe plus statin therapy group, 59 patients (98%) were prescribed ezetimibe 10 mg/d (Table 2). Forty-three patients (72%) were prescribed a high-intensity statin 10 received moderate-intensity (17%) and 7 received low-intensity statin (12%). Most patients were prescribed rosuvastatin (45%), followed by atorvastatin (42%), pravastatin (10%), and simvastatin (3%).

Primary Endpoint

During the 52-week study, more patients met the LDL-C goal of < 70 mg/dL in the alirocumab monotherapy group (70%) than in the ezetimibe plus statin therapy group (57%); however, the difference was not significant (P = .29). Of the patients prescribed alirocumab monotherapy who achieved LDL-C < 70 mg/dL, 15% achieved this goal in 4 to 12 weeks, 40% in 13 to 24 weeks, and 45% in 25 to 52 weeks. In the ezetimibe plus statin therapy group, 28% of patients achieved LDL-C < 70 mg/dL in 4 to 12 weeks, 31% in 13 to 24 weeks, and 41% in 25 to 52 weeks (Table 3).

Secondary Endpoints

During weeks 4 to 52 of treatment, the mean percentage change decreased in LDL-C (37.7% vs 21.4%; P = .01), TC (24.7% vs 12.5%; P = .01), and TG (0.9% vs 7.0%; P = .28) in the alirocumab monotherapy group and the ezetimibe plus statin therapy group, respectively (Table 4). The mean percentage change increased in HDL-C by 3.6% in the alirocumab monotherapy group and 1.8% in the ezetimibe plus statin therapy group (P = .36). During the study, ASCVD events occurred in 1 patient (5%) in the alirocumab monotherapy group and 3 patients (5%) in the ezetimibe plus statin therapy group (P = .99). The patient in the alirocumab monotherapy group had unstable angina 1 month after taking alirocumab. One patient in the ezetimibe plus statin therapy group had coronary artery disease and 2 patients had coronary heart disease that required stents during the 52-week period. There was 1 patient in each group who reported an AE that led to treatment discontinuation (P = .41). One patient stopped alirocumab after a trial of 2 months due to intolerance, but no specific AE was reported in the CPRS. In the ezetimibe plus statin therapy group, 1 patient requested to discontinue ezetimibe after a trial of 3 months without a specific reason noted in the medical record.

DISCUSSION

This study found no statistically significant difference in the incidence of reaching an LDL-C goal of < 70 mg/dL after alirocumab monotherapy initiation compared with ezetimibe plus statin therapy. This occurred despite baseline LDL-C being lower in the ezetimibe plus statin therapy group, which required a smaller reduction in LDL-C to reach the primary goal. Most patients on alirocumab monotherapy were prescribed a lower initial dose of 75 mg every 2 weeks. Of those patients, 30% did not achieve the LDL-C goal < 70 mg/dL. Thus, a higher dose may have led to more patients achieving the LDL-C goal.

Secondary endpoints, including mean percentage change in HDL-C and TG and incidence of ASCVD events during 52 weeks of treatment, were not statistically significant. The mean percentage increase in HDL-C was negligible in both groups, while the mean percentage reduction in TG favored the ezetimibe plus statin therapy group. In the ezetimibe plus statin therapy group, patients who also took fenofibrate experienced a significant reduction in TG while none of the patients in the alirocumab group were prescribed fenofibrate. Although the alirocumab monotherapy group had a statistically significant greater reduction in LDL-C and TC compared with those prescribed ezetimibe plus statin, the mean baseline LDL-C and TC were significantly greater in the alirocumab monotherapy group, which could contribute to higher reductions in LDL-C and TC after alirocumab monotherapy.Based on the available literature, we expected greater reductions in LDL-C in both study groups compared with statin therapy alone.^8,9 However, it was unclear whether the LDL-C and TC reductions were clinically significant.

Limitations

The study design did not permit randomization prior to the treatments, restricting our ability to account for some confounding factors, such as diet, exercise, other antihyperlipidemic medication, and medication adherence, which may have affected LDL-C, HDL-C, TG, and TC levels. Differences in baseline characteristics—particularly major risk factors, such as hypertension, diabetes, and tobacco use—also could have confounding affect on lipid levels and ASCVD events. Additionally, patients prescribed alirocumab monotherapy may have switched from statin or ezetimibe therapy, and the washout period was not reviewed or recorded, which could have affected the lipid panel results.

The small sample size of this study also may have limited the ability to detect significant differences between groups. A direct comparison of alirocumab monotherapy vs ezetimibe plus statin therapy has not been performed, making it difficult to prospectively evaluate what sample size would be needed to power this study. A posthoc analysis was used to calculate power, which was found to be only 17%. Many patients were excluded due to a lack of laboratory results within the study period, contributing to the small sample size. 

Another limitation was the reliance on documentation in CPRS and JLV. For example, having documentation of the specific AEs for the 2 patients who discontinued alirocumab or ezetimibe could have helped determine the severity of the AEs. Several patients were followed by non-VA clinicians, which could have contributed to limited documentation in the CPRS and JLV. It is difficult to draw any conclusions regarding ASCVD events and AEs that led to treatment discontinuation between alirocumab monotherapy and ezetimibe plus statin therapy based on the results of this retrospective study due to the limited number of events within the 52-week period.

CONCLUSIONS

This study found that there was no statistically significant difference in LDL-C reduction to < 70 mg/dL between alirocumab monotherapy and ezetimibe plus statin therapy in a small population of veterans with ASCVD, with a higher percentage of participants in both groups achieving that goal in 25 to 52 weeks. There also was no significant difference in percentage change in HDL-C or TG or in incidence of ASCVD events and AEs leading to treatment discontinuation. However, there was a statistically significant difference in percentage reduction for LDL-C and TC during 52 weeks of alirocumab monotherapy vs ezetimibe plus statin therapy.

Although there was no significant difference in LDL-C reduction to < 70 mg/dL, targeting this goal in patients with ASCVD is still clinically warranted. This study does not support a change in current VA criteria for use of alirocumab or a change in current guidelines for secondary prevention of ASCVD. Still, this study does indicate that the efficacy of alirocumab monotherapy is similar to that of ezetimibe plus statin therapy in patients with a history of ASCVD and may be useful in clinical settings when an alternative to ezetimibe plus statin therapy is needed. Alirocumab also may be more effective in lowering LDL-C and TC than ezetimibe plus statin therapy in veterans with ASCVD and could be added to statin therapy or ezetimibe when additional LDL-C or TC reduction is needed.

Individuals with diabetes are at risk for developing foot ulcers or full-thickness defects in the epithelium of the foot. These defects can lead to bacterial invasion and foot infection, potentially resulting in leg amputation (Figure 1). Effective treatment to prevent leg amputation, known as limb salvage, requires management across multiple medical specialties including podiatry, vascular surgery, and infectious diseases. The multidisciplinary team approach to limb salvage was introduced in Boston in 1928 and has been the prevailing approach to this cross-specialty medical problem for at least a decade.^1,2

The Michael E. DeBakey Veterans Affairs Medical Center (MEDVAMC) has established an inpatient limb salvage program—a group of dedicated clinicians working collaboratively to provide evidence-guided management of patients hospitalized with foot ulcers, foot gangrene or any superimposed infection with the goal of avoiding leg amputations. We have seen a significant and durable reduction in the incidence of leg amputations among veterans at MEDVAMC.

This article describes the evolution and outcomes of the MEDVAMC limb salvage program over more than a decade. It includes changes to team structure and workflow, as well as past and present successes and challenges. The eAppendix provides a narrative summary with examples of how our clinical practice and research efforts have informed one another and how these findings are applied to clinical management. This process is part of the larger efforts of the Veterans Health Administration (VHA) to create a learning health system in which “internal data and experience are systematically integrated with external evidence, and that knowledge is put into practice.”³

Methods

Data from the VHA Support Service Center were used to obtain monthly major (leg) and minor (toe and partial foot) amputation records at MEDVAMC from October 2000 through May 2023. Yearly totals for the number of persons with diabetes and foot ulcers at MEDVAMC were also obtained from the support service center. Annual patient population sizes and number of persons with foot ulcers were converted to monthly estimates using cubic spline interpolation. Rates were calculated as 12-month rolling averages. Trend lines were created with locally weighted running line smoothing that used a span α of 0.1.

We characterized the patient population using data from cohorts of veterans treated for foot ulcers and foot infections at MEDVAMC. To compare the contemporary veteran population with nonveteran inpatients treated for foot ulcers and foot infections at other hospitals, we created a 2:1 nonveteran to veteran cohort matched by sex and zip code, using publicly available hospital admission data from the Texas Department of Health and State Health Services. Veterans used for this cohort comparison are consistent with the 100 consecutive patients who underwent angiography for limb salvage in 2022.

This research was approved by the Baylor College of Medicine Institutional Review Board (protocol H-34858) and the MEDVAMC Research Committee (IRBNet protocol 15A12. HB). All analyses used deidentified data in the R programming language version 4.2.2 using RStudio version 2022.06.0 Build 421.

Program Description

MEDVAMC is a 350-bed teaching hospital located in central Houston. Its hospital system includes 11 outpatient clinics, ranging from 28 to 126 miles (eAppendix, Supplemental Figure A) from MEDVAMC. MEDVAMC provides vascular, orthopedic, and podiatric surgery services, as well as many other highly specialized services such as liver and heart transplants. The hospital’s risk-adjusted rates of operative morbidity and mortality (observed-to-expected ratios) are significantly lower than expected.

Despite this, the incidence rate of leg amputations at MEDVAMC in early 2011 was nearly 3-times higher than the VHA average. The inpatient management of veterans with infected foot ulcers was fragmented, with the general, orthopedic, and vascular surgery teams separately providing siloed care. Delays in treatment were common. There was much service- and practitioner-level practice heterogeneity. No diagnostic or treatment protocols were used, and standard treatment components were sporadically provided.

Patient Population

Compared to the matched non-VHA patient cohort (Supplemental Table 1), veterans treated at MEDVAMC for limb salvage are older. Nearly half (46%) identify as Black, which is associated with a 2-fold higher riskadjusted rate of leg amputations.⁴ MEDVAMC patients also have significantly higher rates of diabetes, chronic kidney disease, and systolic heart failure. About 22% travel > 40 miles for treatment at MEDVAMC, double that of the matched cohort (10.7%). Additionally, 35% currently smoke and 37% have moderate to severe peripheral artery disease (PAD).⁵

Program Design

In late 2011, the MEDVAMC vascular surgery team led limb salvage efforts by implementing a single team model, which involved assuming the primary role of managing foot ulcers for all veterans, both infected and uninfected (eAppendix, Supplemental Figure B). Consultations were directed to a dedicated limb salvage pager. The vascular team provided interdisciplinary limb salvage management across the spectrum of disease, including the surgical treatment of infection, assessment for PAD, open surgical operations and endovascular interventions to treat PAD, and foot reconstruction (debridement, minor or partial foot amputations, and skin grafting). This care was complemented by frequent consultation with the infectious disease, vascular medicine, podiatry, and geriatric wound care teams. This approach streamlined the delivery of consistent multidisciplinary care.

This collaborative effort aimed to develop ideal multidisciplinary care plans through research spanning the spectrum of the diabetic foot infection disease process (eAppendix, Supplemental Table 1). Some of the most impactful practices were: (1) a proclivity towards surgical treatment of foot infections, especially osteomyelitis⁵; (2) improved identification of PAD^6,7; (3) early surgical closure of foot wounds following revascularization^8,9; and (4) palliative wound care as an alternative to leg amputation in veterans who are not candidates for revascularization and limb salvage.¹⁰ Initally, the vascular surgery team held monthly multidisciplinary limb salvage meetings to coordinate patient management, identify ways to streamline care and avoid waste, discuss research findings, and review the 12-month rolling average of the MEDVAMC leg amputation incidence rate.

During the study period, the MEDVAMC vascular surgery team consisted of 2 to 5 board certified vascular or general surgeons, 2 or 3 nurse practitioners, and 3 vascular ultrasound technologists. Associated specialists included 2 podiatrists, 3 geriatricians with wound care certification, as well as additional infectious diseases, vascular medicine, orthopedics, and general surgery specialists.

Program Assessment

We noted a significant and sustained decrease in the MEDVAMC leg amputation rate after implementing multidisciplinary meetings and a single- team model from early 2012 through 2017 (Figure 2). The amputation incidence rate decreased steadily over the period from a maximum of 160 per 100,000 per year in February 2012 to a nadir of 66 per 100,000 per year in April 2017, an overall 60% decrease. Increases were noted in early 2018 after ceasing the single- team model, and in the summer of 2022, following periods of bed shortages after the onset of the COVID-19 pandemic. Tracking this metric allowed clinicians to make course corrections.

The decreased leg amputation rate at MEDVAMC does not seem to be mirroring national or regional trends. During this 10-year period, the VHA annualized amputation rate decreased minimally, from 58 to 54 per 100,000 (eAppendix Supplemental Figure C). Leg amputation incidence at non-VHA hospitals in Texas slightly increased over the same period.¹¹

Value was also reflected in other metrics. MEDVAMC improved safety through a bundled strategy that reduced the risk-adjusted rate of surgical wound infections by 95%.¹² MEDVAMC prioritized limb salvage when selecting patients for angiography and nearly eliminated using stent-grafts, cryopreserved allogeneic saphenous vein grafts, and expensive surgical and endovascular implants, which were identified as more expensive and less effective than other options (Figure 3).^13-15 The MEDVAMC team achieved a > 90% patient trust rating on the Veterans Signals survey in fiscal years 2021 and 2022.

Challenges

A significant increase in the patient-physician ratio occurred 5 years into the program. In 2016, 2 vascular surgeons left MEDVAMC and a planned renovation of 1 of the 2 vascular surgery-assigned hybrid working facilities began even as the number of MEDVAMC patients with diabetes grew 120% (from 89,400 to 107,746 between 2010 and 2016), and the incidence rate of foot ulcers grew 300% (from 392 in 2010 to 1183 in 2016 per 100,000). The net result was a higher clinical workload among the remaining vascular surgeons with less operating room availability.

To stabilize surgeon retention, MEDVAMC reverted from the single team model back to inpatient care being distributed among general surgery, orthopedic surgery, and vascular surgery. After noting an increase in the leg amputation incidence rate, we adjusted the focus from multidisciplinary to interdisciplinary care (ie, majority of limb salvage clinical care can be provided by practitioners of any involved specialties). We worked to establish a local, written, interdisciplinary consensus on evaluating and managing veterans with nonhealing foot ulcers to mitigate the loss of a consolidated inpatient approach. Despite frequent staff turnover, ≥ 1 physician or surgeon from the core specialties of vascular surgery, podiatry, and infectious diseases remained throughout the study period.

The COVID-19 pandemic caused a shortage of hospital beds. This was followed by more bed shortages due to decreased nursing staff. Our health care system also had a period of restricted outpatient encounters early in the pandemic. During this time, we noted a delayed presentation of veterans with advanced infections and another increase in leg amputation incidence rate.

Like many health systems, MEDVAMC pivoted to telephone- and video-based outpatient encounters. Our team also used publicly available Texas hospitalization data to identify zip codes with particularly high leg amputation incidence rates, and > 3500 educational mailings to veterans categorized as moderate and high risk for leg amputation in these zip codes. These mailings provided information on recognizing foot ulcers and infections, emphasized timely evaluation, and named the MEDVAMC vascular surgery team as a point-of-contact. More recently, we have seen a further decrease in the MEDVAMC incidences of leg amputation to its lowest rate in > 20 years.

Discussion

A learning organization that directs its research based on clinical observations and informs its clinical care with research findings can produce palpable improvements in outcomes. Understanding the disease process and trying to better understand management across the entire range of this disease process has allowed our team to make consistent and systematic changes in care (Table). Consolidating inpatient care in a single team model seems to have been effective in reducing amputation rates among veterans with diabetes. The role the MEDVAMC vascular surgery team served for limb salvage patients may have been particularly beneficial because of the large impact untreated or unidentified PAD can have and because of the high prevalence of PAD among the limb salvage population seen at MEDVAMC. To be sustainable, though, a single-team model needs resources. A multiteam model can also be effective if the degree of multidisciplinary involvement for any given veteran is appropriate to the individual's clinical needs, teams are engaged and willing to contribute in a defined role within their specialty, and lines of communication remain open.

The primary challenge at MEDVAMC has been, and will continue to be, the retention of physicians and surgeons. MEDVAMC has excellent leadership and a collegial working environment, but better access to operating rooms for elective and time-sensitive operations, additional clinical staff support, and higher salary at non-VA positions have been the basis for many of physicians— especially surgeons—leaving MEDVAMC. Despite high staff turnover and a constant flow of resident and fellow trainees, MEDVAMC has been able to keep the clinical approach relatively consistent due to the use of written protocols and continuity of care as ≥ 1 physician or surgeon from each of the 4 main teams remained engaged with limb salvage throughout the entire period.

Going forward, we will work to ensure that all requirements of the 2022 Prevention of Amputation in Veterans Everywhere directive are incorporated into care.⁸ We plan to standardize MEDVAMC management algorithms further, both to streamline care and reduce the opportunity for disparities in treatment. More prophylactic podiatric procedures, surgical forms of offloading, and a shared multidisciplinary clinic space may also further help patients.

Conclusions

The introduction of multidisciplinary limb salvage at MEDVAMC has led to significant and sustained reductions in leg amputation incidence. These reductions do not seem dependent upon a specific team structure for inpatient care. To improve patient outcomes, efforts should focus on making improvements across the entire disease spectrum. For limb salvage, this includes primary prevention of foot ulcers, the treatment of foot infections, identification and management of PAD, surgical reconstruction/optimal wound healing, and care for patients who undergo leg amputation.

Individuals with diabetes are at risk for developing foot ulcers or full-thickness defects in the epithelium of the foot. These defects can lead to bacterial invasion and foot infection, potentially resulting in leg amputation (Figure 1). Effective treatment to prevent leg amputation, known as limb salvage, requires management across multiple medical specialties including podiatry, vascular surgery, and infectious diseases. The multidisciplinary team approach to limb salvage was introduced in Boston in 1928 and has been the prevailing approach to this cross-specialty medical problem for at least a decade.^1,2

The Michael E. DeBakey Veterans Affairs Medical Center (MEDVAMC) has established an inpatient limb salvage program—a group of dedicated clinicians working collaboratively to provide evidence-guided management of patients hospitalized with foot ulcers, foot gangrene or any superimposed infection with the goal of avoiding leg amputations. We have seen a significant and durable reduction in the incidence of leg amputations among veterans at MEDVAMC.

This article describes the evolution and outcomes of the MEDVAMC limb salvage program over more than a decade. It includes changes to team structure and workflow, as well as past and present successes and challenges. The eAppendix provides a narrative summary with examples of how our clinical practice and research efforts have informed one another and how these findings are applied to clinical management. This process is part of the larger efforts of the Veterans Health Administration (VHA) to create a learning health system in which “internal data and experience are systematically integrated with external evidence, and that knowledge is put into practice.”³

Methods

Data from the VHA Support Service Center were used to obtain monthly major (leg) and minor (toe and partial foot) amputation records at MEDVAMC from October 2000 through May 2023. Yearly totals for the number of persons with diabetes and foot ulcers at MEDVAMC were also obtained from the support service center. Annual patient population sizes and number of persons with foot ulcers were converted to monthly estimates using cubic spline interpolation. Rates were calculated as 12-month rolling averages. Trend lines were created with locally weighted running line smoothing that used a span α of 0.1.

We characterized the patient population using data from cohorts of veterans treated for foot ulcers and foot infections at MEDVAMC. To compare the contemporary veteran population with nonveteran inpatients treated for foot ulcers and foot infections at other hospitals, we created a 2:1 nonveteran to veteran cohort matched by sex and zip code, using publicly available hospital admission data from the Texas Department of Health and State Health Services. Veterans used for this cohort comparison are consistent with the 100 consecutive patients who underwent angiography for limb salvage in 2022.

This research was approved by the Baylor College of Medicine Institutional Review Board (protocol H-34858) and the MEDVAMC Research Committee (IRBNet protocol 15A12. HB). All analyses used deidentified data in the R programming language version 4.2.2 using RStudio version 2022.06.0 Build 421.

Program Description

MEDVAMC is a 350-bed teaching hospital located in central Houston. Its hospital system includes 11 outpatient clinics, ranging from 28 to 126 miles (eAppendix, Supplemental Figure A) from MEDVAMC. MEDVAMC provides vascular, orthopedic, and podiatric surgery services, as well as many other highly specialized services such as liver and heart transplants. The hospital’s risk-adjusted rates of operative morbidity and mortality (observed-to-expected ratios) are significantly lower than expected.

Despite this, the incidence rate of leg amputations at MEDVAMC in early 2011 was nearly 3-times higher than the VHA average. The inpatient management of veterans with infected foot ulcers was fragmented, with the general, orthopedic, and vascular surgery teams separately providing siloed care. Delays in treatment were common. There was much service- and practitioner-level practice heterogeneity. No diagnostic or treatment protocols were used, and standard treatment components were sporadically provided.

Patient Population

Compared to the matched non-VHA patient cohort (Supplemental Table 1), veterans treated at MEDVAMC for limb salvage are older. Nearly half (46%) identify as Black, which is associated with a 2-fold higher riskadjusted rate of leg amputations.⁴ MEDVAMC patients also have significantly higher rates of diabetes, chronic kidney disease, and systolic heart failure. About 22% travel > 40 miles for treatment at MEDVAMC, double that of the matched cohort (10.7%). Additionally, 35% currently smoke and 37% have moderate to severe peripheral artery disease (PAD).⁵

Program Design

In late 2011, the MEDVAMC vascular surgery team led limb salvage efforts by implementing a single team model, which involved assuming the primary role of managing foot ulcers for all veterans, both infected and uninfected (eAppendix, Supplemental Figure B). Consultations were directed to a dedicated limb salvage pager. The vascular team provided interdisciplinary limb salvage management across the spectrum of disease, including the surgical treatment of infection, assessment for PAD, open surgical operations and endovascular interventions to treat PAD, and foot reconstruction (debridement, minor or partial foot amputations, and skin grafting). This care was complemented by frequent consultation with the infectious disease, vascular medicine, podiatry, and geriatric wound care teams. This approach streamlined the delivery of consistent multidisciplinary care.

This collaborative effort aimed to develop ideal multidisciplinary care plans through research spanning the spectrum of the diabetic foot infection disease process (eAppendix, Supplemental Table 1). Some of the most impactful practices were: (1) a proclivity towards surgical treatment of foot infections, especially osteomyelitis⁵; (2) improved identification of PAD^6,7; (3) early surgical closure of foot wounds following revascularization^8,9; and (4) palliative wound care as an alternative to leg amputation in veterans who are not candidates for revascularization and limb salvage.¹⁰ Initally, the vascular surgery team held monthly multidisciplinary limb salvage meetings to coordinate patient management, identify ways to streamline care and avoid waste, discuss research findings, and review the 12-month rolling average of the MEDVAMC leg amputation incidence rate.

During the study period, the MEDVAMC vascular surgery team consisted of 2 to 5 board certified vascular or general surgeons, 2 or 3 nurse practitioners, and 3 vascular ultrasound technologists. Associated specialists included 2 podiatrists, 3 geriatricians with wound care certification, as well as additional infectious diseases, vascular medicine, orthopedics, and general surgery specialists.

Program Assessment

We noted a significant and sustained decrease in the MEDVAMC leg amputation rate after implementing multidisciplinary meetings and a single- team model from early 2012 through 2017 (Figure 2). The amputation incidence rate decreased steadily over the period from a maximum of 160 per 100,000 per year in February 2012 to a nadir of 66 per 100,000 per year in April 2017, an overall 60% decrease. Increases were noted in early 2018 after ceasing the single- team model, and in the summer of 2022, following periods of bed shortages after the onset of the COVID-19 pandemic. Tracking this metric allowed clinicians to make course corrections.

The decreased leg amputation rate at MEDVAMC does not seem to be mirroring national or regional trends. During this 10-year period, the VHA annualized amputation rate decreased minimally, from 58 to 54 per 100,000 (eAppendix Supplemental Figure C). Leg amputation incidence at non-VHA hospitals in Texas slightly increased over the same period.¹¹

Value was also reflected in other metrics. MEDVAMC improved safety through a bundled strategy that reduced the risk-adjusted rate of surgical wound infections by 95%.¹² MEDVAMC prioritized limb salvage when selecting patients for angiography and nearly eliminated using stent-grafts, cryopreserved allogeneic saphenous vein grafts, and expensive surgical and endovascular implants, which were identified as more expensive and less effective than other options (Figure 3).^13-15 The MEDVAMC team achieved a > 90% patient trust rating on the Veterans Signals survey in fiscal years 2021 and 2022.

Challenges

A significant increase in the patient-physician ratio occurred 5 years into the program. In 2016, 2 vascular surgeons left MEDVAMC and a planned renovation of 1 of the 2 vascular surgery-assigned hybrid working facilities began even as the number of MEDVAMC patients with diabetes grew 120% (from 89,400 to 107,746 between 2010 and 2016), and the incidence rate of foot ulcers grew 300% (from 392 in 2010 to 1183 in 2016 per 100,000). The net result was a higher clinical workload among the remaining vascular surgeons with less operating room availability.

To stabilize surgeon retention, MEDVAMC reverted from the single team model back to inpatient care being distributed among general surgery, orthopedic surgery, and vascular surgery. After noting an increase in the leg amputation incidence rate, we adjusted the focus from multidisciplinary to interdisciplinary care (ie, majority of limb salvage clinical care can be provided by practitioners of any involved specialties). We worked to establish a local, written, interdisciplinary consensus on evaluating and managing veterans with nonhealing foot ulcers to mitigate the loss of a consolidated inpatient approach. Despite frequent staff turnover, ≥ 1 physician or surgeon from the core specialties of vascular surgery, podiatry, and infectious diseases remained throughout the study period.

The COVID-19 pandemic caused a shortage of hospital beds. This was followed by more bed shortages due to decreased nursing staff. Our health care system also had a period of restricted outpatient encounters early in the pandemic. During this time, we noted a delayed presentation of veterans with advanced infections and another increase in leg amputation incidence rate.

Like many health systems, MEDVAMC pivoted to telephone- and video-based outpatient encounters. Our team also used publicly available Texas hospitalization data to identify zip codes with particularly high leg amputation incidence rates, and > 3500 educational mailings to veterans categorized as moderate and high risk for leg amputation in these zip codes. These mailings provided information on recognizing foot ulcers and infections, emphasized timely evaluation, and named the MEDVAMC vascular surgery team as a point-of-contact. More recently, we have seen a further decrease in the MEDVAMC incidences of leg amputation to its lowest rate in > 20 years.

Discussion

A learning organization that directs its research based on clinical observations and informs its clinical care with research findings can produce palpable improvements in outcomes. Understanding the disease process and trying to better understand management across the entire range of this disease process has allowed our team to make consistent and systematic changes in care (Table). Consolidating inpatient care in a single team model seems to have been effective in reducing amputation rates among veterans with diabetes. The role the MEDVAMC vascular surgery team served for limb salvage patients may have been particularly beneficial because of the large impact untreated or unidentified PAD can have and because of the high prevalence of PAD among the limb salvage population seen at MEDVAMC. To be sustainable, though, a single-team model needs resources. A multiteam model can also be effective if the degree of multidisciplinary involvement for any given veteran is appropriate to the individual's clinical needs, teams are engaged and willing to contribute in a defined role within their specialty, and lines of communication remain open.

The primary challenge at MEDVAMC has been, and will continue to be, the retention of physicians and surgeons. MEDVAMC has excellent leadership and a collegial working environment, but better access to operating rooms for elective and time-sensitive operations, additional clinical staff support, and higher salary at non-VA positions have been the basis for many of physicians— especially surgeons—leaving MEDVAMC. Despite high staff turnover and a constant flow of resident and fellow trainees, MEDVAMC has been able to keep the clinical approach relatively consistent due to the use of written protocols and continuity of care as ≥ 1 physician or surgeon from each of the 4 main teams remained engaged with limb salvage throughout the entire period.

Going forward, we will work to ensure that all requirements of the 2022 Prevention of Amputation in Veterans Everywhere directive are incorporated into care.⁸ We plan to standardize MEDVAMC management algorithms further, both to streamline care and reduce the opportunity for disparities in treatment. More prophylactic podiatric procedures, surgical forms of offloading, and a shared multidisciplinary clinic space may also further help patients.

Conclusions

The introduction of multidisciplinary limb salvage at MEDVAMC has led to significant and sustained reductions in leg amputation incidence. These reductions do not seem dependent upon a specific team structure for inpatient care. To improve patient outcomes, efforts should focus on making improvements across the entire disease spectrum. For limb salvage, this includes primary prevention of foot ulcers, the treatment of foot infections, identification and management of PAD, surgical reconstruction/optimal wound healing, and care for patients who undergo leg amputation.

Individuals with diabetes are at risk for developing foot ulcers or full-thickness defects in the epithelium of the foot. These defects can lead to bacterial invasion and foot infection, potentially resulting in leg amputation (Figure 1). Effective treatment to prevent leg amputation, known as limb salvage, requires management across multiple medical specialties including podiatry, vascular surgery, and infectious diseases. The multidisciplinary team approach to limb salvage was introduced in Boston in 1928 and has been the prevailing approach to this cross-specialty medical problem for at least a decade.^1,2

The Michael E. DeBakey Veterans Affairs Medical Center (MEDVAMC) has established an inpatient limb salvage program—a group of dedicated clinicians working collaboratively to provide evidence-guided management of patients hospitalized with foot ulcers, foot gangrene or any superimposed infection with the goal of avoiding leg amputations. We have seen a significant and durable reduction in the incidence of leg amputations among veterans at MEDVAMC.

This article describes the evolution and outcomes of the MEDVAMC limb salvage program over more than a decade. It includes changes to team structure and workflow, as well as past and present successes and challenges. The eAppendix provides a narrative summary with examples of how our clinical practice and research efforts have informed one another and how these findings are applied to clinical management. This process is part of the larger efforts of the Veterans Health Administration (VHA) to create a learning health system in which “internal data and experience are systematically integrated with external evidence, and that knowledge is put into practice.”³

Methods

Data from the VHA Support Service Center were used to obtain monthly major (leg) and minor (toe and partial foot) amputation records at MEDVAMC from October 2000 through May 2023. Yearly totals for the number of persons with diabetes and foot ulcers at MEDVAMC were also obtained from the support service center. Annual patient population sizes and number of persons with foot ulcers were converted to monthly estimates using cubic spline interpolation. Rates were calculated as 12-month rolling averages. Trend lines were created with locally weighted running line smoothing that used a span α of 0.1.

We characterized the patient population using data from cohorts of veterans treated for foot ulcers and foot infections at MEDVAMC. To compare the contemporary veteran population with nonveteran inpatients treated for foot ulcers and foot infections at other hospitals, we created a 2:1 nonveteran to veteran cohort matched by sex and zip code, using publicly available hospital admission data from the Texas Department of Health and State Health Services. Veterans used for this cohort comparison are consistent with the 100 consecutive patients who underwent angiography for limb salvage in 2022.

This research was approved by the Baylor College of Medicine Institutional Review Board (protocol H-34858) and the MEDVAMC Research Committee (IRBNet protocol 15A12. HB). All analyses used deidentified data in the R programming language version 4.2.2 using RStudio version 2022.06.0 Build 421.

Program Description

MEDVAMC is a 350-bed teaching hospital located in central Houston. Its hospital system includes 11 outpatient clinics, ranging from 28 to 126 miles (eAppendix, Supplemental Figure A) from MEDVAMC. MEDVAMC provides vascular, orthopedic, and podiatric surgery services, as well as many other highly specialized services such as liver and heart transplants. The hospital’s risk-adjusted rates of operative morbidity and mortality (observed-to-expected ratios) are significantly lower than expected.

Despite this, the incidence rate of leg amputations at MEDVAMC in early 2011 was nearly 3-times higher than the VHA average. The inpatient management of veterans with infected foot ulcers was fragmented, with the general, orthopedic, and vascular surgery teams separately providing siloed care. Delays in treatment were common. There was much service- and practitioner-level practice heterogeneity. No diagnostic or treatment protocols were used, and standard treatment components were sporadically provided.

Patient Population

Compared to the matched non-VHA patient cohort (Supplemental Table 1), veterans treated at MEDVAMC for limb salvage are older. Nearly half (46%) identify as Black, which is associated with a 2-fold higher riskadjusted rate of leg amputations.⁴ MEDVAMC patients also have significantly higher rates of diabetes, chronic kidney disease, and systolic heart failure. About 22% travel > 40 miles for treatment at MEDVAMC, double that of the matched cohort (10.7%). Additionally, 35% currently smoke and 37% have moderate to severe peripheral artery disease (PAD).⁵

Program Design

In late 2011, the MEDVAMC vascular surgery team led limb salvage efforts by implementing a single team model, which involved assuming the primary role of managing foot ulcers for all veterans, both infected and uninfected (eAppendix, Supplemental Figure B). Consultations were directed to a dedicated limb salvage pager. The vascular team provided interdisciplinary limb salvage management across the spectrum of disease, including the surgical treatment of infection, assessment for PAD, open surgical operations and endovascular interventions to treat PAD, and foot reconstruction (debridement, minor or partial foot amputations, and skin grafting). This care was complemented by frequent consultation with the infectious disease, vascular medicine, podiatry, and geriatric wound care teams. This approach streamlined the delivery of consistent multidisciplinary care.

This collaborative effort aimed to develop ideal multidisciplinary care plans through research spanning the spectrum of the diabetic foot infection disease process (eAppendix, Supplemental Table 1). Some of the most impactful practices were: (1) a proclivity towards surgical treatment of foot infections, especially osteomyelitis⁵; (2) improved identification of PAD^6,7; (3) early surgical closure of foot wounds following revascularization^8,9; and (4) palliative wound care as an alternative to leg amputation in veterans who are not candidates for revascularization and limb salvage.¹⁰ Initally, the vascular surgery team held monthly multidisciplinary limb salvage meetings to coordinate patient management, identify ways to streamline care and avoid waste, discuss research findings, and review the 12-month rolling average of the MEDVAMC leg amputation incidence rate.

During the study period, the MEDVAMC vascular surgery team consisted of 2 to 5 board certified vascular or general surgeons, 2 or 3 nurse practitioners, and 3 vascular ultrasound technologists. Associated specialists included 2 podiatrists, 3 geriatricians with wound care certification, as well as additional infectious diseases, vascular medicine, orthopedics, and general surgery specialists.

Program Assessment

We noted a significant and sustained decrease in the MEDVAMC leg amputation rate after implementing multidisciplinary meetings and a single- team model from early 2012 through 2017 (Figure 2). The amputation incidence rate decreased steadily over the period from a maximum of 160 per 100,000 per year in February 2012 to a nadir of 66 per 100,000 per year in April 2017, an overall 60% decrease. Increases were noted in early 2018 after ceasing the single- team model, and in the summer of 2022, following periods of bed shortages after the onset of the COVID-19 pandemic. Tracking this metric allowed clinicians to make course corrections.

The decreased leg amputation rate at MEDVAMC does not seem to be mirroring national or regional trends. During this 10-year period, the VHA annualized amputation rate decreased minimally, from 58 to 54 per 100,000 (eAppendix Supplemental Figure C). Leg amputation incidence at non-VHA hospitals in Texas slightly increased over the same period.¹¹

Value was also reflected in other metrics. MEDVAMC improved safety through a bundled strategy that reduced the risk-adjusted rate of surgical wound infections by 95%.¹² MEDVAMC prioritized limb salvage when selecting patients for angiography and nearly eliminated using stent-grafts, cryopreserved allogeneic saphenous vein grafts, and expensive surgical and endovascular implants, which were identified as more expensive and less effective than other options (Figure 3).^13-15 The MEDVAMC team achieved a > 90% patient trust rating on the Veterans Signals survey in fiscal years 2021 and 2022.

Challenges

A significant increase in the patient-physician ratio occurred 5 years into the program. In 2016, 2 vascular surgeons left MEDVAMC and a planned renovation of 1 of the 2 vascular surgery-assigned hybrid working facilities began even as the number of MEDVAMC patients with diabetes grew 120% (from 89,400 to 107,746 between 2010 and 2016), and the incidence rate of foot ulcers grew 300% (from 392 in 2010 to 1183 in 2016 per 100,000). The net result was a higher clinical workload among the remaining vascular surgeons with less operating room availability.

To stabilize surgeon retention, MEDVAMC reverted from the single team model back to inpatient care being distributed among general surgery, orthopedic surgery, and vascular surgery. After noting an increase in the leg amputation incidence rate, we adjusted the focus from multidisciplinary to interdisciplinary care (ie, majority of limb salvage clinical care can be provided by practitioners of any involved specialties). We worked to establish a local, written, interdisciplinary consensus on evaluating and managing veterans with nonhealing foot ulcers to mitigate the loss of a consolidated inpatient approach. Despite frequent staff turnover, ≥ 1 physician or surgeon from the core specialties of vascular surgery, podiatry, and infectious diseases remained throughout the study period.

The COVID-19 pandemic caused a shortage of hospital beds. This was followed by more bed shortages due to decreased nursing staff. Our health care system also had a period of restricted outpatient encounters early in the pandemic. During this time, we noted a delayed presentation of veterans with advanced infections and another increase in leg amputation incidence rate.

Like many health systems, MEDVAMC pivoted to telephone- and video-based outpatient encounters. Our team also used publicly available Texas hospitalization data to identify zip codes with particularly high leg amputation incidence rates, and > 3500 educational mailings to veterans categorized as moderate and high risk for leg amputation in these zip codes. These mailings provided information on recognizing foot ulcers and infections, emphasized timely evaluation, and named the MEDVAMC vascular surgery team as a point-of-contact. More recently, we have seen a further decrease in the MEDVAMC incidences of leg amputation to its lowest rate in > 20 years.

Discussion

A learning organization that directs its research based on clinical observations and informs its clinical care with research findings can produce palpable improvements in outcomes. Understanding the disease process and trying to better understand management across the entire range of this disease process has allowed our team to make consistent and systematic changes in care (Table). Consolidating inpatient care in a single team model seems to have been effective in reducing amputation rates among veterans with diabetes. The role the MEDVAMC vascular surgery team served for limb salvage patients may have been particularly beneficial because of the large impact untreated or unidentified PAD can have and because of the high prevalence of PAD among the limb salvage population seen at MEDVAMC. To be sustainable, though, a single-team model needs resources. A multiteam model can also be effective if the degree of multidisciplinary involvement for any given veteran is appropriate to the individual's clinical needs, teams are engaged and willing to contribute in a defined role within their specialty, and lines of communication remain open.

The primary challenge at MEDVAMC has been, and will continue to be, the retention of physicians and surgeons. MEDVAMC has excellent leadership and a collegial working environment, but better access to operating rooms for elective and time-sensitive operations, additional clinical staff support, and higher salary at non-VA positions have been the basis for many of physicians— especially surgeons—leaving MEDVAMC. Despite high staff turnover and a constant flow of resident and fellow trainees, MEDVAMC has been able to keep the clinical approach relatively consistent due to the use of written protocols and continuity of care as ≥ 1 physician or surgeon from each of the 4 main teams remained engaged with limb salvage throughout the entire period.

Going forward, we will work to ensure that all requirements of the 2022 Prevention of Amputation in Veterans Everywhere directive are incorporated into care.⁸ We plan to standardize MEDVAMC management algorithms further, both to streamline care and reduce the opportunity for disparities in treatment. More prophylactic podiatric procedures, surgical forms of offloading, and a shared multidisciplinary clinic space may also further help patients.

Conclusions

The introduction of multidisciplinary limb salvage at MEDVAMC has led to significant and sustained reductions in leg amputation incidence. These reductions do not seem dependent upon a specific team structure for inpatient care. To improve patient outcomes, efforts should focus on making improvements across the entire disease spectrum. For limb salvage, this includes primary prevention of foot ulcers, the treatment of foot infections, identification and management of PAD, surgical reconstruction/optimal wound healing, and care for patients who undergo leg amputation.

Psoriasis is a chronic multisystemic inflammatory skin disease with a worldwide prevalence of 2% to 3%.¹ Psoriasis can be accompanied by other conditions such as psoriatic arthritis, obesity, metabolic syndrome, diabetes mellitus, hypertension, dyslipidemia, atherosclerotic disease, inflammatory bowel disease, and anxiety/depression. It is important to manage comorbidities of psoriasis in addition to treating the cutaneous manifestations of the disease.¹

Obesity is a major public health concern worldwide. Numerous observational and epidemiologic studies have reported a high prevalence of obesity among patients with psoriasis.² Current evidence indicates that obesity may initiate or worsen psoriasis; furthermore, it is important to note that obesity may negatively impact the effectiveness of psoriasis-specific treatments or increase the incidence of adverse effects. Therefore, managing obesity is crucial in the treatment of psoriasis.³ Numerous studies have investigated the association between psoriasis and obesity, and they commonly conclude that both conditions share the same genetic metabolic pathways.^2-4 However, it is important to consider environmental factors such as dietary habits, smoking, alcohol consumption, and a sedentary lifestyle—all of which are associated with psoriasis and also can contribute to the development of obesity.⁵ Because of the effects of obesity in psoriasis patients, factors that impact the development of obesity have become a popular research topic.

Eating disorders (EDs) are a crucial risk factor for both developing and maintaining obesity. In particular, two EDs that are associated with obesity include binge eating disorder and bulimia nervosa.⁶ According to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition,⁷ binge eating disorder can be diagnosed when a patient has at least 1 episode of binge eating per week over a 3-month period. Bulimia nervosa can be diagnosed when a patient is excessively concerned with their body weight and shape and engages in behaviors to prevent weight gain (eg, forced vomiting, excessive use of laxatives).⁷ Psychiatrists who specialize in EDs make diagnoses based on these criteria. In daily practice, there are several quick and simple questionnaires available to screen for EDs that can be used by nonpsychiatrist physicians, including the commonly used 26-item Eating Attitudes Test (EAT-26).⁸ The EAT-26 has been used to screen for EDs in patients with inflammatory disorders.⁹

The aim of this study was to screen for EDs in patients with psoriasis to identify potential risk factors for development of obesity.

Materials and Methods

This study included patients with psoriasis who were screened for EDs at a tertiary dermatology clinic in Turkey between January 2021 and December 2023. This study was approved by the local ethics committee and was in accordance with the Declaration of Helsinki (decision number E-93471371-514.99-225000079).

Study Design and Patient Inclusion Criteria—This quantitative cross-sectional study utilized EAT-26, Dermatology Life Quality Index (DLQI), Attitude Scale for Healthy Nutrition (ASHN), and Depression Anxiety Stress Scale-21 (DASS-21) scores. All the questionnaire scales used in the study were adapted and validated in Turkey.^8,10-12 The inclusion criteria consisted of being older than 18 years of age, being literate, having psoriasis for at least 1 year that was not treated topically or systemically, and having no psychiatric diseases outside an ED. The questionnaires were presented in written format following the clinical examination. Literacy was an inclusion criterion in this study due to the absence of auxiliary health personnel.

Study Variables—The study variables included age, sex, marital status (single/divorced or married), education status (primary/secondary school or high school/university), employment status (employed or unemployed/retired), body mass index (BMI), smoking status, alcohol-consumption status, Psoriasis Area Severity Index score, presence of nail psoriasis and psoriatic arthritis, duration of psoriasis, family history of psoriasis, EAT-26 score, ASHN score, DLQI score, and DASS-21 score. Body mass index was calculated by taking a participant’s weight in kilograms and dividing it by their height in meters squared. The BMI values were classified into 3 categories: normal (18.5–24.9 kg/m²), overweight (25.0–29.9 kg/m²), and obese (≥30 kg/m²).¹³

Questionnaires—The EAT-26 questionnaire includes 26 questions that are used to detect EDs. Responses to each question include Likert-type answer options (ie, “always,” “usually,” “often,” “sometimes,” “rarely,” and “never.”) Patients with scores of 20 points or higher (range, 0–78) are classified as high risk for EDs.⁸ In our study, EAT-26 scores were grouped into 2 categories: patients scoring less than 20 points and those scoring 20 points or higher.

The DLQI questionnaire includes 10 questions to measure dermatologic symptoms and qualiy of life. Responses to each question include Likert-type answer options (ie, “not at all,” “a little,” “a lot,” or “very much.”) On the DLQI scale, the higher the score, the lower the quality of life (score range, 0–30).¹⁰

The ASHN questionnaire includes 21 questions that measure attitudes toward healthy nutrition with 5 possible answer options (“strongly disagree,” “disagree,” “undecided,” “agree,” and “strongly agree”). On this scale, higher scores indicate the participant is more knowledgeable about healthy nutrition (score range, 0–78).¹¹

The DASS-21 questionnaire includes 21 questions that measure the severity of a range of symptoms common to depression, anxiety, and stress. Responses include Likert-type answer options (eg, “never,” “sometimes,” “often,” and “almost always.”) On this scale, a higher score (range of 0–21 for each) indicates higher levels of depression, anxiety, and stress.¹²

Statistical Analysis—Descriptive statistics were analyzed using SPSS software version 22.0 (IBM). The Shapiro-Wilk test was applied to determine whether the data were normally distributed. For categorical variables, frequency differences among groups were compared using the Pearson χ² test. A t test was used to compare the means of 2 independent groups with a normal distribution. One-way analysis of variance and Tukey Honest Significant Difference post hoc analysis were used to test whether there was a statistically significant difference among the normally distributed means of independent groups. Pearson correlation analysis was used to determine whether there was a linear relationship between 2 numeric measurements and, if so, to determine the direction and severity of this relationship. P<.05 indicated statistical significance in this study.

Results

Study Participant Demographics—This study included 82 participants with a mean age of 44.3 years; 52.4% (43/82) were female, and 85.4% (70/82) were married. The questionnaire took an average of 4.2 minutes for participants to complete. A total of 57.3% (47/82) of patients had completed primary/secondary education and 59.8% (49/82) were employed. The mean BMI was 28.1 kg/m². According to the BMI classification, 26.8% (22/82) participants had a normal weight, 36.6% (30/82) were overweight, and 43.9% (36/82) were obese. A total of 48.8% (40/82) of participants smoked, and 4.9% (4/82) consumed alcohol. The mean Psoriasis Area and Severity Index score was 5.4. A total of 54.9% (45/82) of participants had nail psoriasis, and 24.4% (20/82) had psoriatic arthritis. The mean duration of psoriasis was 153 months. A total of 29.3% (24/82) of participants had a positive family history of psoriasis. The mean EAT-26 score was 11.1. A total of 12.2% (10/82) of participants had an EAT-26 score of 20 points or higher and were considered at high risk for an ED. The mean ASHN score was 72.9; the mean DLQI score was 5.5; and on the DASS-21 scale, mean scores for depression, anxiety, and stress were 6.3, 8.7, and 10.0, respectively (Table).

Comparative Evaluation of the BMI Groups—The only statistically significant differences among the 3 BMI groups were related to marital status, EAT-26 score, and anxiety and stress scores (P=.02, <.01, <.01, and <.01, respectively)(eTable 1). The number of single/divorced participants in the overweight group was significantly (P=.02) greater than in the normal weight group. The mean EAT-26 score for the normal weight group was significantly (P<.01) lower than for the overweight and obese groups; there was no significant difference in mean EAT-26 scores between the overweight and obese groups. The mean anxiety score was significantly (P<.01) lower in the normal weight group compared with the overweight and obese groups. There was no significant difference between the overweight and obese groups according to the mean depression score. The mean stress and anxiety scores were significantly (P<.01) lower in the normal weight group than in the overweight and obese groups. There was no significant difference between the overweight and obese groups according to the mean anxiety score.

Comment

Eating disorders are psychiatric conditions that require a multidisciplinary approach. Nonpsychiatric medical departments may be involved due to the severe consequences (eg, various skin changes¹⁴) of these disorders. Psoriasis is not known to be directly affected by the presence of an ED; however, it is possible that EDs could indirectly affect patients with psoriasis by influencing obesity. Therefore, this study aimed to examine the relationship between ED risk factors and obesity in this population.

The relationship between psoriasis and obesity has been a popular research topic in dermatology since the 1990s.¹⁵ Epidemiologic and observational studies have reported that patients with psoriasis are more likely to be overweight or have obesity, which is an independent risk factor for psoriasis.^3,16 However, the causal relationship between psoriasis and obesity remains unclear. In a comprehensive review, Barros et al¹⁷ emphasized the causal relationship between obesity and psoriasis under several headings. Firstly, a higher BMI increases the risk for psoriasis by promoting cytokine release and immune system dysregulation. Secondly, a Western diet (eg, processed foods and fast food) triggers obesity and psoriasis by increasing adipose tissue. Thirdly, the alteration of the skin and gut microbiota triggers chronic inflammation as a result of bacterial translocation in patients with obesity. Fourthly, a high-fat diet and palmitic acid disrupt the intestinal integrity of the gut and increase the risk for psoriasis and obesity by triggering chronic inflammation of bacterial fragments that pass into the blood. Finally, the decrease in the amount of adiponectin and the increase in the amount of leptin in patients with obesity may cause psoriasis by increasing proinflammatory cytokines, which are similar to those involved in the pathogenesis of psoriasis.¹⁷ Additionally, psoriatic inflammation can cause insulin resistance and metabolic dysfunction, leading to obesity.¹⁸ The relationship between psoriasis and obesity cannot be solely explained by metabolic pathways. Smoking, alcohol consumption, and a sedentary lifestyle all are associated with psoriasis and also can contribute to obesity.⁵ Our study revealed no significant difference in smoking or alcohol consumption between the normal weight and overweight/obesity groups. Based on our data, we determined that smoking and alcohol consumption did not affect obesity in our patients with psoriasis.

Observational and epidemiologic studies have shown that patients with psoriasis experience increased rates of depression, anxiety, and stress.¹⁹ In studies of pathogenesis, a connection between depression and psoriatic inflammation has been established.²⁰ It is known that inflammatory cytokines similar to those in psoriasis are involved in the development of obesity.¹⁸ In addition, depression and anxiety can lead to binge eating, unhealthy food choices, and a more sedentary lifestyle.⁵ All of these variables may contribute to the associations between depression and anxiety with psoriasis and obesity. Zafiriou et al²¹ conducted a study to investigate the relationship between psoriasis, obesity, and depression through inflammatory pathways with a focus on the importance of IL-17. Data showing that IL-17–producing Th17-cell subgroups play a considerable role in the development of obesity and depression prompted the authors to suggest that psoriasis, obesity, and ­anxiety/depression may be interconnected manifestations of immune dysregulation, potentially linked to IL-17 and its associated cells.²¹ Mrowietz et al²² also suggested that metabolic inflammation may contribute to obesity and depression in patients with psoriasis and highlighted the importance of several cytokines, including tumor necrosis factor α, IL-6, IL-8, IL-17, and IL-23. Our study revealed no significant differences in depression scores between BMI groups. Another meta-analysis reported conflicting findings on the incidence of depression in obese patients with psoriasis.²³ Some of the studies had a small number of participants. Compared to depression, anxiety has received less attention in studies of patients with obesity with psoriasis. However, these studies have shown a positive correlation between anxiety scores and BMI in patients with psoriasis.^24,25 In our study, similar to the findings of previous studies, overweight patients and those with obesitywho have psoriasis had significantly (P<.01) greater anxiety and stress scores than did normal weight patients with psoriasis.

Obesity should be assessed in patients with psoriasis via a biopsychosocial approach that takes into account genetic, behavioral, and environmental factors.²⁶ Eating disorders are considered to be one of the factors contributing to obesity. Numerous studies in the literature have demonstrated a greater incidence of EDs in patients with obesity vs those without obesity.^5,6,27 Obesity and EDs have a bidirectional relationship: individuals with obesity are at risk for EDs due to body dissatisfaction, dieting habits, and depressive states. Conversely, poor eating behaviors in individuals with a normal weight can lead to obesity.²⁸

There are few studies in the literature exploring the relationship between psoriasis and EDs. Crosta et al²⁹ demonstrated that patients with psoriasis had impaired results on ED screening tests and that these scores deteriorated further as BMI increased. Moreover, Altunay et al³⁰ demonstrated that patients with psoriasis and metabolic syndrome had higher scores on the ED screening test. In this study, patients with higher scores also exhibited high levels of anxiety.³⁰ In our study, similar to the findings of previous studies, patients with psoriasis who were overweight or had obesity had significantly (P<.01) greater EAT-26 scores than those in the normal weight group. Patients with high EAT-26 scores also exhibited elevated levels of depression, anxiety, and stress. Additionally, EAT-26 scores were positively correlated with BMI, anxiety, depression, and stress scores. Our study as well as other studies in the literature indicate that additional research is needed to determine the associations between EDs and obesity in psoriasis.

Conclusion

Managing obesity is crucial for patients with psoriasis. This study showed that EAT-26 scores were higher in patients with psoriasis who were overweight or had obesity than in those who were normal weight. Participants with high EAT-26 scores (≥20 points) were more likely to be female and have higher anxiety and stress scores. In addition, EAT-26 scores were positively correlated with BMI as well as depression, anxiety, and stress scores. Eating disorders may contribute to the development of obesity in patients with psoriasis. Although our study was limited by a small sample size, the results suggest that there is a need for large-scale multicenter studies to investigate the relationship between psoriasis and EDs.

Psoriasis is a chronic multisystemic inflammatory skin disease with a worldwide prevalence of 2% to 3%.¹ Psoriasis can be accompanied by other conditions such as psoriatic arthritis, obesity, metabolic syndrome, diabetes mellitus, hypertension, dyslipidemia, atherosclerotic disease, inflammatory bowel disease, and anxiety/depression. It is important to manage comorbidities of psoriasis in addition to treating the cutaneous manifestations of the disease.¹

Obesity is a major public health concern worldwide. Numerous observational and epidemiologic studies have reported a high prevalence of obesity among patients with psoriasis.² Current evidence indicates that obesity may initiate or worsen psoriasis; furthermore, it is important to note that obesity may negatively impact the effectiveness of psoriasis-specific treatments or increase the incidence of adverse effects. Therefore, managing obesity is crucial in the treatment of psoriasis.³ Numerous studies have investigated the association between psoriasis and obesity, and they commonly conclude that both conditions share the same genetic metabolic pathways.^2-4 However, it is important to consider environmental factors such as dietary habits, smoking, alcohol consumption, and a sedentary lifestyle—all of which are associated with psoriasis and also can contribute to the development of obesity.⁵ Because of the effects of obesity in psoriasis patients, factors that impact the development of obesity have become a popular research topic.

Eating disorders (EDs) are a crucial risk factor for both developing and maintaining obesity. In particular, two EDs that are associated with obesity include binge eating disorder and bulimia nervosa.⁶ According to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition,⁷ binge eating disorder can be diagnosed when a patient has at least 1 episode of binge eating per week over a 3-month period. Bulimia nervosa can be diagnosed when a patient is excessively concerned with their body weight and shape and engages in behaviors to prevent weight gain (eg, forced vomiting, excessive use of laxatives).⁷ Psychiatrists who specialize in EDs make diagnoses based on these criteria. In daily practice, there are several quick and simple questionnaires available to screen for EDs that can be used by nonpsychiatrist physicians, including the commonly used 26-item Eating Attitudes Test (EAT-26).⁸ The EAT-26 has been used to screen for EDs in patients with inflammatory disorders.⁹

The aim of this study was to screen for EDs in patients with psoriasis to identify potential risk factors for development of obesity.

Materials and Methods

This study included patients with psoriasis who were screened for EDs at a tertiary dermatology clinic in Turkey between January 2021 and December 2023. This study was approved by the local ethics committee and was in accordance with the Declaration of Helsinki (decision number E-93471371-514.99-225000079).

Study Design and Patient Inclusion Criteria—This quantitative cross-sectional study utilized EAT-26, Dermatology Life Quality Index (DLQI), Attitude Scale for Healthy Nutrition (ASHN), and Depression Anxiety Stress Scale-21 (DASS-21) scores. All the questionnaire scales used in the study were adapted and validated in Turkey.^8,10-12 The inclusion criteria consisted of being older than 18 years of age, being literate, having psoriasis for at least 1 year that was not treated topically or systemically, and having no psychiatric diseases outside an ED. The questionnaires were presented in written format following the clinical examination. Literacy was an inclusion criterion in this study due to the absence of auxiliary health personnel.

Study Variables—The study variables included age, sex, marital status (single/divorced or married), education status (primary/secondary school or high school/university), employment status (employed or unemployed/retired), body mass index (BMI), smoking status, alcohol-consumption status, Psoriasis Area Severity Index score, presence of nail psoriasis and psoriatic arthritis, duration of psoriasis, family history of psoriasis, EAT-26 score, ASHN score, DLQI score, and DASS-21 score. Body mass index was calculated by taking a participant’s weight in kilograms and dividing it by their height in meters squared. The BMI values were classified into 3 categories: normal (18.5–24.9 kg/m²), overweight (25.0–29.9 kg/m²), and obese (≥30 kg/m²).¹³

Questionnaires—The EAT-26 questionnaire includes 26 questions that are used to detect EDs. Responses to each question include Likert-type answer options (ie, “always,” “usually,” “often,” “sometimes,” “rarely,” and “never.”) Patients with scores of 20 points or higher (range, 0–78) are classified as high risk for EDs.⁸ In our study, EAT-26 scores were grouped into 2 categories: patients scoring less than 20 points and those scoring 20 points or higher.

The DLQI questionnaire includes 10 questions to measure dermatologic symptoms and qualiy of life. Responses to each question include Likert-type answer options (ie, “not at all,” “a little,” “a lot,” or “very much.”) On the DLQI scale, the higher the score, the lower the quality of life (score range, 0–30).¹⁰

The ASHN questionnaire includes 21 questions that measure attitudes toward healthy nutrition with 5 possible answer options (“strongly disagree,” “disagree,” “undecided,” “agree,” and “strongly agree”). On this scale, higher scores indicate the participant is more knowledgeable about healthy nutrition (score range, 0–78).¹¹

The DASS-21 questionnaire includes 21 questions that measure the severity of a range of symptoms common to depression, anxiety, and stress. Responses include Likert-type answer options (eg, “never,” “sometimes,” “often,” and “almost always.”) On this scale, a higher score (range of 0–21 for each) indicates higher levels of depression, anxiety, and stress.¹²

Statistical Analysis—Descriptive statistics were analyzed using SPSS software version 22.0 (IBM). The Shapiro-Wilk test was applied to determine whether the data were normally distributed. For categorical variables, frequency differences among groups were compared using the Pearson χ² test. A t test was used to compare the means of 2 independent groups with a normal distribution. One-way analysis of variance and Tukey Honest Significant Difference post hoc analysis were used to test whether there was a statistically significant difference among the normally distributed means of independent groups. Pearson correlation analysis was used to determine whether there was a linear relationship between 2 numeric measurements and, if so, to determine the direction and severity of this relationship. P<.05 indicated statistical significance in this study.

Results

Study Participant Demographics—This study included 82 participants with a mean age of 44.3 years; 52.4% (43/82) were female, and 85.4% (70/82) were married. The questionnaire took an average of 4.2 minutes for participants to complete. A total of 57.3% (47/82) of patients had completed primary/secondary education and 59.8% (49/82) were employed. The mean BMI was 28.1 kg/m². According to the BMI classification, 26.8% (22/82) participants had a normal weight, 36.6% (30/82) were overweight, and 43.9% (36/82) were obese. A total of 48.8% (40/82) of participants smoked, and 4.9% (4/82) consumed alcohol. The mean Psoriasis Area and Severity Index score was 5.4. A total of 54.9% (45/82) of participants had nail psoriasis, and 24.4% (20/82) had psoriatic arthritis. The mean duration of psoriasis was 153 months. A total of 29.3% (24/82) of participants had a positive family history of psoriasis. The mean EAT-26 score was 11.1. A total of 12.2% (10/82) of participants had an EAT-26 score of 20 points or higher and were considered at high risk for an ED. The mean ASHN score was 72.9; the mean DLQI score was 5.5; and on the DASS-21 scale, mean scores for depression, anxiety, and stress were 6.3, 8.7, and 10.0, respectively (Table).

Comparative Evaluation of the BMI Groups—The only statistically significant differences among the 3 BMI groups were related to marital status, EAT-26 score, and anxiety and stress scores (P=.02, <.01, <.01, and <.01, respectively)(eTable 1). The number of single/divorced participants in the overweight group was significantly (P=.02) greater than in the normal weight group. The mean EAT-26 score for the normal weight group was significantly (P<.01) lower than for the overweight and obese groups; there was no significant difference in mean EAT-26 scores between the overweight and obese groups. The mean anxiety score was significantly (P<.01) lower in the normal weight group compared with the overweight and obese groups. There was no significant difference between the overweight and obese groups according to the mean depression score. The mean stress and anxiety scores were significantly (P<.01) lower in the normal weight group than in the overweight and obese groups. There was no significant difference between the overweight and obese groups according to the mean anxiety score.

Comment

Eating disorders are psychiatric conditions that require a multidisciplinary approach. Nonpsychiatric medical departments may be involved due to the severe consequences (eg, various skin changes¹⁴) of these disorders. Psoriasis is not known to be directly affected by the presence of an ED; however, it is possible that EDs could indirectly affect patients with psoriasis by influencing obesity. Therefore, this study aimed to examine the relationship between ED risk factors and obesity in this population.

The relationship between psoriasis and obesity has been a popular research topic in dermatology since the 1990s.¹⁵ Epidemiologic and observational studies have reported that patients with psoriasis are more likely to be overweight or have obesity, which is an independent risk factor for psoriasis.^3,16 However, the causal relationship between psoriasis and obesity remains unclear. In a comprehensive review, Barros et al¹⁷ emphasized the causal relationship between obesity and psoriasis under several headings. Firstly, a higher BMI increases the risk for psoriasis by promoting cytokine release and immune system dysregulation. Secondly, a Western diet (eg, processed foods and fast food) triggers obesity and psoriasis by increasing adipose tissue. Thirdly, the alteration of the skin and gut microbiota triggers chronic inflammation as a result of bacterial translocation in patients with obesity. Fourthly, a high-fat diet and palmitic acid disrupt the intestinal integrity of the gut and increase the risk for psoriasis and obesity by triggering chronic inflammation of bacterial fragments that pass into the blood. Finally, the decrease in the amount of adiponectin and the increase in the amount of leptin in patients with obesity may cause psoriasis by increasing proinflammatory cytokines, which are similar to those involved in the pathogenesis of psoriasis.¹⁷ Additionally, psoriatic inflammation can cause insulin resistance and metabolic dysfunction, leading to obesity.¹⁸ The relationship between psoriasis and obesity cannot be solely explained by metabolic pathways. Smoking, alcohol consumption, and a sedentary lifestyle all are associated with psoriasis and also can contribute to obesity.⁵ Our study revealed no significant difference in smoking or alcohol consumption between the normal weight and overweight/obesity groups. Based on our data, we determined that smoking and alcohol consumption did not affect obesity in our patients with psoriasis.

Observational and epidemiologic studies have shown that patients with psoriasis experience increased rates of depression, anxiety, and stress.¹⁹ In studies of pathogenesis, a connection between depression and psoriatic inflammation has been established.²⁰ It is known that inflammatory cytokines similar to those in psoriasis are involved in the development of obesity.¹⁸ In addition, depression and anxiety can lead to binge eating, unhealthy food choices, and a more sedentary lifestyle.⁵ All of these variables may contribute to the associations between depression and anxiety with psoriasis and obesity. Zafiriou et al²¹ conducted a study to investigate the relationship between psoriasis, obesity, and depression through inflammatory pathways with a focus on the importance of IL-17. Data showing that IL-17–producing Th17-cell subgroups play a considerable role in the development of obesity and depression prompted the authors to suggest that psoriasis, obesity, and ­anxiety/depression may be interconnected manifestations of immune dysregulation, potentially linked to IL-17 and its associated cells.²¹ Mrowietz et al²² also suggested that metabolic inflammation may contribute to obesity and depression in patients with psoriasis and highlighted the importance of several cytokines, including tumor necrosis factor α, IL-6, IL-8, IL-17, and IL-23. Our study revealed no significant differences in depression scores between BMI groups. Another meta-analysis reported conflicting findings on the incidence of depression in obese patients with psoriasis.²³ Some of the studies had a small number of participants. Compared to depression, anxiety has received less attention in studies of patients with obesity with psoriasis. However, these studies have shown a positive correlation between anxiety scores and BMI in patients with psoriasis.^24,25 In our study, similar to the findings of previous studies, overweight patients and those with obesitywho have psoriasis had significantly (P<.01) greater anxiety and stress scores than did normal weight patients with psoriasis.

Obesity should be assessed in patients with psoriasis via a biopsychosocial approach that takes into account genetic, behavioral, and environmental factors.²⁶ Eating disorders are considered to be one of the factors contributing to obesity. Numerous studies in the literature have demonstrated a greater incidence of EDs in patients with obesity vs those without obesity.^5,6,27 Obesity and EDs have a bidirectional relationship: individuals with obesity are at risk for EDs due to body dissatisfaction, dieting habits, and depressive states. Conversely, poor eating behaviors in individuals with a normal weight can lead to obesity.²⁸

There are few studies in the literature exploring the relationship between psoriasis and EDs. Crosta et al²⁹ demonstrated that patients with psoriasis had impaired results on ED screening tests and that these scores deteriorated further as BMI increased. Moreover, Altunay et al³⁰ demonstrated that patients with psoriasis and metabolic syndrome had higher scores on the ED screening test. In this study, patients with higher scores also exhibited high levels of anxiety.³⁰ In our study, similar to the findings of previous studies, patients with psoriasis who were overweight or had obesity had significantly (P<.01) greater EAT-26 scores than those in the normal weight group. Patients with high EAT-26 scores also exhibited elevated levels of depression, anxiety, and stress. Additionally, EAT-26 scores were positively correlated with BMI, anxiety, depression, and stress scores. Our study as well as other studies in the literature indicate that additional research is needed to determine the associations between EDs and obesity in psoriasis.

Conclusion

Managing obesity is crucial for patients with psoriasis. This study showed that EAT-26 scores were higher in patients with psoriasis who were overweight or had obesity than in those who were normal weight. Participants with high EAT-26 scores (≥20 points) were more likely to be female and have higher anxiety and stress scores. In addition, EAT-26 scores were positively correlated with BMI as well as depression, anxiety, and stress scores. Eating disorders may contribute to the development of obesity in patients with psoriasis. Although our study was limited by a small sample size, the results suggest that there is a need for large-scale multicenter studies to investigate the relationship between psoriasis and EDs.

Psoriasis is a chronic multisystemic inflammatory skin disease with a worldwide prevalence of 2% to 3%.¹ Psoriasis can be accompanied by other conditions such as psoriatic arthritis, obesity, metabolic syndrome, diabetes mellitus, hypertension, dyslipidemia, atherosclerotic disease, inflammatory bowel disease, and anxiety/depression. It is important to manage comorbidities of psoriasis in addition to treating the cutaneous manifestations of the disease.¹

Obesity is a major public health concern worldwide. Numerous observational and epidemiologic studies have reported a high prevalence of obesity among patients with psoriasis.² Current evidence indicates that obesity may initiate or worsen psoriasis; furthermore, it is important to note that obesity may negatively impact the effectiveness of psoriasis-specific treatments or increase the incidence of adverse effects. Therefore, managing obesity is crucial in the treatment of psoriasis.³ Numerous studies have investigated the association between psoriasis and obesity, and they commonly conclude that both conditions share the same genetic metabolic pathways.^2-4 However, it is important to consider environmental factors such as dietary habits, smoking, alcohol consumption, and a sedentary lifestyle—all of which are associated with psoriasis and also can contribute to the development of obesity.⁵ Because of the effects of obesity in psoriasis patients, factors that impact the development of obesity have become a popular research topic.

Eating disorders (EDs) are a crucial risk factor for both developing and maintaining obesity. In particular, two EDs that are associated with obesity include binge eating disorder and bulimia nervosa.⁶ According to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition,⁷ binge eating disorder can be diagnosed when a patient has at least 1 episode of binge eating per week over a 3-month period. Bulimia nervosa can be diagnosed when a patient is excessively concerned with their body weight and shape and engages in behaviors to prevent weight gain (eg, forced vomiting, excessive use of laxatives).⁷ Psychiatrists who specialize in EDs make diagnoses based on these criteria. In daily practice, there are several quick and simple questionnaires available to screen for EDs that can be used by nonpsychiatrist physicians, including the commonly used 26-item Eating Attitudes Test (EAT-26).⁸ The EAT-26 has been used to screen for EDs in patients with inflammatory disorders.⁹

The aim of this study was to screen for EDs in patients with psoriasis to identify potential risk factors for development of obesity.

Materials and Methods

This study included patients with psoriasis who were screened for EDs at a tertiary dermatology clinic in Turkey between January 2021 and December 2023. This study was approved by the local ethics committee and was in accordance with the Declaration of Helsinki (decision number E-93471371-514.99-225000079).

Study Design and Patient Inclusion Criteria—This quantitative cross-sectional study utilized EAT-26, Dermatology Life Quality Index (DLQI), Attitude Scale for Healthy Nutrition (ASHN), and Depression Anxiety Stress Scale-21 (DASS-21) scores. All the questionnaire scales used in the study were adapted and validated in Turkey.^8,10-12 The inclusion criteria consisted of being older than 18 years of age, being literate, having psoriasis for at least 1 year that was not treated topically or systemically, and having no psychiatric diseases outside an ED. The questionnaires were presented in written format following the clinical examination. Literacy was an inclusion criterion in this study due to the absence of auxiliary health personnel.

Study Variables—The study variables included age, sex, marital status (single/divorced or married), education status (primary/secondary school or high school/university), employment status (employed or unemployed/retired), body mass index (BMI), smoking status, alcohol-consumption status, Psoriasis Area Severity Index score, presence of nail psoriasis and psoriatic arthritis, duration of psoriasis, family history of psoriasis, EAT-26 score, ASHN score, DLQI score, and DASS-21 score. Body mass index was calculated by taking a participant’s weight in kilograms and dividing it by their height in meters squared. The BMI values were classified into 3 categories: normal (18.5–24.9 kg/m²), overweight (25.0–29.9 kg/m²), and obese (≥30 kg/m²).¹³

Questionnaires—The EAT-26 questionnaire includes 26 questions that are used to detect EDs. Responses to each question include Likert-type answer options (ie, “always,” “usually,” “often,” “sometimes,” “rarely,” and “never.”) Patients with scores of 20 points or higher (range, 0–78) are classified as high risk for EDs.⁸ In our study, EAT-26 scores were grouped into 2 categories: patients scoring less than 20 points and those scoring 20 points or higher.

The DLQI questionnaire includes 10 questions to measure dermatologic symptoms and qualiy of life. Responses to each question include Likert-type answer options (ie, “not at all,” “a little,” “a lot,” or “very much.”) On the DLQI scale, the higher the score, the lower the quality of life (score range, 0–30).¹⁰

The ASHN questionnaire includes 21 questions that measure attitudes toward healthy nutrition with 5 possible answer options (“strongly disagree,” “disagree,” “undecided,” “agree,” and “strongly agree”). On this scale, higher scores indicate the participant is more knowledgeable about healthy nutrition (score range, 0–78).¹¹

The DASS-21 questionnaire includes 21 questions that measure the severity of a range of symptoms common to depression, anxiety, and stress. Responses include Likert-type answer options (eg, “never,” “sometimes,” “often,” and “almost always.”) On this scale, a higher score (range of 0–21 for each) indicates higher levels of depression, anxiety, and stress.¹²

Statistical Analysis—Descriptive statistics were analyzed using SPSS software version 22.0 (IBM). The Shapiro-Wilk test was applied to determine whether the data were normally distributed. For categorical variables, frequency differences among groups were compared using the Pearson χ² test. A t test was used to compare the means of 2 independent groups with a normal distribution. One-way analysis of variance and Tukey Honest Significant Difference post hoc analysis were used to test whether there was a statistically significant difference among the normally distributed means of independent groups. Pearson correlation analysis was used to determine whether there was a linear relationship between 2 numeric measurements and, if so, to determine the direction and severity of this relationship. P<.05 indicated statistical significance in this study.

Results

Study Participant Demographics—This study included 82 participants with a mean age of 44.3 years; 52.4% (43/82) were female, and 85.4% (70/82) were married. The questionnaire took an average of 4.2 minutes for participants to complete. A total of 57.3% (47/82) of patients had completed primary/secondary education and 59.8% (49/82) were employed. The mean BMI was 28.1 kg/m². According to the BMI classification, 26.8% (22/82) participants had a normal weight, 36.6% (30/82) were overweight, and 43.9% (36/82) were obese. A total of 48.8% (40/82) of participants smoked, and 4.9% (4/82) consumed alcohol. The mean Psoriasis Area and Severity Index score was 5.4. A total of 54.9% (45/82) of participants had nail psoriasis, and 24.4% (20/82) had psoriatic arthritis. The mean duration of psoriasis was 153 months. A total of 29.3% (24/82) of participants had a positive family history of psoriasis. The mean EAT-26 score was 11.1. A total of 12.2% (10/82) of participants had an EAT-26 score of 20 points or higher and were considered at high risk for an ED. The mean ASHN score was 72.9; the mean DLQI score was 5.5; and on the DASS-21 scale, mean scores for depression, anxiety, and stress were 6.3, 8.7, and 10.0, respectively (Table).

Comparative Evaluation of the BMI Groups—The only statistically significant differences among the 3 BMI groups were related to marital status, EAT-26 score, and anxiety and stress scores (P=.02, <.01, <.01, and <.01, respectively)(eTable 1). The number of single/divorced participants in the overweight group was significantly (P=.02) greater than in the normal weight group. The mean EAT-26 score for the normal weight group was significantly (P<.01) lower than for the overweight and obese groups; there was no significant difference in mean EAT-26 scores between the overweight and obese groups. The mean anxiety score was significantly (P<.01) lower in the normal weight group compared with the overweight and obese groups. There was no significant difference between the overweight and obese groups according to the mean depression score. The mean stress and anxiety scores were significantly (P<.01) lower in the normal weight group than in the overweight and obese groups. There was no significant difference between the overweight and obese groups according to the mean anxiety score.

Comment

Eating disorders are psychiatric conditions that require a multidisciplinary approach. Nonpsychiatric medical departments may be involved due to the severe consequences (eg, various skin changes¹⁴) of these disorders. Psoriasis is not known to be directly affected by the presence of an ED; however, it is possible that EDs could indirectly affect patients with psoriasis by influencing obesity. Therefore, this study aimed to examine the relationship between ED risk factors and obesity in this population.

The relationship between psoriasis and obesity has been a popular research topic in dermatology since the 1990s.¹⁵ Epidemiologic and observational studies have reported that patients with psoriasis are more likely to be overweight or have obesity, which is an independent risk factor for psoriasis.^3,16 However, the causal relationship between psoriasis and obesity remains unclear. In a comprehensive review, Barros et al¹⁷ emphasized the causal relationship between obesity and psoriasis under several headings. Firstly, a higher BMI increases the risk for psoriasis by promoting cytokine release and immune system dysregulation. Secondly, a Western diet (eg, processed foods and fast food) triggers obesity and psoriasis by increasing adipose tissue. Thirdly, the alteration of the skin and gut microbiota triggers chronic inflammation as a result of bacterial translocation in patients with obesity. Fourthly, a high-fat diet and palmitic acid disrupt the intestinal integrity of the gut and increase the risk for psoriasis and obesity by triggering chronic inflammation of bacterial fragments that pass into the blood. Finally, the decrease in the amount of adiponectin and the increase in the amount of leptin in patients with obesity may cause psoriasis by increasing proinflammatory cytokines, which are similar to those involved in the pathogenesis of psoriasis.¹⁷ Additionally, psoriatic inflammation can cause insulin resistance and metabolic dysfunction, leading to obesity.¹⁸ The relationship between psoriasis and obesity cannot be solely explained by metabolic pathways. Smoking, alcohol consumption, and a sedentary lifestyle all are associated with psoriasis and also can contribute to obesity.⁵ Our study revealed no significant difference in smoking or alcohol consumption between the normal weight and overweight/obesity groups. Based on our data, we determined that smoking and alcohol consumption did not affect obesity in our patients with psoriasis.

Observational and epidemiologic studies have shown that patients with psoriasis experience increased rates of depression, anxiety, and stress.¹⁹ In studies of pathogenesis, a connection between depression and psoriatic inflammation has been established.²⁰ It is known that inflammatory cytokines similar to those in psoriasis are involved in the development of obesity.¹⁸ In addition, depression and anxiety can lead to binge eating, unhealthy food choices, and a more sedentary lifestyle.⁵ All of these variables may contribute to the associations between depression and anxiety with psoriasis and obesity. Zafiriou et al²¹ conducted a study to investigate the relationship between psoriasis, obesity, and depression through inflammatory pathways with a focus on the importance of IL-17. Data showing that IL-17–producing Th17-cell subgroups play a considerable role in the development of obesity and depression prompted the authors to suggest that psoriasis, obesity, and ­anxiety/depression may be interconnected manifestations of immune dysregulation, potentially linked to IL-17 and its associated cells.²¹ Mrowietz et al²² also suggested that metabolic inflammation may contribute to obesity and depression in patients with psoriasis and highlighted the importance of several cytokines, including tumor necrosis factor α, IL-6, IL-8, IL-17, and IL-23. Our study revealed no significant differences in depression scores between BMI groups. Another meta-analysis reported conflicting findings on the incidence of depression in obese patients with psoriasis.²³ Some of the studies had a small number of participants. Compared to depression, anxiety has received less attention in studies of patients with obesity with psoriasis. However, these studies have shown a positive correlation between anxiety scores and BMI in patients with psoriasis.^24,25 In our study, similar to the findings of previous studies, overweight patients and those with obesitywho have psoriasis had significantly (P<.01) greater anxiety and stress scores than did normal weight patients with psoriasis.

Obesity should be assessed in patients with psoriasis via a biopsychosocial approach that takes into account genetic, behavioral, and environmental factors.²⁶ Eating disorders are considered to be one of the factors contributing to obesity. Numerous studies in the literature have demonstrated a greater incidence of EDs in patients with obesity vs those without obesity.^5,6,27 Obesity and EDs have a bidirectional relationship: individuals with obesity are at risk for EDs due to body dissatisfaction, dieting habits, and depressive states. Conversely, poor eating behaviors in individuals with a normal weight can lead to obesity.²⁸

There are few studies in the literature exploring the relationship between psoriasis and EDs. Crosta et al²⁹ demonstrated that patients with psoriasis had impaired results on ED screening tests and that these scores deteriorated further as BMI increased. Moreover, Altunay et al³⁰ demonstrated that patients with psoriasis and metabolic syndrome had higher scores on the ED screening test. In this study, patients with higher scores also exhibited high levels of anxiety.³⁰ In our study, similar to the findings of previous studies, patients with psoriasis who were overweight or had obesity had significantly (P<.01) greater EAT-26 scores than those in the normal weight group. Patients with high EAT-26 scores also exhibited elevated levels of depression, anxiety, and stress. Additionally, EAT-26 scores were positively correlated with BMI, anxiety, depression, and stress scores. Our study as well as other studies in the literature indicate that additional research is needed to determine the associations between EDs and obesity in psoriasis.

Conclusion

Managing obesity is crucial for patients with psoriasis. This study showed that EAT-26 scores were higher in patients with psoriasis who were overweight or had obesity than in those who were normal weight. Participants with high EAT-26 scores (≥20 points) were more likely to be female and have higher anxiety and stress scores. In addition, EAT-26 scores were positively correlated with BMI as well as depression, anxiety, and stress scores. Eating disorders may contribute to the development of obesity in patients with psoriasis. Although our study was limited by a small sample size, the results suggest that there is a need for large-scale multicenter studies to investigate the relationship between psoriasis and EDs.

More than 37 million Americans have diabetes mellitus (DM), and approximately 90% have type 2 DM (T2DM), including about 25% of veterans.^1,2 The current guidelines suggest that therapy depends on a patient's comorbidities, management needs, and patient-centered treatment factors.³ About 1 in 3 adults with DM have chronic kidney disease (CKD), defined as the presence of kidney damage or an estimated glomerular filtration rate (eGFR) < 60 mL/min per 1.73 m², persisting for ≥ 3 months.⁴

Sodium-glucose cotransporter-2 (SGLT-2) inhibitors are a class of antihyperglycemic agents acting on the SGLT-2 proteins expressed in the renal proximal convoluted tubules. They exert their effects by preventing the reabsorption of filtered glucose from the tubular lumen. There are 4 SGLT-2 inhibitors approved by the US Food and Drug Administration: canagliflozin, dapagliflozin, empagliflozin, and ertugliflozin. Empagliflozin is currently the preferred SGLT-2 inhibitor on the US Department of Veterans Affairs (VA) formulary.

According to the American Diabetes Association guidelines, empagliflozin is considered when an individual has or is at risk for atherosclerotic cardiovascular disease, heart failure, and CKD.³ SGLT-2 inhibitors are a favorable option due to their low risk for hypoglycemia while also promoting weight loss. The EMPEROR-Reduced trial demonstrated that, in addition to benefits for patients with heart failure, empagliflozin also slowed the progressive decline in kidney function in those with and without DM.⁵ The purpose of this study was to evaluate the effectiveness of empagliflozin on hemoglobin A_1c (HbA_1c) levels in patients with CKD at the Hershel “Woody” Williams VA Medical Center (HWWVAMC) in Huntington, West Virginia, along with other laboratory test markers.

Methods

The Marshall University Institutional Review Board #1 (Medical) and the HWWVAMC institutional review board and research and development committee each reviewed and approved this study. A retrospective chart review was conducted on patients diagnosed with T2DM and stage 3 CKD who were prescribed empagliflozin for DM management between January 1, 2015, and October 1, 2022, yielding 1771 patients. Data were obtained through the VHA Corporate Data Warehouse (CDW) and stored on the VA Informatics and Computing Infrastructure (VINCI) research server.

Patients were included if they were aged 18 to 89 years, prescribed empagliflozin by a VA clinician for the treatment of T2DM, had an eGFR between 30 and 59 mL/min/1.73 m², and had an initial HbA_1c between 7% and 10%. Using further random sampling, patients were either excluded or divided into, those with stage 3a CKD and those with stage 3b CKD. The primary endpoint of this study was the change in HbA_1c levels in patients with stage 3b CKD (eGFR 30-44 mL/min/1.73 m²) compared with stage 3a (eGFR 45-59 mL/min/1.73 m²) after 12 months. The secondary endpoints included effects on renal function, weight, blood pressure, incidence of adverse drug events, and cardiovascular events. Of the excluded, 38 had HbA_1c < 7%, 30 had HbA_1c ≥ 10%, 21 did not have data at 1-year mark, 15 had the medication discontinued due to decline in renal function, 14 discontinued their medication without documented reason, 10 discontinued their medication due to adverse drug reactions (ADRs), 12 had eGFR > 60 mL/ min/1.73 m², 9 died within 1 year of initiation, 4 had eGFR < 30 mL/min/1.73 m², 1 had no baseline eGFR, and 1 was the spouse of a veteran.

Statistical Analysis

All statistical analyses were performed using STATA v.15. We used t tests to examine changes within each group, along with paired t tests to compare the 2 groups. Two-sample t tests were used to analyze the continuous data at both the primary and secondary endpoints.

Results

Of the 1771 patients included in the initial data set, a randomized sample of 255 charts were reviewed, 155 were excluded, and 100 were included. Fifty patients, had stage 3a CKD and 50 had stage 3b CKD. Baseline demographics were similar between the stage 3a and 3b groups (Table 1). Both groups were predominantly White and male, with mean age > 70 years.

The primary endpoint was the differences in HbA_1c levels over time and between groups for patients with stage 3a and stage 3b CKD 1 year after initiation of empagliflozin. The starting doses of empagliflozin were either 12.5 mg or 25.0 mg. For both groups, the changes in HbA_1c levels were statistically significant (Table 2). HbA_1c levels dropped 0.65% for the stage 3a group and 0.48% for the 3b group. When compared to one another, the results were not statistically significant (P = .51).

Secondary Endpoint

There was no statistically significant difference in serum creatinine levels within each group between baselines and 1 year later for the stage 3a (P = .21) and stage 3b (P = .22) groups, or when compared to each other (P = .67). There were statistically significant changes in weight for patients in the stage 3a group (P < .05), but not for stage 3b group (P = .06) or when compared to each other (P = .41). A statistically significant change in systolic blood pressure was observed for the stage 3a group (P = .003), but not the stage 3b group (P = .16) or when compared to each other (P = .27). There were statistically significant changes in diastolic blood pressure within the stage 3a group (P = .04), but not within the stage 3b group (P = .61) or when compared to each other (P = .31).

Ten patients discontinued empagliflozin before the 1-year mark due to ADRs, including dizziness, increased incidence of urinary tract infections, rash, and tachycardia (Table 3). Additionally, 3 ADRs resulted in the empagliflozin discontinuation after 1 year (Table 3).

Discussion

This study showed a statistically significant change in HbA_1c levels for patients with stage 3a and stage 3b CKD. With eGFR levels in these 2 groups > 30 mL/min/1.73 m², patients were able to achieve glycemic benefits. There were no significant changes to the serum creatinine levels. Both groups saw statistically significant changes in weight loss within their own group; however, there were no statistically significant changes when compared to each other. With both systolic and diastolic blood pressure, the stage 3a group had statistically significant changes.

The EMPA-REG BP study demonstrated that empagliflozin was associated with significant and clinically meaningful reductions in blood pressure and HbA_1c levels compared with placebo and was well tolerated in patients with T2DM and hypertension.^6,7,8

Limitations

This study had a retrospective study design, which resulted in missing information for many patients and higher rates of exclusion. The population was predominantly older, White, and male and may not reflect other populations. The starting doses of empagliflozin varied between the groups. The VA employs tablet splitting for some patients, and the available doses were either 10.0 mg, 12.5 mg, or 25.0 mg. Some prescribers start veterans at lower doses and gradually increase to the higher dose of 25.0 mg, adding to the variability in starting doses.

Patients with eGFR < 30 mL/min/1.73 m² make it difficult to determine any potential benefit in this population. The EMPA-KIDNEY trial demonstrated that the benefits of empagliflozin treatment were consistent among patients with or without DM and regardless of eGFR at randomization.⁹ Furthermore, many veterans had an initial HbA_1c levels outside the inclusion criteria range, which was a factor in the smaller sample size.

Conclusions

While the reduction in HbA_1c levels was less in patients with stage 3b CKD compared to patients stage 3a CKD, all patients experienced a benefit. The overall incidence of ADRs was low in the study population, showing empagliflozin as a favorable choice for those with T2DM and CKD. Based on the findings of this study, empagliflozin is a potentially beneficial option for reducing HbA_1c levels in patients with CKD.

More than 37 million Americans have diabetes mellitus (DM), and approximately 90% have type 2 DM (T2DM), including about 25% of veterans.^1,2 The current guidelines suggest that therapy depends on a patient's comorbidities, management needs, and patient-centered treatment factors.³ About 1 in 3 adults with DM have chronic kidney disease (CKD), defined as the presence of kidney damage or an estimated glomerular filtration rate (eGFR) < 60 mL/min per 1.73 m², persisting for ≥ 3 months.⁴

Sodium-glucose cotransporter-2 (SGLT-2) inhibitors are a class of antihyperglycemic agents acting on the SGLT-2 proteins expressed in the renal proximal convoluted tubules. They exert their effects by preventing the reabsorption of filtered glucose from the tubular lumen. There are 4 SGLT-2 inhibitors approved by the US Food and Drug Administration: canagliflozin, dapagliflozin, empagliflozin, and ertugliflozin. Empagliflozin is currently the preferred SGLT-2 inhibitor on the US Department of Veterans Affairs (VA) formulary.

According to the American Diabetes Association guidelines, empagliflozin is considered when an individual has or is at risk for atherosclerotic cardiovascular disease, heart failure, and CKD.³ SGLT-2 inhibitors are a favorable option due to their low risk for hypoglycemia while also promoting weight loss. The EMPEROR-Reduced trial demonstrated that, in addition to benefits for patients with heart failure, empagliflozin also slowed the progressive decline in kidney function in those with and without DM.⁵ The purpose of this study was to evaluate the effectiveness of empagliflozin on hemoglobin A_1c (HbA_1c) levels in patients with CKD at the Hershel “Woody” Williams VA Medical Center (HWWVAMC) in Huntington, West Virginia, along with other laboratory test markers.

Methods

The Marshall University Institutional Review Board #1 (Medical) and the HWWVAMC institutional review board and research and development committee each reviewed and approved this study. A retrospective chart review was conducted on patients diagnosed with T2DM and stage 3 CKD who were prescribed empagliflozin for DM management between January 1, 2015, and October 1, 2022, yielding 1771 patients. Data were obtained through the VHA Corporate Data Warehouse (CDW) and stored on the VA Informatics and Computing Infrastructure (VINCI) research server.

Patients were included if they were aged 18 to 89 years, prescribed empagliflozin by a VA clinician for the treatment of T2DM, had an eGFR between 30 and 59 mL/min/1.73 m², and had an initial HbA_1c between 7% and 10%. Using further random sampling, patients were either excluded or divided into, those with stage 3a CKD and those with stage 3b CKD. The primary endpoint of this study was the change in HbA_1c levels in patients with stage 3b CKD (eGFR 30-44 mL/min/1.73 m²) compared with stage 3a (eGFR 45-59 mL/min/1.73 m²) after 12 months. The secondary endpoints included effects on renal function, weight, blood pressure, incidence of adverse drug events, and cardiovascular events. Of the excluded, 38 had HbA_1c < 7%, 30 had HbA_1c ≥ 10%, 21 did not have data at 1-year mark, 15 had the medication discontinued due to decline in renal function, 14 discontinued their medication without documented reason, 10 discontinued their medication due to adverse drug reactions (ADRs), 12 had eGFR > 60 mL/ min/1.73 m², 9 died within 1 year of initiation, 4 had eGFR < 30 mL/min/1.73 m², 1 had no baseline eGFR, and 1 was the spouse of a veteran.

Statistical Analysis

All statistical analyses were performed using STATA v.15. We used t tests to examine changes within each group, along with paired t tests to compare the 2 groups. Two-sample t tests were used to analyze the continuous data at both the primary and secondary endpoints.

Results

Of the 1771 patients included in the initial data set, a randomized sample of 255 charts were reviewed, 155 were excluded, and 100 were included. Fifty patients, had stage 3a CKD and 50 had stage 3b CKD. Baseline demographics were similar between the stage 3a and 3b groups (Table 1). Both groups were predominantly White and male, with mean age > 70 years.

The primary endpoint was the differences in HbA_1c levels over time and between groups for patients with stage 3a and stage 3b CKD 1 year after initiation of empagliflozin. The starting doses of empagliflozin were either 12.5 mg or 25.0 mg. For both groups, the changes in HbA_1c levels were statistically significant (Table 2). HbA_1c levels dropped 0.65% for the stage 3a group and 0.48% for the 3b group. When compared to one another, the results were not statistically significant (P = .51).

Secondary Endpoint

There was no statistically significant difference in serum creatinine levels within each group between baselines and 1 year later for the stage 3a (P = .21) and stage 3b (P = .22) groups, or when compared to each other (P = .67). There were statistically significant changes in weight for patients in the stage 3a group (P < .05), but not for stage 3b group (P = .06) or when compared to each other (P = .41). A statistically significant change in systolic blood pressure was observed for the stage 3a group (P = .003), but not the stage 3b group (P = .16) or when compared to each other (P = .27). There were statistically significant changes in diastolic blood pressure within the stage 3a group (P = .04), but not within the stage 3b group (P = .61) or when compared to each other (P = .31).

Ten patients discontinued empagliflozin before the 1-year mark due to ADRs, including dizziness, increased incidence of urinary tract infections, rash, and tachycardia (Table 3). Additionally, 3 ADRs resulted in the empagliflozin discontinuation after 1 year (Table 3).

Discussion

This study showed a statistically significant change in HbA_1c levels for patients with stage 3a and stage 3b CKD. With eGFR levels in these 2 groups > 30 mL/min/1.73 m², patients were able to achieve glycemic benefits. There were no significant changes to the serum creatinine levels. Both groups saw statistically significant changes in weight loss within their own group; however, there were no statistically significant changes when compared to each other. With both systolic and diastolic blood pressure, the stage 3a group had statistically significant changes.

The EMPA-REG BP study demonstrated that empagliflozin was associated with significant and clinically meaningful reductions in blood pressure and HbA_1c levels compared with placebo and was well tolerated in patients with T2DM and hypertension.^6,7,8

Limitations

This study had a retrospective study design, which resulted in missing information for many patients and higher rates of exclusion. The population was predominantly older, White, and male and may not reflect other populations. The starting doses of empagliflozin varied between the groups. The VA employs tablet splitting for some patients, and the available doses were either 10.0 mg, 12.5 mg, or 25.0 mg. Some prescribers start veterans at lower doses and gradually increase to the higher dose of 25.0 mg, adding to the variability in starting doses.

Patients with eGFR < 30 mL/min/1.73 m² make it difficult to determine any potential benefit in this population. The EMPA-KIDNEY trial demonstrated that the benefits of empagliflozin treatment were consistent among patients with or without DM and regardless of eGFR at randomization.⁹ Furthermore, many veterans had an initial HbA_1c levels outside the inclusion criteria range, which was a factor in the smaller sample size.

Conclusions

While the reduction in HbA_1c levels was less in patients with stage 3b CKD compared to patients stage 3a CKD, all patients experienced a benefit. The overall incidence of ADRs was low in the study population, showing empagliflozin as a favorable choice for those with T2DM and CKD. Based on the findings of this study, empagliflozin is a potentially beneficial option for reducing HbA_1c levels in patients with CKD.

More than 37 million Americans have diabetes mellitus (DM), and approximately 90% have type 2 DM (T2DM), including about 25% of veterans.^1,2 The current guidelines suggest that therapy depends on a patient's comorbidities, management needs, and patient-centered treatment factors.³ About 1 in 3 adults with DM have chronic kidney disease (CKD), defined as the presence of kidney damage or an estimated glomerular filtration rate (eGFR) < 60 mL/min per 1.73 m², persisting for ≥ 3 months.⁴

Sodium-glucose cotransporter-2 (SGLT-2) inhibitors are a class of antihyperglycemic agents acting on the SGLT-2 proteins expressed in the renal proximal convoluted tubules. They exert their effects by preventing the reabsorption of filtered glucose from the tubular lumen. There are 4 SGLT-2 inhibitors approved by the US Food and Drug Administration: canagliflozin, dapagliflozin, empagliflozin, and ertugliflozin. Empagliflozin is currently the preferred SGLT-2 inhibitor on the US Department of Veterans Affairs (VA) formulary.

According to the American Diabetes Association guidelines, empagliflozin is considered when an individual has or is at risk for atherosclerotic cardiovascular disease, heart failure, and CKD.³ SGLT-2 inhibitors are a favorable option due to their low risk for hypoglycemia while also promoting weight loss. The EMPEROR-Reduced trial demonstrated that, in addition to benefits for patients with heart failure, empagliflozin also slowed the progressive decline in kidney function in those with and without DM.⁵ The purpose of this study was to evaluate the effectiveness of empagliflozin on hemoglobin A_1c (HbA_1c) levels in patients with CKD at the Hershel “Woody” Williams VA Medical Center (HWWVAMC) in Huntington, West Virginia, along with other laboratory test markers.

Methods

The Marshall University Institutional Review Board #1 (Medical) and the HWWVAMC institutional review board and research and development committee each reviewed and approved this study. A retrospective chart review was conducted on patients diagnosed with T2DM and stage 3 CKD who were prescribed empagliflozin for DM management between January 1, 2015, and October 1, 2022, yielding 1771 patients. Data were obtained through the VHA Corporate Data Warehouse (CDW) and stored on the VA Informatics and Computing Infrastructure (VINCI) research server.

Patients were included if they were aged 18 to 89 years, prescribed empagliflozin by a VA clinician for the treatment of T2DM, had an eGFR between 30 and 59 mL/min/1.73 m², and had an initial HbA_1c between 7% and 10%. Using further random sampling, patients were either excluded or divided into, those with stage 3a CKD and those with stage 3b CKD. The primary endpoint of this study was the change in HbA_1c levels in patients with stage 3b CKD (eGFR 30-44 mL/min/1.73 m²) compared with stage 3a (eGFR 45-59 mL/min/1.73 m²) after 12 months. The secondary endpoints included effects on renal function, weight, blood pressure, incidence of adverse drug events, and cardiovascular events. Of the excluded, 38 had HbA_1c < 7%, 30 had HbA_1c ≥ 10%, 21 did not have data at 1-year mark, 15 had the medication discontinued due to decline in renal function, 14 discontinued their medication without documented reason, 10 discontinued their medication due to adverse drug reactions (ADRs), 12 had eGFR > 60 mL/ min/1.73 m², 9 died within 1 year of initiation, 4 had eGFR < 30 mL/min/1.73 m², 1 had no baseline eGFR, and 1 was the spouse of a veteran.

Statistical Analysis

All statistical analyses were performed using STATA v.15. We used t tests to examine changes within each group, along with paired t tests to compare the 2 groups. Two-sample t tests were used to analyze the continuous data at both the primary and secondary endpoints.

Results

Of the 1771 patients included in the initial data set, a randomized sample of 255 charts were reviewed, 155 were excluded, and 100 were included. Fifty patients, had stage 3a CKD and 50 had stage 3b CKD. Baseline demographics were similar between the stage 3a and 3b groups (Table 1). Both groups were predominantly White and male, with mean age > 70 years.

The primary endpoint was the differences in HbA_1c levels over time and between groups for patients with stage 3a and stage 3b CKD 1 year after initiation of empagliflozin. The starting doses of empagliflozin were either 12.5 mg or 25.0 mg. For both groups, the changes in HbA_1c levels were statistically significant (Table 2). HbA_1c levels dropped 0.65% for the stage 3a group and 0.48% for the 3b group. When compared to one another, the results were not statistically significant (P = .51).

Secondary Endpoint

There was no statistically significant difference in serum creatinine levels within each group between baselines and 1 year later for the stage 3a (P = .21) and stage 3b (P = .22) groups, or when compared to each other (P = .67). There were statistically significant changes in weight for patients in the stage 3a group (P < .05), but not for stage 3b group (P = .06) or when compared to each other (P = .41). A statistically significant change in systolic blood pressure was observed for the stage 3a group (P = .003), but not the stage 3b group (P = .16) or when compared to each other (P = .27). There were statistically significant changes in diastolic blood pressure within the stage 3a group (P = .04), but not within the stage 3b group (P = .61) or when compared to each other (P = .31).

Ten patients discontinued empagliflozin before the 1-year mark due to ADRs, including dizziness, increased incidence of urinary tract infections, rash, and tachycardia (Table 3). Additionally, 3 ADRs resulted in the empagliflozin discontinuation after 1 year (Table 3).

Discussion

This study showed a statistically significant change in HbA_1c levels for patients with stage 3a and stage 3b CKD. With eGFR levels in these 2 groups > 30 mL/min/1.73 m², patients were able to achieve glycemic benefits. There were no significant changes to the serum creatinine levels. Both groups saw statistically significant changes in weight loss within their own group; however, there were no statistically significant changes when compared to each other. With both systolic and diastolic blood pressure, the stage 3a group had statistically significant changes.

The EMPA-REG BP study demonstrated that empagliflozin was associated with significant and clinically meaningful reductions in blood pressure and HbA_1c levels compared with placebo and was well tolerated in patients with T2DM and hypertension.^6,7,8

Limitations

This study had a retrospective study design, which resulted in missing information for many patients and higher rates of exclusion. The population was predominantly older, White, and male and may not reflect other populations. The starting doses of empagliflozin varied between the groups. The VA employs tablet splitting for some patients, and the available doses were either 10.0 mg, 12.5 mg, or 25.0 mg. Some prescribers start veterans at lower doses and gradually increase to the higher dose of 25.0 mg, adding to the variability in starting doses.

Patients with eGFR < 30 mL/min/1.73 m² make it difficult to determine any potential benefit in this population. The EMPA-KIDNEY trial demonstrated that the benefits of empagliflozin treatment were consistent among patients with or without DM and regardless of eGFR at randomization.⁹ Furthermore, many veterans had an initial HbA_1c levels outside the inclusion criteria range, which was a factor in the smaller sample size.

Conclusions

While the reduction in HbA_1c levels was less in patients with stage 3b CKD compared to patients stage 3a CKD, all patients experienced a benefit. The overall incidence of ADRs was low in the study population, showing empagliflozin as a favorable choice for those with T2DM and CKD. Based on the findings of this study, empagliflozin is a potentially beneficial option for reducing HbA_1c levels in patients with CKD.

As large-scale natural disasters become more common, health care coalitions and the engagement of health systems with local, state, and federal public health departments have effectively bolstered communities’ resilience via collective sharing and distribution of resources.¹ These resources may include supplies and the dissemination of emergency information, education, and training.² The COVID-19 pandemic demonstrated that larger health care systems including hospital networks and nursing homes are better connected to health care coalition resources than smaller, independent systems, such as community home health agencies.³ This leaves some organizations on their own to meet requirements that maintain continuity of care and support their patients and staff throughout a natural disaster.

Home health care workers play important roles in the care of older adults.⁴ Older adults experience high levels of disability and comorbidities that put them at risk during emergencies; they often require support from paid, family, and neighborhood caregivers to live independently.⁵ More than 9.3 million US adults receive paid care from 2.6 million home health care workers (eg, home health aides and personal care assistants).⁶ Many of these individuals are hired through small independent home health agencies (HHAs), while others may work directly for an individual. When neighborhood resources and family caregiving are disrupted during emergencies, the critical services these workers administer become even more essential to ensuring continued access to medical care and social services.

The importance of these services was underscored by the Centers for Medicare and Medicaid Services 2017 inclusion of HHAs in federal emergency preparedness guidelines.^7,8 The fractured and decentralized nature of the home health care industry means many HHAs struggle to maintain continuous care during emergencies and protect their staff. HHAs, and health care workers in the home, are often isolated, under-resourced, and disconnected from broader emergency planning efforts. Additionally, home care jobs are largely part-time, unstable, and low paying, making the workers themselves vulnerable during emergencies.^3,9-13

This is a significant issue for the Veterans Health Administration (VHA), which annually purchases 10.5 million home health care worker visits for 150,000 veterans from community-based HHAs to enable those individuals to live independently. Figure 1 illustrates the existing structure of directly provided and contracted VHA services for community-dwelling veterans, highlighting the circle of care around the veteran.^8,9 Home health care workers anchored health care teams during the COVID-19 pandemic, observing and reporting on patients’ well-being to family caregivers, primary care practitioners, and HHAs. They also provided critical emotional support and companionship to patients isolated from family and friends.⁹ These workers also exposed themselves and their families to considerable risk and often lacked the protection afforded by personal protective equipment (PPE) in accordance with infection prevention guidance.^3,12

Through a combination of its national and local health care networks, the VHA has a robust and well-positioned emergency infrastructure to supportcommunity-dwelling older adults during disasters.¹⁴ This network is supported by the VHA Office of Emergency Management, which shares resources and guidance with local emergency managers at each facility as well as individual programs such as the VHA Home Based Primary Care (HBPC) program, which provides 38,000 seriously ill veterans with home medical visits.¹⁵ Working closely with their local and national hospital networks and emergency managers, individual VHA HBPC programs were able to maintain the safety of staff and continuity of care for patients enrolled in HBPC by rapidly administering COVID-19 vaccines to patients, caregivers, and staff, and providing emergency assistance during the 2017 hurricane season.^16,17 These efforts were successful because HBPC practitioners and their patients, had access to a level of emergency-related information, resources, and technology that are often out of reach for individual community-based health care practitioners (HCPs). The US Department of Veterans Affairs (VA) also supports local communities through its Fourth Mission, which provides emergency resources to non-VHA health care facilities (ie, hospitals and nursing homes) during national emergencies and natural disasters.¹⁷ Although there has been an expansion in the definition of shared resources, such as extending behavioral health support to local communities, the VHA has not historically provided these resources to HHAs.¹⁴

This study examines opportunities to leverage VHA emergency management resources to support contracted HHAs and inform other large health system emergency planning efforts. The findings from the exploratory phase are described in this article. We interviewed VHA emergency managers, HBPC and VA staff who coordinate home health care worker services, as well as administrators at contracted HHAs within a Veterans Integrated Services Network (VISN). These findings will inform the second (single-site pilot study) and third (feasibility study) phases. Our intent was to (1) better understand the relationships between VA medical centers (VAMCs) and their contracted HHAs; (2) identify existing VHA emergency protocols to support community-dwelling older adults; and (3) determine opportunities to build on existing infrastructure and relationships to better support contracted HHAs and their staff in emergencies.

Methods

The 18 VISNs act as regional systems of care that are loosely connected to better meet local health needs and maximize access to care. This study was conducted at 6 of 9 VAMCs within VISN 2, the New York/New Jersey VHA Health Care Network.¹⁸ VAMCs that serve urban, rural, and mixed urban/rural catchment areas were included.

Each VAMC has an emergency management program led by an emergency manager, an HBPC program led by a program director and medical director, and a community care or purchased care office that has a liaison who manages contracted home health care worker services. The studyfocused on HBPC programs because they are most likely to interact with veterans’ home health care workers in the home and care for community-dwelling veterans during emergencies. Each VHA also contracts with a series of local HHAs that generally have a dedicated staff member who interfaces with the VHA liaison. Our goal was to interview ≥ 1 emergency manager, ≥ 1 HBPC team member, ≥ 1 community care staff person, and ≥ 1 contracted home health agency administrator at each site to gain multiple perspectives from the range of HCPs serving veterans in the community.

Recruitment and Data Collection

The 6 sites were selected in consultation with VISN 2 leadership for their strong HBPC and emergency management programs. To recruit respondents, we contacted VISN and VAMC leads and used our professional networks to identify a sample of multidisciplinary individuals who represent both community care and HBPC programs who were contacted via email.

Since each VAMC is organized differently, we utilized a snowball sampling approach to identify the appropriate contacts.¹⁹ At the completion of each interview, we asked the participant to suggest additional contacts and introduce us to any remaining stakeholders (eg, the emergency manager) at that site or colleagues at other VISN facilities. Because roles vary among VAMCs, we contacted the person who most closely resembled the identified role and asked them to direct us to a more appropriate contact, if necessary. We asked community care managers to identify 1 to 2 agencies serving the highest volume of patients who are veterans at their site and requested interviews with those liaisons. This resulted in the recruitment of key stakeholders from 4 teams across the 6 sites (Table).

A semistructured interview guide was jointly developed based on constructs of interest, including relationships within VAMCs and between VAMCs and HHAs; existing emergency protocols and experience during disasters; and suggestions and opportunities for supporting agencies during emergencies and potential barriers. Two researchers (TWL and EF) who were trained in qualitative methods jointly conducted interviews using the interview guide, with 1 researcher leading and another taking notes and asking clarifying questions.

Interviews were conducted virtually via Microsoft Teams with respondents at their work locations between September 2022 and January 2023. Interviews were audio recorded and transcribed and 2 authors (TWL and ESO) reviewed transcripts for accuracy. Interviews averaged 47 minutes in length (range, 20-59).

The study was reviewed and determined to be exempt by institutional review boards at the James J. Peters VAMC and Greater Los Angeles VAMC. We asked participants for verbal consent to participate and preserved their confidentiality.

Analysis

Data were analyzed via an inductive approach, which involves drawing salient themes rather than imposing preconceived theories.²⁰ Three researchers (TWL, EF, and ES) listened to and discussed 2 staff interviews and tagged text with specific codes (eg, communication between the VHA and HHA, internal communication, and barriers to case fulfillment) so the team could selectively return to the interview text for deeper analysis, allowing for the development of a final codebook. The project team synthesized the findings to identify higher-level themes, drawing comparisons across and within the respondent groups, including within and between health care systems. Throughout the analysis, we maintained analytic memos, documented discussions, and engaged in analyst triangulation to ensure trustworthiness.^21,22 To ensure the analysis accurately reflected the participants’ understanding, we held 2 virtual member-checking sessions with participants to share preliminary findings and conclusions and solicit feedback. Analysis was conducted using ATLAS.ti version 20.

Results

VHA-based participants described internal emergency management systems that are deployed during a disaster to support patients and staff. Agency participants described their own internal emergency management protocols. Respondents discussed how and when the 2 intersected, as well as opportunities for future mutual support. The analysis identified several themes: (1) relationships between VAMC teams; (2) relationships between VHA and HHAs; (3) VHA and agencies responses during emergencies; (4) receptivity and opportunities for extending VHA resources into the community; and (5) barriers and facilitators to deeper engagement.

Relationships Within VHA (n = 17)

Staff at all VHA sites described close relationships between the internal emergency management and HBPC teams. HBPC teams identified patients who were most at risk during emergencies to triage those with the highest medical needs (eg, patients dependent on home infusion, oxygen, or electronic medical devices) and worked alongside emergency managers to develop plans to continue care during an emergency. HBPC representatives were part of their facilities’ local emergency response committees. Due to this close collaboration, VHA emergency managers were familiar with the needs of homebound veterans and caregivers. “I invite our [HBPC] program manager to attend [committee] meetings and … they’re part of the EOC [emergency operations center]," an emergency manager said. “We work together and I’m constantly in contact with that individual, especially during natural disasters and so forth, to ensure that everybody’s prepared in the community.”

On the other hand, community caremanagers—who described frequent interactions with HBPC teams, largely around coordinating and managing non-VHA home care services—were less likely to have direct relationships with their facility emergency managers. For example, when asked if they had a relationship with their emergency manager, a community care manager admitted, “I [only] know who he is.” They also did not report having structured protocols for veteran outreach during emergencies, “because all those veterans who are receiving [home health care worker] services also belong to a primary care team,” and considered the outreach to be the responsibility of the primary care team and HHA.

Relationships Between the VHA and HHAs (n = 17)

Communication between VAMCs and contracted agencies primarily went through community care managers, who described established long-term relationships with agency administrators. Communication was commonly restricted to operational activities, such as processing referrals and occasional troubleshooting. According to a community care manager most communication is “why haven’t you signed my orders?” There was a general sense from participants that communication was promptly answered, problems were addressed, and professional collegiality existed between the agencies as patients were referred and placed for services. One community care manager reported meeting with agencies regularly, noting, “I talk to them pretty much daily.”

If problems arose, community care managers described themselves as “the liaison” between agencies and VHA HCPs who ordered the referrals. This is particularly the case if the agency needed help finding a VHA clinician or addressing differences in care delivery protocols.

Responding During Emergencies (n = 19)

During emergencies, VHA and agency staff described following their own organization’s protocols and communicating with each other only on a case-by-case basis rather than through formal or systematic channels and had little knowledge of their counterpart’s emergency protocols. Beyond patient care, there was no evidence of information sharing between VHA and agency staff. Regarding sharing information with their local community, an HBPC Program Director said, “it’s almost like the VHA had become siloed” and operated on its own without engaging with community health systems or emergency managers.

Beyond the guidance provided by state departments of public health, HHAs described collaborating with other agencies in their network and relying on their informal professional network to manage the volume of information and updates they followed during emergencies like the COVID-19 pandemic. One agency administrator did not frequently communicate with VHA partners during the pandemic but explained that the local public health department helped work through challenges. However, “we realized pretty quickly they were overloaded and there was only so much they could do.” The agency administrator turned to a “sister agency” and local hospitals, noting, “Wherever you have connections in the field or in the industry, you know you’re going to reach out to people for guidance on policies and… protocol.”

Opportunities for Extending VHA Resources to the Community (n = 16)

All VHA emergency managers were receptive to extending support to community-based HCPS and, in some cases, felt strongly that they were an essential part of veterans’ care networks. Emergency managers offered examples for how they supportedcommunity-based HCPs, such as helping those in the VAMC medical foster home program develop and evaluate emergency plans. Many said they had not explicitly considered HHAs before (Appendix).

Emergency managers also described how supporting community-based HCPs could be considered within the scope of the VHA role and mission, specifically the Fourth Mission. “I think that we should be making our best effort to make sure that we’re also providing that same level [of protection] to the people taking care of the veteran [as our VHA staff],” an emergency manager said. “It’s our responsibility to provide the best for the staff that are going into those homes to take care of that patient.”

In many cases, emergency managers had already developed practical tools that could be easily shared outside the VHA, including weather alerts, trainings, emergency plan templates, and lists of community resources and shelters (Figure 2). A number of these examples built on existing communication channels. One emergency manager said that the extension of resources could be an opportunity to decrease the perceived isolation of home health care workers through regular training for agencies that are providing health care aides, so that they know that “some bigger folks are keeping an eye on it.”

On the agency side, participants noted that some HHAs could benefit more from support than others. While some agencies are well staffed and have good protocols and keep up to date, “There are smaller agencies, agencies that are starting up that may not have the resources to just disseminate all the information. Those are the agencies [that] could well benefit from the VHA,” an HBPC medical director explained. Agency administrators suggested several areas where they would welcome support, including a deeper understanding of available community resources and access to PPE for staff. Regarding informational resources, an administrator said, “Anytime we can get information, it’s good to have it come to you and not always have to go out searching for it.”

Barriers and Facilitators to Partnering With Community Agencies (n = 16)

A primary barrier regarding resource sharing was potential misalignment between each organization’s policies. HHAs followed state and federal public health guidelines, which sometimes differed from VHA policies. Given that agencies care for both VHA and non-VHA clients, questions also arose around how agencies would prioritize information from the VHA, if they were already receiving information from other sources. When asked about information sharing, both VHA staff and agencies agreed staff time to support any additional activities should be weighed against the value of the information gained.

Six participants also shared that education around emergency preparedness could be an opportunity to bridge gaps between VAMCs and their surrounding communities. One local Chief of Community Care noted, “Any opportunity to just give information is going to make it a lot better for the veteran patient … to have something that’s a little more robust.”

Two emergency managers noted the need to be sensitive in the way they engaged with partners, respecting and building on the work that agencies were already doing in this area to ensure VHA was seen as a trusted partner and resource rather than trying to impose new policies or rules on community-based HCPs. “I know that like all leadership in various organizations, there’s a little bit of bristling going on when other people try and tell them what to do,” an HBPC medical director said. “However, if it is established that as a sort of greater level like a state level or a federal level, that VHA can be a resource. I think that as long as that’s recognized by their own professional organizations within each state, then I think that that would be a tremendous advantage to many agencies.”

In terms of sharing physical resources, emergency managers raised concerns around potential liability, although they also acknowledged this issue was important enough to think about potential workarounds. As one emergency manager said, “I want to know that my PPE is not compromised in any way shape or form and that I am in charge of that PPE, so to rely upon going to a home and hoping that [the PPE] wasn’t compromised … would kind of make me a little uneasy.” This emergency manager suggested possible solutions, such as creating a sealed PPE package to give directly to an aide.

Discussion

As the prevalence of climate-related disasters increases, the need to ensure the safety and independence of older adults during emergencies grows more urgent. Health systems must think beyond the direct services they provide and consider the community resources upon which their patients rely. While relationships did not formally exist between VHA emergency managers and community home health HCPs in the sample analyzed in this article, there is precedent and interest in supporting contracted home health agencies caring for veterans in the community. Although not historically part of the VA Fourth Mission, creating a pipeline of support for contracted HHAs by leveraging existing relationships and resources can potentially strengthen its mission to protect older veterans in emergencies, help them age safely in place, and provide a model for health systems to collaborate with community-based HCPs around emergency planning and response (Figure 3).²³

Existing research on the value of health care coalitions highlights the need for established and growing partnerships with a focus on ensuring they are value-added, which echoes concerns we heard in interviews.²⁴ Investment in community partnerships not only includes sharing supplies but also relying on bidirectional support that can be a trusted form of timely information.^1,25 The findings in this study exhibit strong communication practices within the VHA during periods of nonemergency and underscore the untapped value of the pre-existing relationship between VAMCs and their contracted HHAs as an area of potential growth for health care coalitions.

Sharing resources in a way that does not put new demands on partners contributes to the sustainability and value-added nature of coalitions. Examples include establishing new low-investment practices (ie, information sharing) that support capacity and compliance with existing requirements rather than create new responsibilities for either member of the coalition. The relationship between the VHA emergency managers and the VHA HBPC program can act as a guide. The emergency managers interviewed for this study are currently engaged with HBPC programs and therefore understand the needs of homebound older adults and their caregivers. Extending the information already available to the HBPC teams via existing channels strengthens workforce practices and increased security for the shared patient, even without direct relationships between emergency managers and agencies. It is important to understand the limitations of these practices, including concerns around conflicting federal and state mandates, legal concerns around the liability of sharing physical resources (such as PPE), and awareness that the objective is not for the VHA to increase burdens (eg, increasing compliance requirements) but rather to serve as a resource for a mutual population in a shared community.

Offering training and practical resources to HHA home health care workers can help them meet disaster preparedness requirements. This is particularly important considering the growing home care workforce shortages, a topic mentioned by all HBPC and community care participants interviewed for this study.^26,27 Home health care workers report feeling underprepared and isolated while on the job in normal conditions, a sentiment exacerbated by the COVID-19 pandemic.^3,10 Supporting these individuals may help them feel more prepared and connected to their work, improving stability and quality of care.

While these issues are priorities within the VHA, there is growing recognition at the state and federal level of the importance of including older adults and their HCPs in disaster preparedness and response.^5,28 The US Department of Health and Human Services, for example, includes older adults and organizations that serve them on its National Advisory Committee on Seniors and Disasters. The Senate version of the 2023 reauthorization of the Pandemic and All-Hazards Preparedness and Response Act included specific provisions to support community-dwelling older adults and people with disabilities, incorporating funding for community organizations to support continuity of services and avoid institutionalization in an emergency.²⁹ Other proposed legislation includes the Real Emergency Access for Aging and Disability Inclusion for Disasters Act, which would ensure the needs of older adults and people with disabilities are explicitly included in all phases of emergency planning and response.³⁰

The VHA expansion of the its VEText program to include disaster response is an effort to more efficiently extend outreach to older and vulnerable patients who are veterans.³¹ Given these growing efforts, the VHA and other health systems have an opportunity to expand internal emergency preparedness efforts to ensure the health and safety of individuals living in the community.

Limitations

VISN 2 has been a target of terrorism and other disasters. In addition to the sites being initially recruited for their strong emergency management protocols, this context may have biased respondents who are favorable to extending their resources into the community. At the time of recruitment, contracted HHAs were still experiencing staff shortages due to the COVID-19 pandemic, which limited the ability of agency staff to participate in interviews. Additionally, while the comprehensive exploration of VISN 2 facilities allows for confidence of the organizational structures described, the qualitative research design and small study sample, the study findings cannot be immediately generalized to all VISNs.

Conclusions

Many older veterans increasingly rely on home health care workers to age safely. The VHA, as a large national health care system and leader in emergency preparedness, could play an important role in supporting home health care workers and ameliorating their sense of isolation during emergencies and natural disasters. Leveraging existing resources and relationships may be a low-cost, low-effort opportunity to build higher-level interventions that support the needs of patients. Future research and work in this field, including the authors’ ongoing work, will expand agency participation and engage agency staff in conceptualizing pilot projects to ensure they are viable and feasible for the field.

As large-scale natural disasters become more common, health care coalitions and the engagement of health systems with local, state, and federal public health departments have effectively bolstered communities’ resilience via collective sharing and distribution of resources.¹ These resources may include supplies and the dissemination of emergency information, education, and training.² The COVID-19 pandemic demonstrated that larger health care systems including hospital networks and nursing homes are better connected to health care coalition resources than smaller, independent systems, such as community home health agencies.³ This leaves some organizations on their own to meet requirements that maintain continuity of care and support their patients and staff throughout a natural disaster.

Home health care workers play important roles in the care of older adults.⁴ Older adults experience high levels of disability and comorbidities that put them at risk during emergencies; they often require support from paid, family, and neighborhood caregivers to live independently.⁵ More than 9.3 million US adults receive paid care from 2.6 million home health care workers (eg, home health aides and personal care assistants).⁶ Many of these individuals are hired through small independent home health agencies (HHAs), while others may work directly for an individual. When neighborhood resources and family caregiving are disrupted during emergencies, the critical services these workers administer become even more essential to ensuring continued access to medical care and social services.

The importance of these services was underscored by the Centers for Medicare and Medicaid Services 2017 inclusion of HHAs in federal emergency preparedness guidelines.^7,8 The fractured and decentralized nature of the home health care industry means many HHAs struggle to maintain continuous care during emergencies and protect their staff. HHAs, and health care workers in the home, are often isolated, under-resourced, and disconnected from broader emergency planning efforts. Additionally, home care jobs are largely part-time, unstable, and low paying, making the workers themselves vulnerable during emergencies.^3,9-13

This is a significant issue for the Veterans Health Administration (VHA), which annually purchases 10.5 million home health care worker visits for 150,000 veterans from community-based HHAs to enable those individuals to live independently. Figure 1 illustrates the existing structure of directly provided and contracted VHA services for community-dwelling veterans, highlighting the circle of care around the veteran.^8,9 Home health care workers anchored health care teams during the COVID-19 pandemic, observing and reporting on patients’ well-being to family caregivers, primary care practitioners, and HHAs. They also provided critical emotional support and companionship to patients isolated from family and friends.⁹ These workers also exposed themselves and their families to considerable risk and often lacked the protection afforded by personal protective equipment (PPE) in accordance with infection prevention guidance.^3,12

Through a combination of its national and local health care networks, the VHA has a robust and well-positioned emergency infrastructure to supportcommunity-dwelling older adults during disasters.¹⁴ This network is supported by the VHA Office of Emergency Management, which shares resources and guidance with local emergency managers at each facility as well as individual programs such as the VHA Home Based Primary Care (HBPC) program, which provides 38,000 seriously ill veterans with home medical visits.¹⁵ Working closely with their local and national hospital networks and emergency managers, individual VHA HBPC programs were able to maintain the safety of staff and continuity of care for patients enrolled in HBPC by rapidly administering COVID-19 vaccines to patients, caregivers, and staff, and providing emergency assistance during the 2017 hurricane season.^16,17 These efforts were successful because HBPC practitioners and their patients, had access to a level of emergency-related information, resources, and technology that are often out of reach for individual community-based health care practitioners (HCPs). The US Department of Veterans Affairs (VA) also supports local communities through its Fourth Mission, which provides emergency resources to non-VHA health care facilities (ie, hospitals and nursing homes) during national emergencies and natural disasters.¹⁷ Although there has been an expansion in the definition of shared resources, such as extending behavioral health support to local communities, the VHA has not historically provided these resources to HHAs.¹⁴

This study examines opportunities to leverage VHA emergency management resources to support contracted HHAs and inform other large health system emergency planning efforts. The findings from the exploratory phase are described in this article. We interviewed VHA emergency managers, HBPC and VA staff who coordinate home health care worker services, as well as administrators at contracted HHAs within a Veterans Integrated Services Network (VISN). These findings will inform the second (single-site pilot study) and third (feasibility study) phases. Our intent was to (1) better understand the relationships between VA medical centers (VAMCs) and their contracted HHAs; (2) identify existing VHA emergency protocols to support community-dwelling older adults; and (3) determine opportunities to build on existing infrastructure and relationships to better support contracted HHAs and their staff in emergencies.

Methods

The 18 VISNs act as regional systems of care that are loosely connected to better meet local health needs and maximize access to care. This study was conducted at 6 of 9 VAMCs within VISN 2, the New York/New Jersey VHA Health Care Network.¹⁸ VAMCs that serve urban, rural, and mixed urban/rural catchment areas were included.

Each VAMC has an emergency management program led by an emergency manager, an HBPC program led by a program director and medical director, and a community care or purchased care office that has a liaison who manages contracted home health care worker services. The studyfocused on HBPC programs because they are most likely to interact with veterans’ home health care workers in the home and care for community-dwelling veterans during emergencies. Each VHA also contracts with a series of local HHAs that generally have a dedicated staff member who interfaces with the VHA liaison. Our goal was to interview ≥ 1 emergency manager, ≥ 1 HBPC team member, ≥ 1 community care staff person, and ≥ 1 contracted home health agency administrator at each site to gain multiple perspectives from the range of HCPs serving veterans in the community.

Recruitment and Data Collection

The 6 sites were selected in consultation with VISN 2 leadership for their strong HBPC and emergency management programs. To recruit respondents, we contacted VISN and VAMC leads and used our professional networks to identify a sample of multidisciplinary individuals who represent both community care and HBPC programs who were contacted via email.

Since each VAMC is organized differently, we utilized a snowball sampling approach to identify the appropriate contacts.¹⁹ At the completion of each interview, we asked the participant to suggest additional contacts and introduce us to any remaining stakeholders (eg, the emergency manager) at that site or colleagues at other VISN facilities. Because roles vary among VAMCs, we contacted the person who most closely resembled the identified role and asked them to direct us to a more appropriate contact, if necessary. We asked community care managers to identify 1 to 2 agencies serving the highest volume of patients who are veterans at their site and requested interviews with those liaisons. This resulted in the recruitment of key stakeholders from 4 teams across the 6 sites (Table).

A semistructured interview guide was jointly developed based on constructs of interest, including relationships within VAMCs and between VAMCs and HHAs; existing emergency protocols and experience during disasters; and suggestions and opportunities for supporting agencies during emergencies and potential barriers. Two researchers (TWL and EF) who were trained in qualitative methods jointly conducted interviews using the interview guide, with 1 researcher leading and another taking notes and asking clarifying questions.

Interviews were conducted virtually via Microsoft Teams with respondents at their work locations between September 2022 and January 2023. Interviews were audio recorded and transcribed and 2 authors (TWL and ESO) reviewed transcripts for accuracy. Interviews averaged 47 minutes in length (range, 20-59).

The study was reviewed and determined to be exempt by institutional review boards at the James J. Peters VAMC and Greater Los Angeles VAMC. We asked participants for verbal consent to participate and preserved their confidentiality.

Analysis

Data were analyzed via an inductive approach, which involves drawing salient themes rather than imposing preconceived theories.²⁰ Three researchers (TWL, EF, and ES) listened to and discussed 2 staff interviews and tagged text with specific codes (eg, communication between the VHA and HHA, internal communication, and barriers to case fulfillment) so the team could selectively return to the interview text for deeper analysis, allowing for the development of a final codebook. The project team synthesized the findings to identify higher-level themes, drawing comparisons across and within the respondent groups, including within and between health care systems. Throughout the analysis, we maintained analytic memos, documented discussions, and engaged in analyst triangulation to ensure trustworthiness.^21,22 To ensure the analysis accurately reflected the participants’ understanding, we held 2 virtual member-checking sessions with participants to share preliminary findings and conclusions and solicit feedback. Analysis was conducted using ATLAS.ti version 20.

Results

VHA-based participants described internal emergency management systems that are deployed during a disaster to support patients and staff. Agency participants described their own internal emergency management protocols. Respondents discussed how and when the 2 intersected, as well as opportunities for future mutual support. The analysis identified several themes: (1) relationships between VAMC teams; (2) relationships between VHA and HHAs; (3) VHA and agencies responses during emergencies; (4) receptivity and opportunities for extending VHA resources into the community; and (5) barriers and facilitators to deeper engagement.

Relationships Within VHA (n = 17)

Staff at all VHA sites described close relationships between the internal emergency management and HBPC teams. HBPC teams identified patients who were most at risk during emergencies to triage those with the highest medical needs (eg, patients dependent on home infusion, oxygen, or electronic medical devices) and worked alongside emergency managers to develop plans to continue care during an emergency. HBPC representatives were part of their facilities’ local emergency response committees. Due to this close collaboration, VHA emergency managers were familiar with the needs of homebound veterans and caregivers. “I invite our [HBPC] program manager to attend [committee] meetings and … they’re part of the EOC [emergency operations center]," an emergency manager said. “We work together and I’m constantly in contact with that individual, especially during natural disasters and so forth, to ensure that everybody’s prepared in the community.”

On the other hand, community caremanagers—who described frequent interactions with HBPC teams, largely around coordinating and managing non-VHA home care services—were less likely to have direct relationships with their facility emergency managers. For example, when asked if they had a relationship with their emergency manager, a community care manager admitted, “I [only] know who he is.” They also did not report having structured protocols for veteran outreach during emergencies, “because all those veterans who are receiving [home health care worker] services also belong to a primary care team,” and considered the outreach to be the responsibility of the primary care team and HHA.

Relationships Between the VHA and HHAs (n = 17)

Communication between VAMCs and contracted agencies primarily went through community care managers, who described established long-term relationships with agency administrators. Communication was commonly restricted to operational activities, such as processing referrals and occasional troubleshooting. According to a community care manager most communication is “why haven’t you signed my orders?” There was a general sense from participants that communication was promptly answered, problems were addressed, and professional collegiality existed between the agencies as patients were referred and placed for services. One community care manager reported meeting with agencies regularly, noting, “I talk to them pretty much daily.”

If problems arose, community care managers described themselves as “the liaison” between agencies and VHA HCPs who ordered the referrals. This is particularly the case if the agency needed help finding a VHA clinician or addressing differences in care delivery protocols.

Responding During Emergencies (n = 19)

During emergencies, VHA and agency staff described following their own organization’s protocols and communicating with each other only on a case-by-case basis rather than through formal or systematic channels and had little knowledge of their counterpart’s emergency protocols. Beyond patient care, there was no evidence of information sharing between VHA and agency staff. Regarding sharing information with their local community, an HBPC Program Director said, “it’s almost like the VHA had become siloed” and operated on its own without engaging with community health systems or emergency managers.

Beyond the guidance provided by state departments of public health, HHAs described collaborating with other agencies in their network and relying on their informal professional network to manage the volume of information and updates they followed during emergencies like the COVID-19 pandemic. One agency administrator did not frequently communicate with VHA partners during the pandemic but explained that the local public health department helped work through challenges. However, “we realized pretty quickly they were overloaded and there was only so much they could do.” The agency administrator turned to a “sister agency” and local hospitals, noting, “Wherever you have connections in the field or in the industry, you know you’re going to reach out to people for guidance on policies and… protocol.”

Opportunities for Extending VHA Resources to the Community (n = 16)

All VHA emergency managers were receptive to extending support to community-based HCPS and, in some cases, felt strongly that they were an essential part of veterans’ care networks. Emergency managers offered examples for how they supportedcommunity-based HCPs, such as helping those in the VAMC medical foster home program develop and evaluate emergency plans. Many said they had not explicitly considered HHAs before (Appendix).

Emergency managers also described how supporting community-based HCPs could be considered within the scope of the VHA role and mission, specifically the Fourth Mission. “I think that we should be making our best effort to make sure that we’re also providing that same level [of protection] to the people taking care of the veteran [as our VHA staff],” an emergency manager said. “It’s our responsibility to provide the best for the staff that are going into those homes to take care of that patient.”

In many cases, emergency managers had already developed practical tools that could be easily shared outside the VHA, including weather alerts, trainings, emergency plan templates, and lists of community resources and shelters (Figure 2). A number of these examples built on existing communication channels. One emergency manager said that the extension of resources could be an opportunity to decrease the perceived isolation of home health care workers through regular training for agencies that are providing health care aides, so that they know that “some bigger folks are keeping an eye on it.”

On the agency side, participants noted that some HHAs could benefit more from support than others. While some agencies are well staffed and have good protocols and keep up to date, “There are smaller agencies, agencies that are starting up that may not have the resources to just disseminate all the information. Those are the agencies [that] could well benefit from the VHA,” an HBPC medical director explained. Agency administrators suggested several areas where they would welcome support, including a deeper understanding of available community resources and access to PPE for staff. Regarding informational resources, an administrator said, “Anytime we can get information, it’s good to have it come to you and not always have to go out searching for it.”

Barriers and Facilitators to Partnering With Community Agencies (n = 16)

A primary barrier regarding resource sharing was potential misalignment between each organization’s policies. HHAs followed state and federal public health guidelines, which sometimes differed from VHA policies. Given that agencies care for both VHA and non-VHA clients, questions also arose around how agencies would prioritize information from the VHA, if they were already receiving information from other sources. When asked about information sharing, both VHA staff and agencies agreed staff time to support any additional activities should be weighed against the value of the information gained.

Six participants also shared that education around emergency preparedness could be an opportunity to bridge gaps between VAMCs and their surrounding communities. One local Chief of Community Care noted, “Any opportunity to just give information is going to make it a lot better for the veteran patient … to have something that’s a little more robust.”

Two emergency managers noted the need to be sensitive in the way they engaged with partners, respecting and building on the work that agencies were already doing in this area to ensure VHA was seen as a trusted partner and resource rather than trying to impose new policies or rules on community-based HCPs. “I know that like all leadership in various organizations, there’s a little bit of bristling going on when other people try and tell them what to do,” an HBPC medical director said. “However, if it is established that as a sort of greater level like a state level or a federal level, that VHA can be a resource. I think that as long as that’s recognized by their own professional organizations within each state, then I think that that would be a tremendous advantage to many agencies.”

In terms of sharing physical resources, emergency managers raised concerns around potential liability, although they also acknowledged this issue was important enough to think about potential workarounds. As one emergency manager said, “I want to know that my PPE is not compromised in any way shape or form and that I am in charge of that PPE, so to rely upon going to a home and hoping that [the PPE] wasn’t compromised … would kind of make me a little uneasy.” This emergency manager suggested possible solutions, such as creating a sealed PPE package to give directly to an aide.

Discussion

As the prevalence of climate-related disasters increases, the need to ensure the safety and independence of older adults during emergencies grows more urgent. Health systems must think beyond the direct services they provide and consider the community resources upon which their patients rely. While relationships did not formally exist between VHA emergency managers and community home health HCPs in the sample analyzed in this article, there is precedent and interest in supporting contracted home health agencies caring for veterans in the community. Although not historically part of the VA Fourth Mission, creating a pipeline of support for contracted HHAs by leveraging existing relationships and resources can potentially strengthen its mission to protect older veterans in emergencies, help them age safely in place, and provide a model for health systems to collaborate with community-based HCPs around emergency planning and response (Figure 3).²³

Existing research on the value of health care coalitions highlights the need for established and growing partnerships with a focus on ensuring they are value-added, which echoes concerns we heard in interviews.²⁴ Investment in community partnerships not only includes sharing supplies but also relying on bidirectional support that can be a trusted form of timely information.^1,25 The findings in this study exhibit strong communication practices within the VHA during periods of nonemergency and underscore the untapped value of the pre-existing relationship between VAMCs and their contracted HHAs as an area of potential growth for health care coalitions.

Sharing resources in a way that does not put new demands on partners contributes to the sustainability and value-added nature of coalitions. Examples include establishing new low-investment practices (ie, information sharing) that support capacity and compliance with existing requirements rather than create new responsibilities for either member of the coalition. The relationship between the VHA emergency managers and the VHA HBPC program can act as a guide. The emergency managers interviewed for this study are currently engaged with HBPC programs and therefore understand the needs of homebound older adults and their caregivers. Extending the information already available to the HBPC teams via existing channels strengthens workforce practices and increased security for the shared patient, even without direct relationships between emergency managers and agencies. It is important to understand the limitations of these practices, including concerns around conflicting federal and state mandates, legal concerns around the liability of sharing physical resources (such as PPE), and awareness that the objective is not for the VHA to increase burdens (eg, increasing compliance requirements) but rather to serve as a resource for a mutual population in a shared community.

Offering training and practical resources to HHA home health care workers can help them meet disaster preparedness requirements. This is particularly important considering the growing home care workforce shortages, a topic mentioned by all HBPC and community care participants interviewed for this study.^26,27 Home health care workers report feeling underprepared and isolated while on the job in normal conditions, a sentiment exacerbated by the COVID-19 pandemic.^3,10 Supporting these individuals may help them feel more prepared and connected to their work, improving stability and quality of care.

While these issues are priorities within the VHA, there is growing recognition at the state and federal level of the importance of including older adults and their HCPs in disaster preparedness and response.^5,28 The US Department of Health and Human Services, for example, includes older adults and organizations that serve them on its National Advisory Committee on Seniors and Disasters. The Senate version of the 2023 reauthorization of the Pandemic and All-Hazards Preparedness and Response Act included specific provisions to support community-dwelling older adults and people with disabilities, incorporating funding for community organizations to support continuity of services and avoid institutionalization in an emergency.²⁹ Other proposed legislation includes the Real Emergency Access for Aging and Disability Inclusion for Disasters Act, which would ensure the needs of older adults and people with disabilities are explicitly included in all phases of emergency planning and response.³⁰

The VHA expansion of the its VEText program to include disaster response is an effort to more efficiently extend outreach to older and vulnerable patients who are veterans.³¹ Given these growing efforts, the VHA and other health systems have an opportunity to expand internal emergency preparedness efforts to ensure the health and safety of individuals living in the community.

Limitations

VISN 2 has been a target of terrorism and other disasters. In addition to the sites being initially recruited for their strong emergency management protocols, this context may have biased respondents who are favorable to extending their resources into the community. At the time of recruitment, contracted HHAs were still experiencing staff shortages due to the COVID-19 pandemic, which limited the ability of agency staff to participate in interviews. Additionally, while the comprehensive exploration of VISN 2 facilities allows for confidence of the organizational structures described, the qualitative research design and small study sample, the study findings cannot be immediately generalized to all VISNs.

Conclusions

Many older veterans increasingly rely on home health care workers to age safely. The VHA, as a large national health care system and leader in emergency preparedness, could play an important role in supporting home health care workers and ameliorating their sense of isolation during emergencies and natural disasters. Leveraging existing resources and relationships may be a low-cost, low-effort opportunity to build higher-level interventions that support the needs of patients. Future research and work in this field, including the authors’ ongoing work, will expand agency participation and engage agency staff in conceptualizing pilot projects to ensure they are viable and feasible for the field.

As large-scale natural disasters become more common, health care coalitions and the engagement of health systems with local, state, and federal public health departments have effectively bolstered communities’ resilience via collective sharing and distribution of resources.¹ These resources may include supplies and the dissemination of emergency information, education, and training.² The COVID-19 pandemic demonstrated that larger health care systems including hospital networks and nursing homes are better connected to health care coalition resources than smaller, independent systems, such as community home health agencies.³ This leaves some organizations on their own to meet requirements that maintain continuity of care and support their patients and staff throughout a natural disaster.

Home health care workers play important roles in the care of older adults.⁴ Older adults experience high levels of disability and comorbidities that put them at risk during emergencies; they often require support from paid, family, and neighborhood caregivers to live independently.⁵ More than 9.3 million US adults receive paid care from 2.6 million home health care workers (eg, home health aides and personal care assistants).⁶ Many of these individuals are hired through small independent home health agencies (HHAs), while others may work directly for an individual. When neighborhood resources and family caregiving are disrupted during emergencies, the critical services these workers administer become even more essential to ensuring continued access to medical care and social services.

The importance of these services was underscored by the Centers for Medicare and Medicaid Services 2017 inclusion of HHAs in federal emergency preparedness guidelines.^7,8 The fractured and decentralized nature of the home health care industry means many HHAs struggle to maintain continuous care during emergencies and protect their staff. HHAs, and health care workers in the home, are often isolated, under-resourced, and disconnected from broader emergency planning efforts. Additionally, home care jobs are largely part-time, unstable, and low paying, making the workers themselves vulnerable during emergencies.^3,9-13

This is a significant issue for the Veterans Health Administration (VHA), which annually purchases 10.5 million home health care worker visits for 150,000 veterans from community-based HHAs to enable those individuals to live independently. Figure 1 illustrates the existing structure of directly provided and contracted VHA services for community-dwelling veterans, highlighting the circle of care around the veteran.^8,9 Home health care workers anchored health care teams during the COVID-19 pandemic, observing and reporting on patients’ well-being to family caregivers, primary care practitioners, and HHAs. They also provided critical emotional support and companionship to patients isolated from family and friends.⁹ These workers also exposed themselves and their families to considerable risk and often lacked the protection afforded by personal protective equipment (PPE) in accordance with infection prevention guidance.^3,12

Through a combination of its national and local health care networks, the VHA has a robust and well-positioned emergency infrastructure to supportcommunity-dwelling older adults during disasters.¹⁴ This network is supported by the VHA Office of Emergency Management, which shares resources and guidance with local emergency managers at each facility as well as individual programs such as the VHA Home Based Primary Care (HBPC) program, which provides 38,000 seriously ill veterans with home medical visits.¹⁵ Working closely with their local and national hospital networks and emergency managers, individual VHA HBPC programs were able to maintain the safety of staff and continuity of care for patients enrolled in HBPC by rapidly administering COVID-19 vaccines to patients, caregivers, and staff, and providing emergency assistance during the 2017 hurricane season.^16,17 These efforts were successful because HBPC practitioners and their patients, had access to a level of emergency-related information, resources, and technology that are often out of reach for individual community-based health care practitioners (HCPs). The US Department of Veterans Affairs (VA) also supports local communities through its Fourth Mission, which provides emergency resources to non-VHA health care facilities (ie, hospitals and nursing homes) during national emergencies and natural disasters.¹⁷ Although there has been an expansion in the definition of shared resources, such as extending behavioral health support to local communities, the VHA has not historically provided these resources to HHAs.¹⁴

This study examines opportunities to leverage VHA emergency management resources to support contracted HHAs and inform other large health system emergency planning efforts. The findings from the exploratory phase are described in this article. We interviewed VHA emergency managers, HBPC and VA staff who coordinate home health care worker services, as well as administrators at contracted HHAs within a Veterans Integrated Services Network (VISN). These findings will inform the second (single-site pilot study) and third (feasibility study) phases. Our intent was to (1) better understand the relationships between VA medical centers (VAMCs) and their contracted HHAs; (2) identify existing VHA emergency protocols to support community-dwelling older adults; and (3) determine opportunities to build on existing infrastructure and relationships to better support contracted HHAs and their staff in emergencies.

Methods

The 18 VISNs act as regional systems of care that are loosely connected to better meet local health needs and maximize access to care. This study was conducted at 6 of 9 VAMCs within VISN 2, the New York/New Jersey VHA Health Care Network.¹⁸ VAMCs that serve urban, rural, and mixed urban/rural catchment areas were included.

Each VAMC has an emergency management program led by an emergency manager, an HBPC program led by a program director and medical director, and a community care or purchased care office that has a liaison who manages contracted home health care worker services. The studyfocused on HBPC programs because they are most likely to interact with veterans’ home health care workers in the home and care for community-dwelling veterans during emergencies. Each VHA also contracts with a series of local HHAs that generally have a dedicated staff member who interfaces with the VHA liaison. Our goal was to interview ≥ 1 emergency manager, ≥ 1 HBPC team member, ≥ 1 community care staff person, and ≥ 1 contracted home health agency administrator at each site to gain multiple perspectives from the range of HCPs serving veterans in the community.

Recruitment and Data Collection

The 6 sites were selected in consultation with VISN 2 leadership for their strong HBPC and emergency management programs. To recruit respondents, we contacted VISN and VAMC leads and used our professional networks to identify a sample of multidisciplinary individuals who represent both community care and HBPC programs who were contacted via email.

Since each VAMC is organized differently, we utilized a snowball sampling approach to identify the appropriate contacts.¹⁹ At the completion of each interview, we asked the participant to suggest additional contacts and introduce us to any remaining stakeholders (eg, the emergency manager) at that site or colleagues at other VISN facilities. Because roles vary among VAMCs, we contacted the person who most closely resembled the identified role and asked them to direct us to a more appropriate contact, if necessary. We asked community care managers to identify 1 to 2 agencies serving the highest volume of patients who are veterans at their site and requested interviews with those liaisons. This resulted in the recruitment of key stakeholders from 4 teams across the 6 sites (Table).

A semistructured interview guide was jointly developed based on constructs of interest, including relationships within VAMCs and between VAMCs and HHAs; existing emergency protocols and experience during disasters; and suggestions and opportunities for supporting agencies during emergencies and potential barriers. Two researchers (TWL and EF) who were trained in qualitative methods jointly conducted interviews using the interview guide, with 1 researcher leading and another taking notes and asking clarifying questions.

Interviews were conducted virtually via Microsoft Teams with respondents at their work locations between September 2022 and January 2023. Interviews were audio recorded and transcribed and 2 authors (TWL and ESO) reviewed transcripts for accuracy. Interviews averaged 47 minutes in length (range, 20-59).

The study was reviewed and determined to be exempt by institutional review boards at the James J. Peters VAMC and Greater Los Angeles VAMC. We asked participants for verbal consent to participate and preserved their confidentiality.

Analysis

Data were analyzed via an inductive approach, which involves drawing salient themes rather than imposing preconceived theories.²⁰ Three researchers (TWL, EF, and ES) listened to and discussed 2 staff interviews and tagged text with specific codes (eg, communication between the VHA and HHA, internal communication, and barriers to case fulfillment) so the team could selectively return to the interview text for deeper analysis, allowing for the development of a final codebook. The project team synthesized the findings to identify higher-level themes, drawing comparisons across and within the respondent groups, including within and between health care systems. Throughout the analysis, we maintained analytic memos, documented discussions, and engaged in analyst triangulation to ensure trustworthiness.^21,22 To ensure the analysis accurately reflected the participants’ understanding, we held 2 virtual member-checking sessions with participants to share preliminary findings and conclusions and solicit feedback. Analysis was conducted using ATLAS.ti version 20.

Results

VHA-based participants described internal emergency management systems that are deployed during a disaster to support patients and staff. Agency participants described their own internal emergency management protocols. Respondents discussed how and when the 2 intersected, as well as opportunities for future mutual support. The analysis identified several themes: (1) relationships between VAMC teams; (2) relationships between VHA and HHAs; (3) VHA and agencies responses during emergencies; (4) receptivity and opportunities for extending VHA resources into the community; and (5) barriers and facilitators to deeper engagement.

Relationships Within VHA (n = 17)

Staff at all VHA sites described close relationships between the internal emergency management and HBPC teams. HBPC teams identified patients who were most at risk during emergencies to triage those with the highest medical needs (eg, patients dependent on home infusion, oxygen, or electronic medical devices) and worked alongside emergency managers to develop plans to continue care during an emergency. HBPC representatives were part of their facilities’ local emergency response committees. Due to this close collaboration, VHA emergency managers were familiar with the needs of homebound veterans and caregivers. “I invite our [HBPC] program manager to attend [committee] meetings and … they’re part of the EOC [emergency operations center]," an emergency manager said. “We work together and I’m constantly in contact with that individual, especially during natural disasters and so forth, to ensure that everybody’s prepared in the community.”

On the other hand, community caremanagers—who described frequent interactions with HBPC teams, largely around coordinating and managing non-VHA home care services—were less likely to have direct relationships with their facility emergency managers. For example, when asked if they had a relationship with their emergency manager, a community care manager admitted, “I [only] know who he is.” They also did not report having structured protocols for veteran outreach during emergencies, “because all those veterans who are receiving [home health care worker] services also belong to a primary care team,” and considered the outreach to be the responsibility of the primary care team and HHA.

Relationships Between the VHA and HHAs (n = 17)

Communication between VAMCs and contracted agencies primarily went through community care managers, who described established long-term relationships with agency administrators. Communication was commonly restricted to operational activities, such as processing referrals and occasional troubleshooting. According to a community care manager most communication is “why haven’t you signed my orders?” There was a general sense from participants that communication was promptly answered, problems were addressed, and professional collegiality existed between the agencies as patients were referred and placed for services. One community care manager reported meeting with agencies regularly, noting, “I talk to them pretty much daily.”

If problems arose, community care managers described themselves as “the liaison” between agencies and VHA HCPs who ordered the referrals. This is particularly the case if the agency needed help finding a VHA clinician or addressing differences in care delivery protocols.

Responding During Emergencies (n = 19)

During emergencies, VHA and agency staff described following their own organization’s protocols and communicating with each other only on a case-by-case basis rather than through formal or systematic channels and had little knowledge of their counterpart’s emergency protocols. Beyond patient care, there was no evidence of information sharing between VHA and agency staff. Regarding sharing information with their local community, an HBPC Program Director said, “it’s almost like the VHA had become siloed” and operated on its own without engaging with community health systems or emergency managers.

Beyond the guidance provided by state departments of public health, HHAs described collaborating with other agencies in their network and relying on their informal professional network to manage the volume of information and updates they followed during emergencies like the COVID-19 pandemic. One agency administrator did not frequently communicate with VHA partners during the pandemic but explained that the local public health department helped work through challenges. However, “we realized pretty quickly they were overloaded and there was only so much they could do.” The agency administrator turned to a “sister agency” and local hospitals, noting, “Wherever you have connections in the field or in the industry, you know you’re going to reach out to people for guidance on policies and… protocol.”

Opportunities for Extending VHA Resources to the Community (n = 16)

All VHA emergency managers were receptive to extending support to community-based HCPS and, in some cases, felt strongly that they were an essential part of veterans’ care networks. Emergency managers offered examples for how they supportedcommunity-based HCPs, such as helping those in the VAMC medical foster home program develop and evaluate emergency plans. Many said they had not explicitly considered HHAs before (Appendix).

Emergency managers also described how supporting community-based HCPs could be considered within the scope of the VHA role and mission, specifically the Fourth Mission. “I think that we should be making our best effort to make sure that we’re also providing that same level [of protection] to the people taking care of the veteran [as our VHA staff],” an emergency manager said. “It’s our responsibility to provide the best for the staff that are going into those homes to take care of that patient.”

In many cases, emergency managers had already developed practical tools that could be easily shared outside the VHA, including weather alerts, trainings, emergency plan templates, and lists of community resources and shelters (Figure 2). A number of these examples built on existing communication channels. One emergency manager said that the extension of resources could be an opportunity to decrease the perceived isolation of home health care workers through regular training for agencies that are providing health care aides, so that they know that “some bigger folks are keeping an eye on it.”

On the agency side, participants noted that some HHAs could benefit more from support than others. While some agencies are well staffed and have good protocols and keep up to date, “There are smaller agencies, agencies that are starting up that may not have the resources to just disseminate all the information. Those are the agencies [that] could well benefit from the VHA,” an HBPC medical director explained. Agency administrators suggested several areas where they would welcome support, including a deeper understanding of available community resources and access to PPE for staff. Regarding informational resources, an administrator said, “Anytime we can get information, it’s good to have it come to you and not always have to go out searching for it.”

Barriers and Facilitators to Partnering With Community Agencies (n = 16)

A primary barrier regarding resource sharing was potential misalignment between each organization’s policies. HHAs followed state and federal public health guidelines, which sometimes differed from VHA policies. Given that agencies care for both VHA and non-VHA clients, questions also arose around how agencies would prioritize information from the VHA, if they were already receiving information from other sources. When asked about information sharing, both VHA staff and agencies agreed staff time to support any additional activities should be weighed against the value of the information gained.

Six participants also shared that education around emergency preparedness could be an opportunity to bridge gaps between VAMCs and their surrounding communities. One local Chief of Community Care noted, “Any opportunity to just give information is going to make it a lot better for the veteran patient … to have something that’s a little more robust.”

Two emergency managers noted the need to be sensitive in the way they engaged with partners, respecting and building on the work that agencies were already doing in this area to ensure VHA was seen as a trusted partner and resource rather than trying to impose new policies or rules on community-based HCPs. “I know that like all leadership in various organizations, there’s a little bit of bristling going on when other people try and tell them what to do,” an HBPC medical director said. “However, if it is established that as a sort of greater level like a state level or a federal level, that VHA can be a resource. I think that as long as that’s recognized by their own professional organizations within each state, then I think that that would be a tremendous advantage to many agencies.”

In terms of sharing physical resources, emergency managers raised concerns around potential liability, although they also acknowledged this issue was important enough to think about potential workarounds. As one emergency manager said, “I want to know that my PPE is not compromised in any way shape or form and that I am in charge of that PPE, so to rely upon going to a home and hoping that [the PPE] wasn’t compromised … would kind of make me a little uneasy.” This emergency manager suggested possible solutions, such as creating a sealed PPE package to give directly to an aide.

Discussion

As the prevalence of climate-related disasters increases, the need to ensure the safety and independence of older adults during emergencies grows more urgent. Health systems must think beyond the direct services they provide and consider the community resources upon which their patients rely. While relationships did not formally exist between VHA emergency managers and community home health HCPs in the sample analyzed in this article, there is precedent and interest in supporting contracted home health agencies caring for veterans in the community. Although not historically part of the VA Fourth Mission, creating a pipeline of support for contracted HHAs by leveraging existing relationships and resources can potentially strengthen its mission to protect older veterans in emergencies, help them age safely in place, and provide a model for health systems to collaborate with community-based HCPs around emergency planning and response (Figure 3).²³

Existing research on the value of health care coalitions highlights the need for established and growing partnerships with a focus on ensuring they are value-added, which echoes concerns we heard in interviews.²⁴ Investment in community partnerships not only includes sharing supplies but also relying on bidirectional support that can be a trusted form of timely information.^1,25 The findings in this study exhibit strong communication practices within the VHA during periods of nonemergency and underscore the untapped value of the pre-existing relationship between VAMCs and their contracted HHAs as an area of potential growth for health care coalitions.

Sharing resources in a way that does not put new demands on partners contributes to the sustainability and value-added nature of coalitions. Examples include establishing new low-investment practices (ie, information sharing) that support capacity and compliance with existing requirements rather than create new responsibilities for either member of the coalition. The relationship between the VHA emergency managers and the VHA HBPC program can act as a guide. The emergency managers interviewed for this study are currently engaged with HBPC programs and therefore understand the needs of homebound older adults and their caregivers. Extending the information already available to the HBPC teams via existing channels strengthens workforce practices and increased security for the shared patient, even without direct relationships between emergency managers and agencies. It is important to understand the limitations of these practices, including concerns around conflicting federal and state mandates, legal concerns around the liability of sharing physical resources (such as PPE), and awareness that the objective is not for the VHA to increase burdens (eg, increasing compliance requirements) but rather to serve as a resource for a mutual population in a shared community.

Offering training and practical resources to HHA home health care workers can help them meet disaster preparedness requirements. This is particularly important considering the growing home care workforce shortages, a topic mentioned by all HBPC and community care participants interviewed for this study.^26,27 Home health care workers report feeling underprepared and isolated while on the job in normal conditions, a sentiment exacerbated by the COVID-19 pandemic.^3,10 Supporting these individuals may help them feel more prepared and connected to their work, improving stability and quality of care.

While these issues are priorities within the VHA, there is growing recognition at the state and federal level of the importance of including older adults and their HCPs in disaster preparedness and response.^5,28 The US Department of Health and Human Services, for example, includes older adults and organizations that serve them on its National Advisory Committee on Seniors and Disasters. The Senate version of the 2023 reauthorization of the Pandemic and All-Hazards Preparedness and Response Act included specific provisions to support community-dwelling older adults and people with disabilities, incorporating funding for community organizations to support continuity of services and avoid institutionalization in an emergency.²⁹ Other proposed legislation includes the Real Emergency Access for Aging and Disability Inclusion for Disasters Act, which would ensure the needs of older adults and people with disabilities are explicitly included in all phases of emergency planning and response.³⁰

The VHA expansion of the its VEText program to include disaster response is an effort to more efficiently extend outreach to older and vulnerable patients who are veterans.³¹ Given these growing efforts, the VHA and other health systems have an opportunity to expand internal emergency preparedness efforts to ensure the health and safety of individuals living in the community.

Limitations

VISN 2 has been a target of terrorism and other disasters. In addition to the sites being initially recruited for their strong emergency management protocols, this context may have biased respondents who are favorable to extending their resources into the community. At the time of recruitment, contracted HHAs were still experiencing staff shortages due to the COVID-19 pandemic, which limited the ability of agency staff to participate in interviews. Additionally, while the comprehensive exploration of VISN 2 facilities allows for confidence of the organizational structures described, the qualitative research design and small study sample, the study findings cannot be immediately generalized to all VISNs.

Conclusions

Many older veterans increasingly rely on home health care workers to age safely. The VHA, as a large national health care system and leader in emergency preparedness, could play an important role in supporting home health care workers and ameliorating their sense of isolation during emergencies and natural disasters. Leveraging existing resources and relationships may be a low-cost, low-effort opportunity to build higher-level interventions that support the needs of patients. Future research and work in this field, including the authors’ ongoing work, will expand agency participation and engage agency staff in conceptualizing pilot projects to ensure they are viable and feasible for the field.