Joshua Freeman, MD

Diabetes, especially type II (non-insulin-dependent diabetes mellitus [NIDDM]) is extremely common among primary care patients. It accounts for a large percentage of office visits, and treatments for it are well described. However, caring for patients with diabetes is often a frustrating problem for clinicians.¹ A biopsychosocial understanding^2,3 of diabetes recognizes that medical outcomes, while occurring at the organ and cellular level, are affected by patient behaviors, and that these behaviors are largely determined by the patient’s world-view.^4-9 Physicians and other health professionals may have different world-views than their patients,^10-13 but must be able to communicate their ideas effectively and listen to what their patients are saying.

Ineffective communication has been identified as a barrier to effective treatment of diabetes.^6,14,15 Differing interpretations of medical terms, differences between the clinician’s personal feelings and professional ideas conveyed to the patient, and discrepancies between what the clinician emphasizes and what the patient thinks is important can obstruct effective communication and successful management of the disease.⁹

The authors of several recent articles^16-19 have pointed out the importance of clinician attitude or patient perception of physician behavior (eg, Dietrich¹⁶ argues that “the reaction and attitude of physicians displayed toward the patients at the point of diagnosis were crucial in influencing attitudes toward…the disease [diabetes] and consequently compliance.”), and some research even suggests that the physician’s attitudes toward diabetes management may be more important than actual knowledge of the disease.²⁰ Published literature, however, continues to focus almost exclusively on patient, rather than clinician, beliefs and practices.

The Journal of Family Practice o june 2000 o Vol. 49, No. 6 n 507y medicine, university of texas health science center at san antonio, 7703 floyd curl drive, san antonio, tx 78229-3900.

Methods

Setting

Our project focuses on clinicians and their patients in a busy family practice training site adjacent to a large public hospital in Chicago. The majority of the patient population at this site is poor, predominantly middle-aged or elderly, and sick. Approximately 70% of the patients are African American, 15% are Latino, and 15% are of Asian or Eastern European descent. Two other residency training sites, one a predominantly Latino clinic and the other an African American clinic, were also used to recruit patients and to observe precepting encounters and support groups.

We employed multiple qualitative methods (ie, interviews, informal conversation, and participant observation) to investigate the attitudes and practices of clinicians involved in the care of patients with diabetes. These enabled us to relate statements and actions to their ethnographic or clinical context and to enhance the validity of the data.^21,22 Although qualitative research has low reliability compared with experimental research, it has very strong face validity when it includes an observation component that enables the researcher to compare oral statements with actual practice. The presence of this element makes our study unique. The second methodologic objective was to collect data in multiple venues. This is important because physicians and patients are likely to say different things in different conversational settings; to make valid statements about attitudes and beliefs, therefore, it is necessary to sample broadly.

Semistructured Interviews with Clinicians and Patients

A semistructured interview guide for clinicians and a slightly modified guide for patients were developed to insure general consistency across interviews and to facilitate comparisons within and across groups. However, during the interviewing process, issues raised in earlier interviews were formulated as questions for subsequent interviewees (eg, Is fear an effective impetus to behavior change?), thus using the constant comparative method of Glaser and Strauss.²³ Interviewers were instructed to elicit stories from patients or particularly significant cases from clinicians.

Although the clinician sample was largely based on convenience, an attempt was made to include health professionals with different backgrounds; thus 2 nurses, 2 fellows, one physician’s assistant (PA), and one medical student were interviewed in addition to the family practice residents (n=9). The goal of the patient sample was to recruit and interview a patient mix representative of the ethnicities and sex of the patients in each clinic. Most patient interviews were done in the clinic before or after appointments, while clinician interviews were generally done in an office. Seventeen clinician and 22 patient interviews were taped and transcribed for review.

Participant Observation

The research team also observed 8 clinical precepting sessions involving family physician attendings and residents, students, or PAs who were seeing patients with diabetes. During these sessions the observers took field notes or tape-recorded the sessions; the encounters were written up or transcribed in their entirety and input into Nud.ist (Qualitative Solutions and Research, Melbourne, Australia), a qualitative software package, for subsequent analysis. Following the precepting sessions, the observers talked informally with the attendings about the encounters and other aspects of diabetes care.

Taped Interactions Between Clinicians and Patients

Three encounters between clinicians and patients were audiotaped or videotaped with the informed consent of both parties and transcribed for analysis.

Data Analysis

As transcripts became available they were analyzed for content by members of the study group, who coded the data independently and met to discuss their interpretations in monthly data-analysis sessions. Comments were coded using P (precepting session), I (interview), M (medical consultation), L (lecture), and C (informal conversation). The diversity of professional training in the group (a family physician, a psychologist/family therapist, 3 cultural anthropologists-2 of whom were also family therapists-and a PhD-level nurse brought multiple perspectives to a single data source, a qualitative technique known as researcher (or theoretical) triangulation.²¹ Most of the data collected through these 3 techniques was also input into Nud.ist and indexed according to format (interview, precepting session, and so forth), speaker characteristics (ethnicity, sex, professional status), clinical site, and other attributes.* The results presented here are grouped by themes derived from common content analysis techniques, such as counting, and clustering.²²

Results

The most significant themes revolved around control and frustration. Within these broad analytic categories, important subthemes that distinguish clinician narratives from those provided by patients include: the complexity of the disease; the importance of internal processes, such as nephropathy; long-term outcomes or end points; and profound pessimism. Also, while individual autonomy or control appears to be emphasized in the medical consultation, the impediments to control (social, cultural, and biological) and the importance of medication are noted in private discourses.

Control

A dominant, recurring theme, and a major source of confusion, was the use of the term “control.” This word was frequently used by clinicians and patients, with a wide range of meanings.

Metabolic Control. Both clinicians and patients talked about diabetes (the disease) being “in control” or “out of control,” meaning that the blood sugar or glycohemoglobin was in or out of the appropriate range: “His hemoglobin A1c is 10.8, his urine was 500 mg per deciliter of glucose…diabetes is not very well controlled” (resident, P).

Compliance. Clinicians also talked about the patients themselves being “out of control,” referring to patients’ inability or unwillingness to control their behaviors or diet. A resident uses the word in this sense when she points out that her patient has “no control…[she] had a donut and coffee for breakfast” (P).

Strategic or Situational Control. Finally, patients at times spoke of being in control in ways that clearly demonstrated their need to be in charge of their daily activities.

I was just talking to a doctor, and I said when I check my blood sugar before I eat it’s like 80 or maybe below…. He told me not to do that, and I, I want to do that and keep it low, because when I feel bad I want to go get something to eat and make that bad feeling go away…. For some reason the low blood sugar feeling is a lot easier to deal with than the high blood sugar; I think because I have control. (I)

Frustration

Physicians repeatedly expressed frustration in conjunction with both diabetes and caring for patients with diabetes. An attending physician said, “I think especially NIDDM is frustrating to take care of” (I), while a nurse said, “They’re [people with diabetes], I think, the hardest people to work with.” (I) A PA who works extensively with patients who have human immunodeficiency virus (HIV) remarked, “Diabetes is probably one of the most frustrating things I have to deal with…HIV patients are much easier to deal with.” (I)

Frustration was also tied to specific efforts by clinicians to perform patient education, particularly in the areas of diet and exercise. In this area clinicians highlighted their own lack of preparation or the difficulty they believed patients would have meeting tough dietary recommendations.

I have people who have a hard time following the dietary recommendations. They may not have enough money to buy [fresh] fruits and vegetables…. The first thing…is to talk to them about nutrition,…but we aren’t taught, as doctors, nutrition. (attending, I)

The diet’s the main thing, them following the diet. And it’s hard for me to teach them thinking I would have a hard time with this diet myself. (nurse, I)

I think most of them have 10 or 15 concerns that are ahead of the diabetes, so we’re having to get throughall those things before we hit the behavior change in dealing with diabetes. You know, child-care issues, transportation issues, violence. Everything. You’ve got to find out those things that are ahead of the diabetes. (PA, I)

Another major source of frustration is related to counseling patients about the risks of diabetes. In the face-to-face interviews, several clinicians emphasized the complexity of the disease or the difficulty of making asymptomatic patients aware of the risk of devastating consequences from “uncontrolled” diabetes.

For the most part there’s no symptoms…. They don’t feel better necessarily once they’re being treated. (attending)

It’s frustrating for the patients, because a lot of times they don’t understand…. There’s not necessarily symptoms associated with diabetes. (attending, I)

Diabetes encompasses so many things…their whole life. (nurse, I)

HIV patients see hope when they do something that makes a positive change in their life…whereas diabetics have a harder time getting through to that. (PA, I)

As a result of their frustration with attempts to educate and counsel patients, clinicians often relied on numbers and drugs to attempt to manage diabetes.

It’s hard to talk to them about nutrition, and you try to talk to them about exercise, and again a lot of times that’s hard for our patients who live in urban areas where they don’t want to walk outside, or there’s no place for them to exercise…so quickly you move to drugs. (attending, I)

Thus, an attending with a good “recipe” for teaching the “meds” was greatly appreciated. In an interview, a PA revealed that the head of the diabetes clinic was great at discussing “the meds,” and was a good mentor in clinic, but virtually ignored patient care issues.

I was taught by Dr X, and she is wonderful. She runs a diabetes clinic, and she’s real frank about talking about diabetes…. We talked a lot about treatments and, you know, what are the meds to begin with, but as far as behavior change and how to do any of that kind of stuff-the negotiating, diabetes management-we didn’t talk about that. A real cookbook type of thing…when we got in the clinic I think we saw the subtleties…or not-so-subtleties of treating diabetes. (I)

It appears that the formal teaching is about drugs; the behavioral management is picked up more informally, and therefore, may be perceived by the clinician-in-training as less valued. Thus, instead of education and counseling issues, the case presentations by resident physicians, students, and other clinicians to attending physicians tended to focus on numbers (the patient’s current blood sugar or glycohemoglobin level), compliance with treatment plans for oral hypoglycemic agents or insulin, and whether to increase drug therapy. As this represents the culturally expected behavior of physicians in clinical teaching settings, it took overt encouragement by the attending supervisor to focus their attention on other issues.

Long-Term Outcomes

Another major source of frustration for both clinicians and patients stems from the radical differences in the signs and symptoms that patients and clinicians focus on in treatment. Physicians are primarily concerned with the internal states or processes that lead to serious morbidity and death.

So I just want to go over some of the big complications and see how we can intervene; the biggest concern is diabetic nephropathy, usually very severe retinopathy…nephropathy. (L)

In contrast, the patients we spoke with seemed relatively unconcerned about the effect of diabetes on internal organs and rarely mentioned the heart, pancreas, or kidneys unless they were specifically asked about them. Patients did, however, express grave concern about wounds that would not heal, amputation, and external manifestations of the disease, such as blindness or other symptoms which reveal their condition to the public.

Interviewer: What do you fear most about diabetes?

Patient: Loss of limbs…. Now there is other things, I know kidneys can fail; I know that, and the heart can stop, I know that. But I’m worried about my feet…. Anytime [there’s] a strange color, I’m here. What’s wrong with this? (I)

Pessimism

Finally, physicians in many instances expressed a sense of pessimism or melancholy with of their experience of treating patients or their knowledge of likely medical outcomes.

So nephropathy, we believe, is related to the length of time the person has had diabetes and the control they’ve had over the years. Once a patient has gross proteinuria, end-stage renal failure or kidney transplant is within 5 years. What I’ve heard from the nephrologist is once you have microalbuminuria it’s like a 15-year type of progression from microalbuminuria to end-stage renal disease. The ophthalmologists believe that anyone who’s had diabetes over 10 years, you can detect some degree of retinopathy. (Fourth-year medical student, L)

I think medical students get the experience of diabetes in the hospital frequently, so they see diabetics that are having their second cardiac bypass, or having their legs cut off, or on dialysis. They see the worst end of diabetes instead of seeing the beginning. And so we come into working with diabetics without a lot of hope, and so we sort of go, “this is a horrible disease.” And someone who is diabetic doesn’t want to hear that. (attending, I)

Patients who do not have friends or relatives with severe complications associated with diabetes, however, are often hopeful or at least uncertain about what the future will bring.

Discussion

The Theme of Control

Among the themes revealed through content analysis, control is both the most common and the most complex. Used ubiquitously in both interviews and during observations of clinicians and patients, it has so many connotations that confusion is unavoidable. It is used by patients to refer to things the patient actually does to affect blood sugar as well as other outcomes (like his or her emotional state or the reactions of other people). In the physician narratives, “control” is generally thought of as an objective measurable condition, a range in which a patient’s blood sugar is thought to be relatively healthy. Where personal or subjective factors enter the equation, “compliance” is the operative physician term, although slippage between the 2 terms-compliance and control-is not uncommon. The ambiguity surrounding the term “control” highlights the fact that clinicians are uncertain about the agency or pathophysiology of this disease. Is the patient responsible for poor metabolic control, or do bad outcomes happen for reasons beyond his or her control? Are clinicians and patients in agreement about what might be in the patient’s control? The answers to these questions have major implications for the management of diabetes. If poor control is viewed as a problem of poor compliance (compliance used in the sense of doing something voluntarily and relatively easily), it is easy to blame the patient, especially if he or she is obese, poor, unemployed, or poorly educated. If poor control is not a result of poor compliance, however, or if compliance, as defined by the physician, beyond the practical ability of the patient, then the physician will look for other strategies to increase metabolic control.

Control represents a core value in physician culture^11,24 rather than indisputable scientific wisdom. Tight control of blood sugar levels is a principle on which doctors have acted since long before the recent studies showing its effectiveness for preventing poor long-term outcomes.^25,26

The Themes of Frustration, Long-Term Outcomes, and Pessimism

The frustration experienced by physicians caring for patients with diabetes results from many things, some of which are clearly avoidable. At the beginning of the research process, we hypothesized that this frustration was primarily due to the difference in the frames^27,28 or explanatory models^29,30 that patients and physicians use in thinking about diabetes. This hypothesis was borne out by radical differences in the symptoms and signs the respective parties focus on during treatment, most notably the distinction between internal and external signs.

This hypothesis is also justified by the differences between the perspective of patients who generally do not know where the disease will end up, and the more teleological perspective of the physician who can measure and time the breakdown of organ systems. From the physician’s viewpoint, that breakdown is more likely when control is less than optimal. This perspective leads to a profound pessimism, which not only belies the more optimistic tone of the medical consultation but leads to a series of conundrums. For example, while physicians privately acknowledge their doubts about the benefits of glycemic control or whether it is even possible for their patients will lose weight, they often end up blaming their patients for negative outcomes.

Patients who have symptoms that they can see, focus on those symptoms. They can become confused and frustrated when clinicians focus on less obvious internal conditions. Patients without symptoms feel they are being asked by clinicians to eat unpalatable food or perform unpleasant exercise in the pursuit of a vague hypothetical outcome (eg, avoiding nephropathy). None of the clinicians indicated that they involved their patients in deciding what a desirable outcome would be, but rather stipulated an outcome based on their own beliefs. It is critical that the patients be involved in weighing the risks and benefits of treatment to come up with a practical plan.

It is by no means certain that all patients would choose the same outcomes as desirable. When given the degree of work (and, in their view, discomfort or even suffering) needed to decrease by half an already small probability of ending up on dialysis, it is conceivable that some patients would see such possible outcomes as inadequate justification for denying themselves their favorite foods. Because the treatment plan involves minimal consultation with the patient in pursuit of a goal set by the clinician, it is not surprising that the outcome is often frustration for both.

Conclusions

Clinician attitudes toward diabetes and their patients with diabetes are derived from a number of sources. One source is the culture of medicine, which is largely influenced by a shared set of medical beliefs not always grounded in science or scientific data based on current research. These attitudes and the actions that originate with them are often markedly different from the attitudes of people with diabetes. Even when clinicians voice concern for patient belief issues in the interview setting, they rarely address these beliefs as priority issues in their practice, focusing instead on numbers and drugs. Also, clinician attitudes are usually the sole or primary determinant of what outcome will be pursued. This culturally insular approach leaves patients with little input into their own care plan, and their frequent failure to follow the clinician’s plan leads to major frustration for both parties. Although the macrovascular complications of diabetes (heart disease, stroke) are the major causes of mortality in people with this disease,³¹ in our study they were rarely addressed by providers or patients, which suggests that these well-known and serious conditions were probably not tied to diabetes in the minds of clinicians or patients.

Clinicians should become aware of their own beliefs, understand that they have cultural as well as scientific roots, open themselves to learning the beliefs and attitudes of their patients with diabetes, and recognize that the ultimate decision about the outcome to be pursued is the patient’s, informed by the clinician’s scientific knowledge. In particular, this knowledge includes current clinical guidelines about the relationship of heart disease and blood pressure control with diabetes.^32-35 Clinicians can seek a middle ground between unilaterally imposing external clinical guidelines on an unaccepting patient and adopting a near fatalistic individual relativism that does not seek to help people change behaviors that can adversely affect their health.

Such an approach will allow patients to use information offered by the clinician to identify the outcome they hope to achieve. Then the patient and clinician can work together to design a plan that will realize this desired outcome. It is likely that such an approach will decrease the frustration experienced by both patients and clinicians and enhance the collaborative effort to minimize the impact of diabetes

Acknowledgments

The authors wish to acknowledge the generous financial support of the American Academy of Family Physicians Foundation. Special thanks to Sean O’Sullivan, PhD; Lauretta Quinn, PhD, John Schwartzman, PhD; Steven Zuckerman, PhD; and Eve Pinsker, PhD, for participating in the data analysis sessions, to Anamari Golf, MA, and Dionne Hart for conducting interviews with patients with diabetes, and to Anne Larme, PhD, and Denise Stallcup, MA, for their invaluable feedback on earlier versions of this manuscript. Finally, we would like to thank Stephanie Tillman for coordinating research activities from start to finish and transcribing many of the interviews and precepting sessions.

Diabetes, especially type II (non-insulin-dependent diabetes mellitus [NIDDM]) is extremely common among primary care patients. It accounts for a large percentage of office visits, and treatments for it are well described. However, caring for patients with diabetes is often a frustrating problem for clinicians.¹ A biopsychosocial understanding^2,3 of diabetes recognizes that medical outcomes, while occurring at the organ and cellular level, are affected by patient behaviors, and that these behaviors are largely determined by the patient’s world-view.^4-9 Physicians and other health professionals may have different world-views than their patients,^10-13 but must be able to communicate their ideas effectively and listen to what their patients are saying.

Ineffective communication has been identified as a barrier to effective treatment of diabetes.^6,14,15 Differing interpretations of medical terms, differences between the clinician’s personal feelings and professional ideas conveyed to the patient, and discrepancies between what the clinician emphasizes and what the patient thinks is important can obstruct effective communication and successful management of the disease.⁹

The authors of several recent articles^16-19 have pointed out the importance of clinician attitude or patient perception of physician behavior (eg, Dietrich¹⁶ argues that “the reaction and attitude of physicians displayed toward the patients at the point of diagnosis were crucial in influencing attitudes toward…the disease [diabetes] and consequently compliance.”), and some research even suggests that the physician’s attitudes toward diabetes management may be more important than actual knowledge of the disease.²⁰ Published literature, however, continues to focus almost exclusively on patient, rather than clinician, beliefs and practices.

The Journal of Family Practice o june 2000 o Vol. 49, No. 6 n 507y medicine, university of texas health science center at san antonio, 7703 floyd curl drive, san antonio, tx 78229-3900.

Methods

Setting

Our project focuses on clinicians and their patients in a busy family practice training site adjacent to a large public hospital in Chicago. The majority of the patient population at this site is poor, predominantly middle-aged or elderly, and sick. Approximately 70% of the patients are African American, 15% are Latino, and 15% are of Asian or Eastern European descent. Two other residency training sites, one a predominantly Latino clinic and the other an African American clinic, were also used to recruit patients and to observe precepting encounters and support groups.

We employed multiple qualitative methods (ie, interviews, informal conversation, and participant observation) to investigate the attitudes and practices of clinicians involved in the care of patients with diabetes. These enabled us to relate statements and actions to their ethnographic or clinical context and to enhance the validity of the data.^21,22 Although qualitative research has low reliability compared with experimental research, it has very strong face validity when it includes an observation component that enables the researcher to compare oral statements with actual practice. The presence of this element makes our study unique. The second methodologic objective was to collect data in multiple venues. This is important because physicians and patients are likely to say different things in different conversational settings; to make valid statements about attitudes and beliefs, therefore, it is necessary to sample broadly.

Semistructured Interviews with Clinicians and Patients

A semistructured interview guide for clinicians and a slightly modified guide for patients were developed to insure general consistency across interviews and to facilitate comparisons within and across groups. However, during the interviewing process, issues raised in earlier interviews were formulated as questions for subsequent interviewees (eg, Is fear an effective impetus to behavior change?), thus using the constant comparative method of Glaser and Strauss.²³ Interviewers were instructed to elicit stories from patients or particularly significant cases from clinicians.

Although the clinician sample was largely based on convenience, an attempt was made to include health professionals with different backgrounds; thus 2 nurses, 2 fellows, one physician’s assistant (PA), and one medical student were interviewed in addition to the family practice residents (n=9). The goal of the patient sample was to recruit and interview a patient mix representative of the ethnicities and sex of the patients in each clinic. Most patient interviews were done in the clinic before or after appointments, while clinician interviews were generally done in an office. Seventeen clinician and 22 patient interviews were taped and transcribed for review.

Participant Observation

The research team also observed 8 clinical precepting sessions involving family physician attendings and residents, students, or PAs who were seeing patients with diabetes. During these sessions the observers took field notes or tape-recorded the sessions; the encounters were written up or transcribed in their entirety and input into Nud.ist (Qualitative Solutions and Research, Melbourne, Australia), a qualitative software package, for subsequent analysis. Following the precepting sessions, the observers talked informally with the attendings about the encounters and other aspects of diabetes care.

Taped Interactions Between Clinicians and Patients

Three encounters between clinicians and patients were audiotaped or videotaped with the informed consent of both parties and transcribed for analysis.

Data Analysis

As transcripts became available they were analyzed for content by members of the study group, who coded the data independently and met to discuss their interpretations in monthly data-analysis sessions. Comments were coded using P (precepting session), I (interview), M (medical consultation), L (lecture), and C (informal conversation). The diversity of professional training in the group (a family physician, a psychologist/family therapist, 3 cultural anthropologists-2 of whom were also family therapists-and a PhD-level nurse brought multiple perspectives to a single data source, a qualitative technique known as researcher (or theoretical) triangulation.²¹ Most of the data collected through these 3 techniques was also input into Nud.ist and indexed according to format (interview, precepting session, and so forth), speaker characteristics (ethnicity, sex, professional status), clinical site, and other attributes.* The results presented here are grouped by themes derived from common content analysis techniques, such as counting, and clustering.²²

Results

The most significant themes revolved around control and frustration. Within these broad analytic categories, important subthemes that distinguish clinician narratives from those provided by patients include: the complexity of the disease; the importance of internal processes, such as nephropathy; long-term outcomes or end points; and profound pessimism. Also, while individual autonomy or control appears to be emphasized in the medical consultation, the impediments to control (social, cultural, and biological) and the importance of medication are noted in private discourses.

Control

A dominant, recurring theme, and a major source of confusion, was the use of the term “control.” This word was frequently used by clinicians and patients, with a wide range of meanings.

Metabolic Control. Both clinicians and patients talked about diabetes (the disease) being “in control” or “out of control,” meaning that the blood sugar or glycohemoglobin was in or out of the appropriate range: “His hemoglobin A1c is 10.8, his urine was 500 mg per deciliter of glucose…diabetes is not very well controlled” (resident, P).

Compliance. Clinicians also talked about the patients themselves being “out of control,” referring to patients’ inability or unwillingness to control their behaviors or diet. A resident uses the word in this sense when she points out that her patient has “no control…[she] had a donut and coffee for breakfast” (P).

Strategic or Situational Control. Finally, patients at times spoke of being in control in ways that clearly demonstrated their need to be in charge of their daily activities.

I was just talking to a doctor, and I said when I check my blood sugar before I eat it’s like 80 or maybe below…. He told me not to do that, and I, I want to do that and keep it low, because when I feel bad I want to go get something to eat and make that bad feeling go away…. For some reason the low blood sugar feeling is a lot easier to deal with than the high blood sugar; I think because I have control. (I)

Frustration

Physicians repeatedly expressed frustration in conjunction with both diabetes and caring for patients with diabetes. An attending physician said, “I think especially NIDDM is frustrating to take care of” (I), while a nurse said, “They’re [people with diabetes], I think, the hardest people to work with.” (I) A PA who works extensively with patients who have human immunodeficiency virus (HIV) remarked, “Diabetes is probably one of the most frustrating things I have to deal with…HIV patients are much easier to deal with.” (I)

Frustration was also tied to specific efforts by clinicians to perform patient education, particularly in the areas of diet and exercise. In this area clinicians highlighted their own lack of preparation or the difficulty they believed patients would have meeting tough dietary recommendations.

I have people who have a hard time following the dietary recommendations. They may not have enough money to buy [fresh] fruits and vegetables…. The first thing…is to talk to them about nutrition,…but we aren’t taught, as doctors, nutrition. (attending, I)

The diet’s the main thing, them following the diet. And it’s hard for me to teach them thinking I would have a hard time with this diet myself. (nurse, I)

I think most of them have 10 or 15 concerns that are ahead of the diabetes, so we’re having to get throughall those things before we hit the behavior change in dealing with diabetes. You know, child-care issues, transportation issues, violence. Everything. You’ve got to find out those things that are ahead of the diabetes. (PA, I)

Another major source of frustration is related to counseling patients about the risks of diabetes. In the face-to-face interviews, several clinicians emphasized the complexity of the disease or the difficulty of making asymptomatic patients aware of the risk of devastating consequences from “uncontrolled” diabetes.

For the most part there’s no symptoms…. They don’t feel better necessarily once they’re being treated. (attending)

It’s frustrating for the patients, because a lot of times they don’t understand…. There’s not necessarily symptoms associated with diabetes. (attending, I)

Diabetes encompasses so many things…their whole life. (nurse, I)

HIV patients see hope when they do something that makes a positive change in their life…whereas diabetics have a harder time getting through to that. (PA, I)

As a result of their frustration with attempts to educate and counsel patients, clinicians often relied on numbers and drugs to attempt to manage diabetes.

It’s hard to talk to them about nutrition, and you try to talk to them about exercise, and again a lot of times that’s hard for our patients who live in urban areas where they don’t want to walk outside, or there’s no place for them to exercise…so quickly you move to drugs. (attending, I)

Thus, an attending with a good “recipe” for teaching the “meds” was greatly appreciated. In an interview, a PA revealed that the head of the diabetes clinic was great at discussing “the meds,” and was a good mentor in clinic, but virtually ignored patient care issues.

I was taught by Dr X, and she is wonderful. She runs a diabetes clinic, and she’s real frank about talking about diabetes…. We talked a lot about treatments and, you know, what are the meds to begin with, but as far as behavior change and how to do any of that kind of stuff-the negotiating, diabetes management-we didn’t talk about that. A real cookbook type of thing…when we got in the clinic I think we saw the subtleties…or not-so-subtleties of treating diabetes. (I)

It appears that the formal teaching is about drugs; the behavioral management is picked up more informally, and therefore, may be perceived by the clinician-in-training as less valued. Thus, instead of education and counseling issues, the case presentations by resident physicians, students, and other clinicians to attending physicians tended to focus on numbers (the patient’s current blood sugar or glycohemoglobin level), compliance with treatment plans for oral hypoglycemic agents or insulin, and whether to increase drug therapy. As this represents the culturally expected behavior of physicians in clinical teaching settings, it took overt encouragement by the attending supervisor to focus their attention on other issues.

Long-Term Outcomes

Another major source of frustration for both clinicians and patients stems from the radical differences in the signs and symptoms that patients and clinicians focus on in treatment. Physicians are primarily concerned with the internal states or processes that lead to serious morbidity and death.

So I just want to go over some of the big complications and see how we can intervene; the biggest concern is diabetic nephropathy, usually very severe retinopathy…nephropathy. (L)

In contrast, the patients we spoke with seemed relatively unconcerned about the effect of diabetes on internal organs and rarely mentioned the heart, pancreas, or kidneys unless they were specifically asked about them. Patients did, however, express grave concern about wounds that would not heal, amputation, and external manifestations of the disease, such as blindness or other symptoms which reveal their condition to the public.

Interviewer: What do you fear most about diabetes?

Patient: Loss of limbs…. Now there is other things, I know kidneys can fail; I know that, and the heart can stop, I know that. But I’m worried about my feet…. Anytime [there’s] a strange color, I’m here. What’s wrong with this? (I)

Pessimism

Finally, physicians in many instances expressed a sense of pessimism or melancholy with of their experience of treating patients or their knowledge of likely medical outcomes.

So nephropathy, we believe, is related to the length of time the person has had diabetes and the control they’ve had over the years. Once a patient has gross proteinuria, end-stage renal failure or kidney transplant is within 5 years. What I’ve heard from the nephrologist is once you have microalbuminuria it’s like a 15-year type of progression from microalbuminuria to end-stage renal disease. The ophthalmologists believe that anyone who’s had diabetes over 10 years, you can detect some degree of retinopathy. (Fourth-year medical student, L)

I think medical students get the experience of diabetes in the hospital frequently, so they see diabetics that are having their second cardiac bypass, or having their legs cut off, or on dialysis. They see the worst end of diabetes instead of seeing the beginning. And so we come into working with diabetics without a lot of hope, and so we sort of go, “this is a horrible disease.” And someone who is diabetic doesn’t want to hear that. (attending, I)

Patients who do not have friends or relatives with severe complications associated with diabetes, however, are often hopeful or at least uncertain about what the future will bring.

Discussion

The Theme of Control

Among the themes revealed through content analysis, control is both the most common and the most complex. Used ubiquitously in both interviews and during observations of clinicians and patients, it has so many connotations that confusion is unavoidable. It is used by patients to refer to things the patient actually does to affect blood sugar as well as other outcomes (like his or her emotional state or the reactions of other people). In the physician narratives, “control” is generally thought of as an objective measurable condition, a range in which a patient’s blood sugar is thought to be relatively healthy. Where personal or subjective factors enter the equation, “compliance” is the operative physician term, although slippage between the 2 terms-compliance and control-is not uncommon. The ambiguity surrounding the term “control” highlights the fact that clinicians are uncertain about the agency or pathophysiology of this disease. Is the patient responsible for poor metabolic control, or do bad outcomes happen for reasons beyond his or her control? Are clinicians and patients in agreement about what might be in the patient’s control? The answers to these questions have major implications for the management of diabetes. If poor control is viewed as a problem of poor compliance (compliance used in the sense of doing something voluntarily and relatively easily), it is easy to blame the patient, especially if he or she is obese, poor, unemployed, or poorly educated. If poor control is not a result of poor compliance, however, or if compliance, as defined by the physician, beyond the practical ability of the patient, then the physician will look for other strategies to increase metabolic control.

Control represents a core value in physician culture^11,24 rather than indisputable scientific wisdom. Tight control of blood sugar levels is a principle on which doctors have acted since long before the recent studies showing its effectiveness for preventing poor long-term outcomes.^25,26

The Themes of Frustration, Long-Term Outcomes, and Pessimism

The frustration experienced by physicians caring for patients with diabetes results from many things, some of which are clearly avoidable. At the beginning of the research process, we hypothesized that this frustration was primarily due to the difference in the frames^27,28 or explanatory models^29,30 that patients and physicians use in thinking about diabetes. This hypothesis was borne out by radical differences in the symptoms and signs the respective parties focus on during treatment, most notably the distinction between internal and external signs.

This hypothesis is also justified by the differences between the perspective of patients who generally do not know where the disease will end up, and the more teleological perspective of the physician who can measure and time the breakdown of organ systems. From the physician’s viewpoint, that breakdown is more likely when control is less than optimal. This perspective leads to a profound pessimism, which not only belies the more optimistic tone of the medical consultation but leads to a series of conundrums. For example, while physicians privately acknowledge their doubts about the benefits of glycemic control or whether it is even possible for their patients will lose weight, they often end up blaming their patients for negative outcomes.

Patients who have symptoms that they can see, focus on those symptoms. They can become confused and frustrated when clinicians focus on less obvious internal conditions. Patients without symptoms feel they are being asked by clinicians to eat unpalatable food or perform unpleasant exercise in the pursuit of a vague hypothetical outcome (eg, avoiding nephropathy). None of the clinicians indicated that they involved their patients in deciding what a desirable outcome would be, but rather stipulated an outcome based on their own beliefs. It is critical that the patients be involved in weighing the risks and benefits of treatment to come up with a practical plan.

It is by no means certain that all patients would choose the same outcomes as desirable. When given the degree of work (and, in their view, discomfort or even suffering) needed to decrease by half an already small probability of ending up on dialysis, it is conceivable that some patients would see such possible outcomes as inadequate justification for denying themselves their favorite foods. Because the treatment plan involves minimal consultation with the patient in pursuit of a goal set by the clinician, it is not surprising that the outcome is often frustration for both.

Conclusions

Clinician attitudes toward diabetes and their patients with diabetes are derived from a number of sources. One source is the culture of medicine, which is largely influenced by a shared set of medical beliefs not always grounded in science or scientific data based on current research. These attitudes and the actions that originate with them are often markedly different from the attitudes of people with diabetes. Even when clinicians voice concern for patient belief issues in the interview setting, they rarely address these beliefs as priority issues in their practice, focusing instead on numbers and drugs. Also, clinician attitudes are usually the sole or primary determinant of what outcome will be pursued. This culturally insular approach leaves patients with little input into their own care plan, and their frequent failure to follow the clinician’s plan leads to major frustration for both parties. Although the macrovascular complications of diabetes (heart disease, stroke) are the major causes of mortality in people with this disease,³¹ in our study they were rarely addressed by providers or patients, which suggests that these well-known and serious conditions were probably not tied to diabetes in the minds of clinicians or patients.

Clinicians should become aware of their own beliefs, understand that they have cultural as well as scientific roots, open themselves to learning the beliefs and attitudes of their patients with diabetes, and recognize that the ultimate decision about the outcome to be pursued is the patient’s, informed by the clinician’s scientific knowledge. In particular, this knowledge includes current clinical guidelines about the relationship of heart disease and blood pressure control with diabetes.^32-35 Clinicians can seek a middle ground between unilaterally imposing external clinical guidelines on an unaccepting patient and adopting a near fatalistic individual relativism that does not seek to help people change behaviors that can adversely affect their health.

Such an approach will allow patients to use information offered by the clinician to identify the outcome they hope to achieve. Then the patient and clinician can work together to design a plan that will realize this desired outcome. It is likely that such an approach will decrease the frustration experienced by both patients and clinicians and enhance the collaborative effort to minimize the impact of diabetes

Acknowledgments

The authors wish to acknowledge the generous financial support of the American Academy of Family Physicians Foundation. Special thanks to Sean O’Sullivan, PhD; Lauretta Quinn, PhD, John Schwartzman, PhD; Steven Zuckerman, PhD; and Eve Pinsker, PhD, for participating in the data analysis sessions, to Anamari Golf, MA, and Dionne Hart for conducting interviews with patients with diabetes, and to Anne Larme, PhD, and Denise Stallcup, MA, for their invaluable feedback on earlier versions of this manuscript. Finally, we would like to thank Stephanie Tillman for coordinating research activities from start to finish and transcribing many of the interviews and precepting sessions.

Diabetes, especially type II (non-insulin-dependent diabetes mellitus [NIDDM]) is extremely common among primary care patients. It accounts for a large percentage of office visits, and treatments for it are well described. However, caring for patients with diabetes is often a frustrating problem for clinicians.¹ A biopsychosocial understanding^2,3 of diabetes recognizes that medical outcomes, while occurring at the organ and cellular level, are affected by patient behaviors, and that these behaviors are largely determined by the patient’s world-view.^4-9 Physicians and other health professionals may have different world-views than their patients,^10-13 but must be able to communicate their ideas effectively and listen to what their patients are saying.

Ineffective communication has been identified as a barrier to effective treatment of diabetes.^6,14,15 Differing interpretations of medical terms, differences between the clinician’s personal feelings and professional ideas conveyed to the patient, and discrepancies between what the clinician emphasizes and what the patient thinks is important can obstruct effective communication and successful management of the disease.⁹

The authors of several recent articles^16-19 have pointed out the importance of clinician attitude or patient perception of physician behavior (eg, Dietrich¹⁶ argues that “the reaction and attitude of physicians displayed toward the patients at the point of diagnosis were crucial in influencing attitudes toward…the disease [diabetes] and consequently compliance.”), and some research even suggests that the physician’s attitudes toward diabetes management may be more important than actual knowledge of the disease.²⁰ Published literature, however, continues to focus almost exclusively on patient, rather than clinician, beliefs and practices.

The Journal of Family Practice o june 2000 o Vol. 49, No. 6 n 507y medicine, university of texas health science center at san antonio, 7703 floyd curl drive, san antonio, tx 78229-3900.

Methods

Setting

Our project focuses on clinicians and their patients in a busy family practice training site adjacent to a large public hospital in Chicago. The majority of the patient population at this site is poor, predominantly middle-aged or elderly, and sick. Approximately 70% of the patients are African American, 15% are Latino, and 15% are of Asian or Eastern European descent. Two other residency training sites, one a predominantly Latino clinic and the other an African American clinic, were also used to recruit patients and to observe precepting encounters and support groups.

We employed multiple qualitative methods (ie, interviews, informal conversation, and participant observation) to investigate the attitudes and practices of clinicians involved in the care of patients with diabetes. These enabled us to relate statements and actions to their ethnographic or clinical context and to enhance the validity of the data.^21,22 Although qualitative research has low reliability compared with experimental research, it has very strong face validity when it includes an observation component that enables the researcher to compare oral statements with actual practice. The presence of this element makes our study unique. The second methodologic objective was to collect data in multiple venues. This is important because physicians and patients are likely to say different things in different conversational settings; to make valid statements about attitudes and beliefs, therefore, it is necessary to sample broadly.

Semistructured Interviews with Clinicians and Patients

A semistructured interview guide for clinicians and a slightly modified guide for patients were developed to insure general consistency across interviews and to facilitate comparisons within and across groups. However, during the interviewing process, issues raised in earlier interviews were formulated as questions for subsequent interviewees (eg, Is fear an effective impetus to behavior change?), thus using the constant comparative method of Glaser and Strauss.²³ Interviewers were instructed to elicit stories from patients or particularly significant cases from clinicians.

Although the clinician sample was largely based on convenience, an attempt was made to include health professionals with different backgrounds; thus 2 nurses, 2 fellows, one physician’s assistant (PA), and one medical student were interviewed in addition to the family practice residents (n=9). The goal of the patient sample was to recruit and interview a patient mix representative of the ethnicities and sex of the patients in each clinic. Most patient interviews were done in the clinic before or after appointments, while clinician interviews were generally done in an office. Seventeen clinician and 22 patient interviews were taped and transcribed for review.

Participant Observation

The research team also observed 8 clinical precepting sessions involving family physician attendings and residents, students, or PAs who were seeing patients with diabetes. During these sessions the observers took field notes or tape-recorded the sessions; the encounters were written up or transcribed in their entirety and input into Nud.ist (Qualitative Solutions and Research, Melbourne, Australia), a qualitative software package, for subsequent analysis. Following the precepting sessions, the observers talked informally with the attendings about the encounters and other aspects of diabetes care.

Taped Interactions Between Clinicians and Patients

Three encounters between clinicians and patients were audiotaped or videotaped with the informed consent of both parties and transcribed for analysis.

Data Analysis

As transcripts became available they were analyzed for content by members of the study group, who coded the data independently and met to discuss their interpretations in monthly data-analysis sessions. Comments were coded using P (precepting session), I (interview), M (medical consultation), L (lecture), and C (informal conversation). The diversity of professional training in the group (a family physician, a psychologist/family therapist, 3 cultural anthropologists-2 of whom were also family therapists-and a PhD-level nurse brought multiple perspectives to a single data source, a qualitative technique known as researcher (or theoretical) triangulation.²¹ Most of the data collected through these 3 techniques was also input into Nud.ist and indexed according to format (interview, precepting session, and so forth), speaker characteristics (ethnicity, sex, professional status), clinical site, and other attributes.* The results presented here are grouped by themes derived from common content analysis techniques, such as counting, and clustering.²²

Results

The most significant themes revolved around control and frustration. Within these broad analytic categories, important subthemes that distinguish clinician narratives from those provided by patients include: the complexity of the disease; the importance of internal processes, such as nephropathy; long-term outcomes or end points; and profound pessimism. Also, while individual autonomy or control appears to be emphasized in the medical consultation, the impediments to control (social, cultural, and biological) and the importance of medication are noted in private discourses.

Control

A dominant, recurring theme, and a major source of confusion, was the use of the term “control.” This word was frequently used by clinicians and patients, with a wide range of meanings.

Metabolic Control. Both clinicians and patients talked about diabetes (the disease) being “in control” or “out of control,” meaning that the blood sugar or glycohemoglobin was in or out of the appropriate range: “His hemoglobin A1c is 10.8, his urine was 500 mg per deciliter of glucose…diabetes is not very well controlled” (resident, P).

Compliance. Clinicians also talked about the patients themselves being “out of control,” referring to patients’ inability or unwillingness to control their behaviors or diet. A resident uses the word in this sense when she points out that her patient has “no control…[she] had a donut and coffee for breakfast” (P).

Strategic or Situational Control. Finally, patients at times spoke of being in control in ways that clearly demonstrated their need to be in charge of their daily activities.

I was just talking to a doctor, and I said when I check my blood sugar before I eat it’s like 80 or maybe below…. He told me not to do that, and I, I want to do that and keep it low, because when I feel bad I want to go get something to eat and make that bad feeling go away…. For some reason the low blood sugar feeling is a lot easier to deal with than the high blood sugar; I think because I have control. (I)

Frustration

Physicians repeatedly expressed frustration in conjunction with both diabetes and caring for patients with diabetes. An attending physician said, “I think especially NIDDM is frustrating to take care of” (I), while a nurse said, “They’re [people with diabetes], I think, the hardest people to work with.” (I) A PA who works extensively with patients who have human immunodeficiency virus (HIV) remarked, “Diabetes is probably one of the most frustrating things I have to deal with…HIV patients are much easier to deal with.” (I)

Frustration was also tied to specific efforts by clinicians to perform patient education, particularly in the areas of diet and exercise. In this area clinicians highlighted their own lack of preparation or the difficulty they believed patients would have meeting tough dietary recommendations.

I have people who have a hard time following the dietary recommendations. They may not have enough money to buy [fresh] fruits and vegetables…. The first thing…is to talk to them about nutrition,…but we aren’t taught, as doctors, nutrition. (attending, I)

The diet’s the main thing, them following the diet. And it’s hard for me to teach them thinking I would have a hard time with this diet myself. (nurse, I)

I think most of them have 10 or 15 concerns that are ahead of the diabetes, so we’re having to get throughall those things before we hit the behavior change in dealing with diabetes. You know, child-care issues, transportation issues, violence. Everything. You’ve got to find out those things that are ahead of the diabetes. (PA, I)

Another major source of frustration is related to counseling patients about the risks of diabetes. In the face-to-face interviews, several clinicians emphasized the complexity of the disease or the difficulty of making asymptomatic patients aware of the risk of devastating consequences from “uncontrolled” diabetes.

For the most part there’s no symptoms…. They don’t feel better necessarily once they’re being treated. (attending)

It’s frustrating for the patients, because a lot of times they don’t understand…. There’s not necessarily symptoms associated with diabetes. (attending, I)

Diabetes encompasses so many things…their whole life. (nurse, I)

HIV patients see hope when they do something that makes a positive change in their life…whereas diabetics have a harder time getting through to that. (PA, I)

As a result of their frustration with attempts to educate and counsel patients, clinicians often relied on numbers and drugs to attempt to manage diabetes.

It’s hard to talk to them about nutrition, and you try to talk to them about exercise, and again a lot of times that’s hard for our patients who live in urban areas where they don’t want to walk outside, or there’s no place for them to exercise…so quickly you move to drugs. (attending, I)

Thus, an attending with a good “recipe” for teaching the “meds” was greatly appreciated. In an interview, a PA revealed that the head of the diabetes clinic was great at discussing “the meds,” and was a good mentor in clinic, but virtually ignored patient care issues.

I was taught by Dr X, and she is wonderful. She runs a diabetes clinic, and she’s real frank about talking about diabetes…. We talked a lot about treatments and, you know, what are the meds to begin with, but as far as behavior change and how to do any of that kind of stuff-the negotiating, diabetes management-we didn’t talk about that. A real cookbook type of thing…when we got in the clinic I think we saw the subtleties…or not-so-subtleties of treating diabetes. (I)

It appears that the formal teaching is about drugs; the behavioral management is picked up more informally, and therefore, may be perceived by the clinician-in-training as less valued. Thus, instead of education and counseling issues, the case presentations by resident physicians, students, and other clinicians to attending physicians tended to focus on numbers (the patient’s current blood sugar or glycohemoglobin level), compliance with treatment plans for oral hypoglycemic agents or insulin, and whether to increase drug therapy. As this represents the culturally expected behavior of physicians in clinical teaching settings, it took overt encouragement by the attending supervisor to focus their attention on other issues.

Long-Term Outcomes

Another major source of frustration for both clinicians and patients stems from the radical differences in the signs and symptoms that patients and clinicians focus on in treatment. Physicians are primarily concerned with the internal states or processes that lead to serious morbidity and death.

So I just want to go over some of the big complications and see how we can intervene; the biggest concern is diabetic nephropathy, usually very severe retinopathy…nephropathy. (L)

In contrast, the patients we spoke with seemed relatively unconcerned about the effect of diabetes on internal organs and rarely mentioned the heart, pancreas, or kidneys unless they were specifically asked about them. Patients did, however, express grave concern about wounds that would not heal, amputation, and external manifestations of the disease, such as blindness or other symptoms which reveal their condition to the public.

Interviewer: What do you fear most about diabetes?

Patient: Loss of limbs…. Now there is other things, I know kidneys can fail; I know that, and the heart can stop, I know that. But I’m worried about my feet…. Anytime [there’s] a strange color, I’m here. What’s wrong with this? (I)

Pessimism

Finally, physicians in many instances expressed a sense of pessimism or melancholy with of their experience of treating patients or their knowledge of likely medical outcomes.

So nephropathy, we believe, is related to the length of time the person has had diabetes and the control they’ve had over the years. Once a patient has gross proteinuria, end-stage renal failure or kidney transplant is within 5 years. What I’ve heard from the nephrologist is once you have microalbuminuria it’s like a 15-year type of progression from microalbuminuria to end-stage renal disease. The ophthalmologists believe that anyone who’s had diabetes over 10 years, you can detect some degree of retinopathy. (Fourth-year medical student, L)

I think medical students get the experience of diabetes in the hospital frequently, so they see diabetics that are having their second cardiac bypass, or having their legs cut off, or on dialysis. They see the worst end of diabetes instead of seeing the beginning. And so we come into working with diabetics without a lot of hope, and so we sort of go, “this is a horrible disease.” And someone who is diabetic doesn’t want to hear that. (attending, I)

Patients who do not have friends or relatives with severe complications associated with diabetes, however, are often hopeful or at least uncertain about what the future will bring.

Discussion

The Theme of Control

Among the themes revealed through content analysis, control is both the most common and the most complex. Used ubiquitously in both interviews and during observations of clinicians and patients, it has so many connotations that confusion is unavoidable. It is used by patients to refer to things the patient actually does to affect blood sugar as well as other outcomes (like his or her emotional state or the reactions of other people). In the physician narratives, “control” is generally thought of as an objective measurable condition, a range in which a patient’s blood sugar is thought to be relatively healthy. Where personal or subjective factors enter the equation, “compliance” is the operative physician term, although slippage between the 2 terms-compliance and control-is not uncommon. The ambiguity surrounding the term “control” highlights the fact that clinicians are uncertain about the agency or pathophysiology of this disease. Is the patient responsible for poor metabolic control, or do bad outcomes happen for reasons beyond his or her control? Are clinicians and patients in agreement about what might be in the patient’s control? The answers to these questions have major implications for the management of diabetes. If poor control is viewed as a problem of poor compliance (compliance used in the sense of doing something voluntarily and relatively easily), it is easy to blame the patient, especially if he or she is obese, poor, unemployed, or poorly educated. If poor control is not a result of poor compliance, however, or if compliance, as defined by the physician, beyond the practical ability of the patient, then the physician will look for other strategies to increase metabolic control.

Control represents a core value in physician culture^11,24 rather than indisputable scientific wisdom. Tight control of blood sugar levels is a principle on which doctors have acted since long before the recent studies showing its effectiveness for preventing poor long-term outcomes.^25,26

The Themes of Frustration, Long-Term Outcomes, and Pessimism

The frustration experienced by physicians caring for patients with diabetes results from many things, some of which are clearly avoidable. At the beginning of the research process, we hypothesized that this frustration was primarily due to the difference in the frames^27,28 or explanatory models^29,30 that patients and physicians use in thinking about diabetes. This hypothesis was borne out by radical differences in the symptoms and signs the respective parties focus on during treatment, most notably the distinction between internal and external signs.

This hypothesis is also justified by the differences between the perspective of patients who generally do not know where the disease will end up, and the more teleological perspective of the physician who can measure and time the breakdown of organ systems. From the physician’s viewpoint, that breakdown is more likely when control is less than optimal. This perspective leads to a profound pessimism, which not only belies the more optimistic tone of the medical consultation but leads to a series of conundrums. For example, while physicians privately acknowledge their doubts about the benefits of glycemic control or whether it is even possible for their patients will lose weight, they often end up blaming their patients for negative outcomes.

Patients who have symptoms that they can see, focus on those symptoms. They can become confused and frustrated when clinicians focus on less obvious internal conditions. Patients without symptoms feel they are being asked by clinicians to eat unpalatable food or perform unpleasant exercise in the pursuit of a vague hypothetical outcome (eg, avoiding nephropathy). None of the clinicians indicated that they involved their patients in deciding what a desirable outcome would be, but rather stipulated an outcome based on their own beliefs. It is critical that the patients be involved in weighing the risks and benefits of treatment to come up with a practical plan.

It is by no means certain that all patients would choose the same outcomes as desirable. When given the degree of work (and, in their view, discomfort or even suffering) needed to decrease by half an already small probability of ending up on dialysis, it is conceivable that some patients would see such possible outcomes as inadequate justification for denying themselves their favorite foods. Because the treatment plan involves minimal consultation with the patient in pursuit of a goal set by the clinician, it is not surprising that the outcome is often frustration for both.

Conclusions

Clinician attitudes toward diabetes and their patients with diabetes are derived from a number of sources. One source is the culture of medicine, which is largely influenced by a shared set of medical beliefs not always grounded in science or scientific data based on current research. These attitudes and the actions that originate with them are often markedly different from the attitudes of people with diabetes. Even when clinicians voice concern for patient belief issues in the interview setting, they rarely address these beliefs as priority issues in their practice, focusing instead on numbers and drugs. Also, clinician attitudes are usually the sole or primary determinant of what outcome will be pursued. This culturally insular approach leaves patients with little input into their own care plan, and their frequent failure to follow the clinician’s plan leads to major frustration for both parties. Although the macrovascular complications of diabetes (heart disease, stroke) are the major causes of mortality in people with this disease,³¹ in our study they were rarely addressed by providers or patients, which suggests that these well-known and serious conditions were probably not tied to diabetes in the minds of clinicians or patients.

Clinicians should become aware of their own beliefs, understand that they have cultural as well as scientific roots, open themselves to learning the beliefs and attitudes of their patients with diabetes, and recognize that the ultimate decision about the outcome to be pursued is the patient’s, informed by the clinician’s scientific knowledge. In particular, this knowledge includes current clinical guidelines about the relationship of heart disease and blood pressure control with diabetes.^32-35 Clinicians can seek a middle ground between unilaterally imposing external clinical guidelines on an unaccepting patient and adopting a near fatalistic individual relativism that does not seek to help people change behaviors that can adversely affect their health.

Such an approach will allow patients to use information offered by the clinician to identify the outcome they hope to achieve. Then the patient and clinician can work together to design a plan that will realize this desired outcome. It is likely that such an approach will decrease the frustration experienced by both patients and clinicians and enhance the collaborative effort to minimize the impact of diabetes

Acknowledgments

The authors wish to acknowledge the generous financial support of the American Academy of Family Physicians Foundation. Special thanks to Sean O’Sullivan, PhD; Lauretta Quinn, PhD, John Schwartzman, PhD; Steven Zuckerman, PhD; and Eve Pinsker, PhD, for participating in the data analysis sessions, to Anamari Golf, MA, and Dionne Hart for conducting interviews with patients with diabetes, and to Anne Larme, PhD, and Denise Stallcup, MA, for their invaluable feedback on earlier versions of this manuscript. Finally, we would like to thank Stephanie Tillman for coordinating research activities from start to finish and transcribing many of the interviews and precepting sessions.