Mary Ann Moon

Current methods for profiling individual physicians as to whether they provide low-cost or high-cost care are often inaccurate and produce misleading results, according to study findings.

Health plans use cost profiling to limit how many physicians get in-network contracts and to allot bonuses to the physicians whose “resource use” is lower than average. In each case, there must be a method for determining physicians' costs, yet the accuracy of these methods has never been proven, according to John L. Adams, Ph.D., of Rand Corp., Santa Monica, Calif., and his associates.

Dr. Adams and his colleagues assessed the reliability of current methods of cost profiling using claims data from four Massachusetts insurance companies concerning 1.1 million adult patients treated during 2004-2005. A total of 12,789 physicians were included in the study. They were predominantly men who were board certified, had been trained in the United States, and had been in practice for more than 10 years.

The physicians worked in 28 specialties, including cardiology, endocrinology, gastroenterology, and obstetrics and gynecology. Family physicians, general physicians, and internal medicine physicians comprised approximately one-third of the sample.

The investigators estimated the reliability of cost profiles on a scale of 0-1, with 0 representing completely unreliable profiles and 1 representing completely reliable profiles. They then estimated the proportion of physicians in each specialty whose cost performance would be calculated inaccurately.

Overall, only 41% of physicians across all specialties had cost profile scores of 0.70 or greater, a commonly used threshold of acceptable accuracy. Only 47% of internists, 30% of cardiologists, 41% of family or general physicians, 57% of ob.gyns., 59% of gastroenterologists, and 22% of endocrinologists received scores of 0.70.

Overall, only 9% of physicians in the study had scores of 0.90 or greater, indicating optimal accuracy.

The proportion of physicians who were classified as “lower cost” but who were not in fact lower cost ranged from 29% to 67%, depending on the specialty. Fully 39% of family or general physicians were misclassified as “lower-cost” providers when they were not.

These findings indicate that standard methods of cost profiling are highly unreliable, and that many individuals and groups are basing important decisions on inaccuracies (N. Engl. J. Med. 2010;362:1014-21).

The study findings also suggest that using cost profiles that are based on these unreliable methods will not reduce health care spending.

Disclosures: This study was supported by the Department of Labor, the National Institutes of Health, and the Robert Wood Johnson Foundation. The investigators' conflicts of interest included support from the Physicians Advocacy Institute, Commonwealth Fund, and Ingenix Inc.

My Take

Abandon Flawed Evaluations

The RAND Corporation study verifies the American Medical Association's longstanding contention that there are serious flaws in health insurer programs that attempt to rate physicians based on cost-of-care.

The RAND study shows that physician ratings conducted by insurers can be wrong up to two-thirds of the time for some groups of physicians. Inaccurate information can erode patient confidence and trust in caring physicians, and disrupt patients' longstanding relationships with physicians who have cared for them for years.

Patients should always be able to trust that the information they receive on physicians is valid and reliable, especially when the data are used by insurers to influence or restrict patients' choice of physicians.

Given the potential for irreparable damage to the patient-physician relationship, the AMA calls on the health insurance industry to abandon flawed physician evaluation and ranking programs, and join with the AMA to create constructive programs that produce meaningful data for increasing the quality and efficiency of health care.

J. JAMES ROHACK, M.D., president of the American Medical Association, reported having no conflicts of interest.

Current methods for profiling individual physicians as to whether they provide low-cost or high-cost care are often inaccurate and produce misleading results, according to study findings.

Health plans use cost profiling to limit how many physicians get in-network contracts and to allot bonuses to the physicians whose “resource use” is lower than average. In each case, there must be a method for determining physicians' costs, yet the accuracy of these methods has never been proven, according to John L. Adams, Ph.D., of Rand Corp., Santa Monica, Calif., and his associates.

Dr. Adams and his colleagues assessed the reliability of current methods of cost profiling using claims data from four Massachusetts insurance companies concerning 1.1 million adult patients treated during 2004-2005. A total of 12,789 physicians were included in the study. They were predominantly men who were board certified, had been trained in the United States, and had been in practice for more than 10 years.

The physicians worked in 28 specialties, including cardiology, endocrinology, gastroenterology, and obstetrics and gynecology. Family physicians, general physicians, and internal medicine physicians comprised approximately one-third of the sample.

The investigators estimated the reliability of cost profiles on a scale of 0-1, with 0 representing completely unreliable profiles and 1 representing completely reliable profiles. They then estimated the proportion of physicians in each specialty whose cost performance would be calculated inaccurately.

Overall, only 41% of physicians across all specialties had cost profile scores of 0.70 or greater, a commonly used threshold of acceptable accuracy. Only 47% of internists, 30% of cardiologists, 41% of family or general physicians, 57% of ob.gyns., 59% of gastroenterologists, and 22% of endocrinologists received scores of 0.70.

Overall, only 9% of physicians in the study had scores of 0.90 or greater, indicating optimal accuracy.

The proportion of physicians who were classified as “lower cost” but who were not in fact lower cost ranged from 29% to 67%, depending on the specialty. Fully 39% of family or general physicians were misclassified as “lower-cost” providers when they were not.

These findings indicate that standard methods of cost profiling are highly unreliable, and that many individuals and groups are basing important decisions on inaccuracies (N. Engl. J. Med. 2010;362:1014-21).

The study findings also suggest that using cost profiles that are based on these unreliable methods will not reduce health care spending.

Disclosures: This study was supported by the Department of Labor, the National Institutes of Health, and the Robert Wood Johnson Foundation. The investigators' conflicts of interest included support from the Physicians Advocacy Institute, Commonwealth Fund, and Ingenix Inc.

My Take

Abandon Flawed Evaluations

The RAND Corporation study verifies the American Medical Association's longstanding contention that there are serious flaws in health insurer programs that attempt to rate physicians based on cost-of-care.

The RAND study shows that physician ratings conducted by insurers can be wrong up to two-thirds of the time for some groups of physicians. Inaccurate information can erode patient confidence and trust in caring physicians, and disrupt patients' longstanding relationships with physicians who have cared for them for years.

Patients should always be able to trust that the information they receive on physicians is valid and reliable, especially when the data are used by insurers to influence or restrict patients' choice of physicians.

Given the potential for irreparable damage to the patient-physician relationship, the AMA calls on the health insurance industry to abandon flawed physician evaluation and ranking programs, and join with the AMA to create constructive programs that produce meaningful data for increasing the quality and efficiency of health care.

J. JAMES ROHACK, M.D., president of the American Medical Association, reported having no conflicts of interest.

Current methods for profiling individual physicians as to whether they provide low-cost or high-cost care are often inaccurate and produce misleading results, according to study findings.

Health plans use cost profiling to limit how many physicians get in-network contracts and to allot bonuses to the physicians whose “resource use” is lower than average. In each case, there must be a method for determining physicians' costs, yet the accuracy of these methods has never been proven, according to John L. Adams, Ph.D., of Rand Corp., Santa Monica, Calif., and his associates.

Dr. Adams and his colleagues assessed the reliability of current methods of cost profiling using claims data from four Massachusetts insurance companies concerning 1.1 million adult patients treated during 2004-2005. A total of 12,789 physicians were included in the study. They were predominantly men who were board certified, had been trained in the United States, and had been in practice for more than 10 years.

The physicians worked in 28 specialties, including cardiology, endocrinology, gastroenterology, and obstetrics and gynecology. Family physicians, general physicians, and internal medicine physicians comprised approximately one-third of the sample.

The investigators estimated the reliability of cost profiles on a scale of 0-1, with 0 representing completely unreliable profiles and 1 representing completely reliable profiles. They then estimated the proportion of physicians in each specialty whose cost performance would be calculated inaccurately.

Overall, only 41% of physicians across all specialties had cost profile scores of 0.70 or greater, a commonly used threshold of acceptable accuracy. Only 47% of internists, 30% of cardiologists, 41% of family or general physicians, 57% of ob.gyns., 59% of gastroenterologists, and 22% of endocrinologists received scores of 0.70.

Overall, only 9% of physicians in the study had scores of 0.90 or greater, indicating optimal accuracy.

The proportion of physicians who were classified as “lower cost” but who were not in fact lower cost ranged from 29% to 67%, depending on the specialty. Fully 39% of family or general physicians were misclassified as “lower-cost” providers when they were not.

These findings indicate that standard methods of cost profiling are highly unreliable, and that many individuals and groups are basing important decisions on inaccuracies (N. Engl. J. Med. 2010;362:1014-21).

The study findings also suggest that using cost profiles that are based on these unreliable methods will not reduce health care spending.

Disclosures: This study was supported by the Department of Labor, the National Institutes of Health, and the Robert Wood Johnson Foundation. The investigators' conflicts of interest included support from the Physicians Advocacy Institute, Commonwealth Fund, and Ingenix Inc.

My Take

Abandon Flawed Evaluations

The RAND Corporation study verifies the American Medical Association's longstanding contention that there are serious flaws in health insurer programs that attempt to rate physicians based on cost-of-care.

The RAND study shows that physician ratings conducted by insurers can be wrong up to two-thirds of the time for some groups of physicians. Inaccurate information can erode patient confidence and trust in caring physicians, and disrupt patients' longstanding relationships with physicians who have cared for them for years.

Patients should always be able to trust that the information they receive on physicians is valid and reliable, especially when the data are used by insurers to influence or restrict patients' choice of physicians.

Given the potential for irreparable damage to the patient-physician relationship, the AMA calls on the health insurance industry to abandon flawed physician evaluation and ranking programs, and join with the AMA to create constructive programs that produce meaningful data for increasing the quality and efficiency of health care.

J. JAMES ROHACK, M.D., president of the American Medical Association, reported having no conflicts of interest.

Twelve to 15 million white patients living in the United States have had at least one nonmelanoma skin cancer in their lifetimes, according to estimates based on a new mathematical model.

This figure is approximately twice that of previous estimates based on patient surveys, such as the estimate calculated in the National Health Interview Study, according to Dr. Robert S. Stern of Beth Israel Deaconess Medical Center and Harvard Medical School, Boston.

This difference can be attributed in part to people often incorrectly reporting their skin cancer histories when they are surveyed, falsely believing that basal cell and squamous cell lesions are not cancerous or that all skin cancers can be considered melanoma.

Dr. Stern devised his mathematical model using the same basic data available from national samples, such as the Surveillance, Epidemiology, and End Results (SEER) studies and information from the National Cancer Institute.

His model, however, took into consideration several factors that had not been accounted for in previous estimates, such as the likelihood that patients develop numerous nonmelanoma skin cancers over the course of several years and that “a substantial proportion” of patients with melanoma also have nonmalignant skin cancers.

According to his model, “about 13 million white non-Hispanic U.S. residents (6%) have had more than 22 million nonmelanoma skin cancers” (Arch. Dermatol. 2010;146:279-82).

These estimates put the prevalence of a skin cancer history at a level far higher than that of any other cancer—prevalence that “exceeds that of all other cancers diagnosed since 1975,” he added. “Recent population-based data concerning skin cancer incidence, morbidity, and cost of care are lacking for the most common types of skin cancer, [basal cell carcinoma] and [squamous cell carcinoma].”

Regarding patients' inaccurate reporting of skin cancer histories in surveys and interviews, Dr. Stern noted that many patients equate skin cancer with melanoma. “Hundreds of thousands of patients may be unnecessarily burdened with the belief that they have had a potentially lethal cancer (melanoma) when in fact they have had a skin tumor that is very unlikely to be lethal,” he wrote.

In contrast, patients with breast or prostate cancer are much more likely to accurately report their cancer histories, so incidence and prevalence estimates for these tumors are much more accurate than those for skin cancer, Dr. Stern noted.

This study was funded in part by the National Institutes of Health. Dr. Stern reported being a consultant for Johnson & Johnson, Vertex, and Takeda and an expert witness for Alphapharm, Mutual Pharm, and Johnson & Johnson.

Twelve to 15 million white patients living in the United States have had at least one nonmelanoma skin cancer in their lifetimes, according to estimates based on a new mathematical model.

This figure is approximately twice that of previous estimates based on patient surveys, such as the estimate calculated in the National Health Interview Study, according to Dr. Robert S. Stern of Beth Israel Deaconess Medical Center and Harvard Medical School, Boston.

This difference can be attributed in part to people often incorrectly reporting their skin cancer histories when they are surveyed, falsely believing that basal cell and squamous cell lesions are not cancerous or that all skin cancers can be considered melanoma.

Dr. Stern devised his mathematical model using the same basic data available from national samples, such as the Surveillance, Epidemiology, and End Results (SEER) studies and information from the National Cancer Institute.

His model, however, took into consideration several factors that had not been accounted for in previous estimates, such as the likelihood that patients develop numerous nonmelanoma skin cancers over the course of several years and that “a substantial proportion” of patients with melanoma also have nonmalignant skin cancers.

According to his model, “about 13 million white non-Hispanic U.S. residents (6%) have had more than 22 million nonmelanoma skin cancers” (Arch. Dermatol. 2010;146:279-82).

These estimates put the prevalence of a skin cancer history at a level far higher than that of any other cancer—prevalence that “exceeds that of all other cancers diagnosed since 1975,” he added. “Recent population-based data concerning skin cancer incidence, morbidity, and cost of care are lacking for the most common types of skin cancer, [basal cell carcinoma] and [squamous cell carcinoma].”

Regarding patients' inaccurate reporting of skin cancer histories in surveys and interviews, Dr. Stern noted that many patients equate skin cancer with melanoma. “Hundreds of thousands of patients may be unnecessarily burdened with the belief that they have had a potentially lethal cancer (melanoma) when in fact they have had a skin tumor that is very unlikely to be lethal,” he wrote.

In contrast, patients with breast or prostate cancer are much more likely to accurately report their cancer histories, so incidence and prevalence estimates for these tumors are much more accurate than those for skin cancer, Dr. Stern noted.

This study was funded in part by the National Institutes of Health. Dr. Stern reported being a consultant for Johnson & Johnson, Vertex, and Takeda and an expert witness for Alphapharm, Mutual Pharm, and Johnson & Johnson.

Twelve to 15 million white patients living in the United States have had at least one nonmelanoma skin cancer in their lifetimes, according to estimates based on a new mathematical model.

This figure is approximately twice that of previous estimates based on patient surveys, such as the estimate calculated in the National Health Interview Study, according to Dr. Robert S. Stern of Beth Israel Deaconess Medical Center and Harvard Medical School, Boston.

This difference can be attributed in part to people often incorrectly reporting their skin cancer histories when they are surveyed, falsely believing that basal cell and squamous cell lesions are not cancerous or that all skin cancers can be considered melanoma.

Dr. Stern devised his mathematical model using the same basic data available from national samples, such as the Surveillance, Epidemiology, and End Results (SEER) studies and information from the National Cancer Institute.

His model, however, took into consideration several factors that had not been accounted for in previous estimates, such as the likelihood that patients develop numerous nonmelanoma skin cancers over the course of several years and that “a substantial proportion” of patients with melanoma also have nonmalignant skin cancers.

According to his model, “about 13 million white non-Hispanic U.S. residents (6%) have had more than 22 million nonmelanoma skin cancers” (Arch. Dermatol. 2010;146:279-82).

These estimates put the prevalence of a skin cancer history at a level far higher than that of any other cancer—prevalence that “exceeds that of all other cancers diagnosed since 1975,” he added. “Recent population-based data concerning skin cancer incidence, morbidity, and cost of care are lacking for the most common types of skin cancer, [basal cell carcinoma] and [squamous cell carcinoma].”

Regarding patients' inaccurate reporting of skin cancer histories in surveys and interviews, Dr. Stern noted that many patients equate skin cancer with melanoma. “Hundreds of thousands of patients may be unnecessarily burdened with the belief that they have had a potentially lethal cancer (melanoma) when in fact they have had a skin tumor that is very unlikely to be lethal,” he wrote.

In contrast, patients with breast or prostate cancer are much more likely to accurately report their cancer histories, so incidence and prevalence estimates for these tumors are much more accurate than those for skin cancer, Dr. Stern noted.

This study was funded in part by the National Institutes of Health. Dr. Stern reported being a consultant for Johnson & Johnson, Vertex, and Takeda and an expert witness for Alphapharm, Mutual Pharm, and Johnson & Johnson.

Adding eprotirome to statin therapy further reduced serum LDL cholesterol, non-HDL cholesterol, and apolipoprotein B in a phase II study.

Eprotirome, an investigational thyroid hormone analogue, also decreased levels of two other atherogenic lipids—triglycerides and Lp(a) lipoprotein—which are known to have a comparatively poor response to statins alone, said Dr. Paul W. Ladenson of Johns Hopkins University, Baltimore, and his associates.

The investigators conducted a double-blind trial in 189 patients already taking simvastatin or atorvastatin to determine whether adding eprotirome would decrease levels of atherogenic lipoproteins even further. These subjects continued to have a mean LDL cholesterol level of 116 mg/dL despite statin therapy.

The study subjects were randomly assigned to receive placebo or one of three doses of eprotirome—25, 50, or 100 mcg—in the form of enteric-coated tablets for 12 weeks. The study was sponsored by Karo Bio, maker of eprotirome.

A total of 168 subjects completed the trial.

Serum LDL cholesterol decreased 22% from baseline levels at the lowest dose of eprotirome, 28% at the intermediate dose, and 32% at the high dose, compared with a 7% decrease with placebo.

The proportion of subjects who had an LDL level of less than 100 mg/dL was 36% at the lowest dose of eprotirome, 50% at the intermediate dose, and 57% at the high dose, compared with 6% with placebo.

“Eprotirome was associated with larger decreases in levels of serum LDL cholesterol than would be expected with a doubling of the statin dose,” Dr. Ladenson and his colleagues said (N. Engl. J. Med. 2010;362:906-16).

The drug lowered serum levels of the atherogenic compounds apolipoprotein B, triglycerides, and Lp(a) lipoprotein. However, it also decreased serum levels of the favorable compounds HDL cholesterol and apolipoprotein A-I.

Dr. Ladenson reported that he received consulting fees from Karo Bio and Genzyme.

Adding eprotirome to statin therapy further reduced serum LDL cholesterol, non-HDL cholesterol, and apolipoprotein B in a phase II study.

Eprotirome, an investigational thyroid hormone analogue, also decreased levels of two other atherogenic lipids—triglycerides and Lp(a) lipoprotein—which are known to have a comparatively poor response to statins alone, said Dr. Paul W. Ladenson of Johns Hopkins University, Baltimore, and his associates.

The investigators conducted a double-blind trial in 189 patients already taking simvastatin or atorvastatin to determine whether adding eprotirome would decrease levels of atherogenic lipoproteins even further. These subjects continued to have a mean LDL cholesterol level of 116 mg/dL despite statin therapy.

The study subjects were randomly assigned to receive placebo or one of three doses of eprotirome—25, 50, or 100 mcg—in the form of enteric-coated tablets for 12 weeks. The study was sponsored by Karo Bio, maker of eprotirome.

A total of 168 subjects completed the trial.

Serum LDL cholesterol decreased 22% from baseline levels at the lowest dose of eprotirome, 28% at the intermediate dose, and 32% at the high dose, compared with a 7% decrease with placebo.

The proportion of subjects who had an LDL level of less than 100 mg/dL was 36% at the lowest dose of eprotirome, 50% at the intermediate dose, and 57% at the high dose, compared with 6% with placebo.

“Eprotirome was associated with larger decreases in levels of serum LDL cholesterol than would be expected with a doubling of the statin dose,” Dr. Ladenson and his colleagues said (N. Engl. J. Med. 2010;362:906-16).

The drug lowered serum levels of the atherogenic compounds apolipoprotein B, triglycerides, and Lp(a) lipoprotein. However, it also decreased serum levels of the favorable compounds HDL cholesterol and apolipoprotein A-I.

Dr. Ladenson reported that he received consulting fees from Karo Bio and Genzyme.

Adding eprotirome to statin therapy further reduced serum LDL cholesterol, non-HDL cholesterol, and apolipoprotein B in a phase II study.

Eprotirome, an investigational thyroid hormone analogue, also decreased levels of two other atherogenic lipids—triglycerides and Lp(a) lipoprotein—which are known to have a comparatively poor response to statins alone, said Dr. Paul W. Ladenson of Johns Hopkins University, Baltimore, and his associates.

The investigators conducted a double-blind trial in 189 patients already taking simvastatin or atorvastatin to determine whether adding eprotirome would decrease levels of atherogenic lipoproteins even further. These subjects continued to have a mean LDL cholesterol level of 116 mg/dL despite statin therapy.

The study subjects were randomly assigned to receive placebo or one of three doses of eprotirome—25, 50, or 100 mcg—in the form of enteric-coated tablets for 12 weeks. The study was sponsored by Karo Bio, maker of eprotirome.

A total of 168 subjects completed the trial.

Serum LDL cholesterol decreased 22% from baseline levels at the lowest dose of eprotirome, 28% at the intermediate dose, and 32% at the high dose, compared with a 7% decrease with placebo.

The proportion of subjects who had an LDL level of less than 100 mg/dL was 36% at the lowest dose of eprotirome, 50% at the intermediate dose, and 57% at the high dose, compared with 6% with placebo.

“Eprotirome was associated with larger decreases in levels of serum LDL cholesterol than would be expected with a doubling of the statin dose,” Dr. Ladenson and his colleagues said (N. Engl. J. Med. 2010;362:906-16).

The drug lowered serum levels of the atherogenic compounds apolipoprotein B, triglycerides, and Lp(a) lipoprotein. However, it also decreased serum levels of the favorable compounds HDL cholesterol and apolipoprotein A-I.

Dr. Ladenson reported that he received consulting fees from Karo Bio and Genzyme.

Immunizing children aged 3-15 years in isolated rural communities against influenza conferred substantial immunity to unvaccinated members of the communities, said Dr. Mark Loeb of McMaster University, Hamilton, Ont., and his associates.

“Our findings offer experimental proof to support selective influenza immunization of school-aged children with inactivated influenza vaccine to interrupt influenza transmission,” they wrote (JAMA 2010;303:943-50).

Observational and computer modeling studies have suggested that such an approach might reduce influenza transmission, but randomized clinical trials to confirm this theory have not been feasible because in most settings, it would be unethical to withhold immunization from children in a control group.

However, rural Hutterite colonies in Western Canada offer a unique setting for such a study. These communities of approximately 60-120 Anabaptist residents are relatively isolated from other populations but show significant influenza activity each winter. The members of 46 Hutterite colonies in Alberta, Saskatchewan, and Manitoba agreed to random assignment to receive either immunization for influenza A and B during the 2008-2009 flu season (22 colonies) or to receive hepatitis A vaccination as a control (24 colonies).

Only healthy children aged 3-15 years were immunized. Mean vaccine coverage was 83% in this age group. This resulted in 502 children receiving flu vaccine in a population totaling 1,773 and 445 children receiving hepatitis A vaccine in a population totaling 1,500.

The primary outcomeof s the development of laboratory-confirmed influenza A or B in colony members who did not receive flu vaccine. This occurred in 39 members of colonies assigned to influenza immunization (3%), a rate less than half of the 7.6% rate of influenza infection in control colonies.

“The level of indirect vaccine protectiveness was 61%” overall and 49% among high-risk subjects, Dr. Loeb and his colleagues said.

There were six outbreaks of influenza in the vaccinated colonies, with 3-16 cases in each outbreak. In contrast, there were more than twice as many outbreaks (13) in the control colonies, with 4-26 cases in each outbreak.

Disclosures: This study was supported by the Canadian Institutes for Health Research and the National Institute for Allergy and Infectious Diseases. Sanofi Pasteur donated vaccines used for the study but provided no funding and had no other role. The authors said that they had no conflicts of interest.

Immunizing children aged 3-15 years in isolated rural communities against influenza conferred substantial immunity to unvaccinated members of the communities, said Dr. Mark Loeb of McMaster University, Hamilton, Ont., and his associates.

“Our findings offer experimental proof to support selective influenza immunization of school-aged children with inactivated influenza vaccine to interrupt influenza transmission,” they wrote (JAMA 2010;303:943-50).

Observational and computer modeling studies have suggested that such an approach might reduce influenza transmission, but randomized clinical trials to confirm this theory have not been feasible because in most settings, it would be unethical to withhold immunization from children in a control group.

However, rural Hutterite colonies in Western Canada offer a unique setting for such a study. These communities of approximately 60-120 Anabaptist residents are relatively isolated from other populations but show significant influenza activity each winter. The members of 46 Hutterite colonies in Alberta, Saskatchewan, and Manitoba agreed to random assignment to receive either immunization for influenza A and B during the 2008-2009 flu season (22 colonies) or to receive hepatitis A vaccination as a control (24 colonies).

Only healthy children aged 3-15 years were immunized. Mean vaccine coverage was 83% in this age group. This resulted in 502 children receiving flu vaccine in a population totaling 1,773 and 445 children receiving hepatitis A vaccine in a population totaling 1,500.

The primary outcomeof s the development of laboratory-confirmed influenza A or B in colony members who did not receive flu vaccine. This occurred in 39 members of colonies assigned to influenza immunization (3%), a rate less than half of the 7.6% rate of influenza infection in control colonies.

“The level of indirect vaccine protectiveness was 61%” overall and 49% among high-risk subjects, Dr. Loeb and his colleagues said.

There were six outbreaks of influenza in the vaccinated colonies, with 3-16 cases in each outbreak. In contrast, there were more than twice as many outbreaks (13) in the control colonies, with 4-26 cases in each outbreak.

Disclosures: This study was supported by the Canadian Institutes for Health Research and the National Institute for Allergy and Infectious Diseases. Sanofi Pasteur donated vaccines used for the study but provided no funding and had no other role. The authors said that they had no conflicts of interest.

Immunizing children aged 3-15 years in isolated rural communities against influenza conferred substantial immunity to unvaccinated members of the communities, said Dr. Mark Loeb of McMaster University, Hamilton, Ont., and his associates.

“Our findings offer experimental proof to support selective influenza immunization of school-aged children with inactivated influenza vaccine to interrupt influenza transmission,” they wrote (JAMA 2010;303:943-50).

Observational and computer modeling studies have suggested that such an approach might reduce influenza transmission, but randomized clinical trials to confirm this theory have not been feasible because in most settings, it would be unethical to withhold immunization from children in a control group.

However, rural Hutterite colonies in Western Canada offer a unique setting for such a study. These communities of approximately 60-120 Anabaptist residents are relatively isolated from other populations but show significant influenza activity each winter. The members of 46 Hutterite colonies in Alberta, Saskatchewan, and Manitoba agreed to random assignment to receive either immunization for influenza A and B during the 2008-2009 flu season (22 colonies) or to receive hepatitis A vaccination as a control (24 colonies).

Only healthy children aged 3-15 years were immunized. Mean vaccine coverage was 83% in this age group. This resulted in 502 children receiving flu vaccine in a population totaling 1,773 and 445 children receiving hepatitis A vaccine in a population totaling 1,500.

The primary outcomeof s the development of laboratory-confirmed influenza A or B in colony members who did not receive flu vaccine. This occurred in 39 members of colonies assigned to influenza immunization (3%), a rate less than half of the 7.6% rate of influenza infection in control colonies.

“The level of indirect vaccine protectiveness was 61%” overall and 49% among high-risk subjects, Dr. Loeb and his colleagues said.

There were six outbreaks of influenza in the vaccinated colonies, with 3-16 cases in each outbreak. In contrast, there were more than twice as many outbreaks (13) in the control colonies, with 4-26 cases in each outbreak.

Disclosures: This study was supported by the Canadian Institutes for Health Research and the National Institute for Allergy and Infectious Diseases. Sanofi Pasteur donated vaccines used for the study but provided no funding and had no other role. The authors said that they had no conflicts of interest.

Nateglinide, an insulin secretagogue that lowers postprandial glucose, failed to prevent the development of diabetes and related cardiovascular events among high-risk patients in a large international clinical trial.

The angiotensin-receptor blocker valsartan also failed to prevent cardiovascular events in the same trial, but it did induce an unexpected relative reduction of 14% in the incidence of diabetes, according to the Nateglinide and Valsartan in Impaired Glucose Tolerance Outcomes Research (NAVIGATOR) Study Group.

The results were published online in the New England Journal of Medicine, simultaneously with the planned presentation at the annual meeting of the American College of Cardiology in Atlanta.

In an editorial comment accompanying the two reports, Dr. David M. Nathan of Massachusetts General Hospital, Boston, said, “The authors suggest that the prevention of diabetes with valsartan might make it a preferred drug as compared with antihypertensive drugs that potentially worsen glycemia.”

Instead, the study findings show that “for now we should steer away from these two drugs” when attempting to forestall diabetes and cardiovascular complications in high-risk patients, Dr. Nathan said (N. Engl. J. Med. 2010 March 14 [doi:10.1056/NEJMe1002322

In NAVIGATOR, 9,306 patients who had impaired glucose tolerance and either known cardiovascular disease or cardiovascular risk factors were randomly assigned to take 60 mg oral nateglinide before meals three times daily, a placebo, or in a 2-by-2 factorial design, oral valsartan or a placebo.

Nateglinide was studied to determine whether it would slow progression to diabetes by restoring a more physiologic insulin response to meals. However, during a mean follow-up of about 6 years, progression to diabetes occurred in 36% of the nateglinide group and 34% of the placebo group, a nonsignificant difference, said Dr. Rury R. Holman of Oxford (England) University's Centre for Diabetes, Endocrinology, and Metabolism, and his associates said (N. Engl. J. Med. 2010 March 14 [doi:10.1056/NEJMoa1001122

Similarly, a composite cardiovascular outcome event occurred in 14% of the nateglinide group and 15% of the placebo group, a nonsignificant difference. There also were no differences between the two groups in any of the individual components of the composite cardiovascular outcome, including mortality rates.

The valsartan results were reported in a separate article. Unexpectedly, the angiotensin-receptor blocker had no effect on combined cardiovascular outcomes. Also unexpectedly, it reduced the incidence of diabetes by 14% relative to placebo, said Dr. Robert M. Califf of the Duke Translational Medicine Institute in Durham, N.C., and his NAVIGATOR colleagues.

It is possible that valsartan did not improve cardiovascular outcomes as it should have because most risk factors were already well controlled, since study subjects were allowed to take nonstudy medications such as ACE inhibitors, they said (N. Engl. J. Med. 2010 March 14 [doi:10.1056/NEJMoa1001121

Also, a “substantial proportion” of study subjects discontinued valsartan during the trial, which may have further mitigated its beneficial effects, they said.

In his editorial comment, Dr. Nathan agreed that “the high rates of loss to follow-up (13%), use of off-study ACE inhibitors or ARBs among participants assigned to placebo (24%), and nonadherence to valsartan (34% by study end) could explain the absence of an effect on cardiovascular disease.”

He went on to question the use of valsartan in the study in the first place. “The rationale behind the choice of valsartan to inhibit the renin-angiotensin axis is less clear, other than the fact that both nateglinide and valsartan are manufactured by the pharmaceutical sponsor, which also designed the study,” he said.

Nateglinide, an insulin secretagogue that lowers postprandial glucose, failed to prevent the development of diabetes and related cardiovascular events among high-risk patients in a large international clinical trial.

The angiotensin-receptor blocker valsartan also failed to prevent cardiovascular events in the same trial, but it did induce an unexpected relative reduction of 14% in the incidence of diabetes, according to the Nateglinide and Valsartan in Impaired Glucose Tolerance Outcomes Research (NAVIGATOR) Study Group.

The results were published online in the New England Journal of Medicine, simultaneously with the planned presentation at the annual meeting of the American College of Cardiology in Atlanta.

In an editorial comment accompanying the two reports, Dr. David M. Nathan of Massachusetts General Hospital, Boston, said, “The authors suggest that the prevention of diabetes with valsartan might make it a preferred drug as compared with antihypertensive drugs that potentially worsen glycemia.”

Instead, the study findings show that “for now we should steer away from these two drugs” when attempting to forestall diabetes and cardiovascular complications in high-risk patients, Dr. Nathan said (N. Engl. J. Med. 2010 March 14 [doi:10.1056/NEJMe1002322

In NAVIGATOR, 9,306 patients who had impaired glucose tolerance and either known cardiovascular disease or cardiovascular risk factors were randomly assigned to take 60 mg oral nateglinide before meals three times daily, a placebo, or in a 2-by-2 factorial design, oral valsartan or a placebo.

Nateglinide was studied to determine whether it would slow progression to diabetes by restoring a more physiologic insulin response to meals. However, during a mean follow-up of about 6 years, progression to diabetes occurred in 36% of the nateglinide group and 34% of the placebo group, a nonsignificant difference, said Dr. Rury R. Holman of Oxford (England) University's Centre for Diabetes, Endocrinology, and Metabolism, and his associates said (N. Engl. J. Med. 2010 March 14 [doi:10.1056/NEJMoa1001122

Similarly, a composite cardiovascular outcome event occurred in 14% of the nateglinide group and 15% of the placebo group, a nonsignificant difference. There also were no differences between the two groups in any of the individual components of the composite cardiovascular outcome, including mortality rates.

The valsartan results were reported in a separate article. Unexpectedly, the angiotensin-receptor blocker had no effect on combined cardiovascular outcomes. Also unexpectedly, it reduced the incidence of diabetes by 14% relative to placebo, said Dr. Robert M. Califf of the Duke Translational Medicine Institute in Durham, N.C., and his NAVIGATOR colleagues.

It is possible that valsartan did not improve cardiovascular outcomes as it should have because most risk factors were already well controlled, since study subjects were allowed to take nonstudy medications such as ACE inhibitors, they said (N. Engl. J. Med. 2010 March 14 [doi:10.1056/NEJMoa1001121

Also, a “substantial proportion” of study subjects discontinued valsartan during the trial, which may have further mitigated its beneficial effects, they said.

In his editorial comment, Dr. Nathan agreed that “the high rates of loss to follow-up (13%), use of off-study ACE inhibitors or ARBs among participants assigned to placebo (24%), and nonadherence to valsartan (34% by study end) could explain the absence of an effect on cardiovascular disease.”

He went on to question the use of valsartan in the study in the first place. “The rationale behind the choice of valsartan to inhibit the renin-angiotensin axis is less clear, other than the fact that both nateglinide and valsartan are manufactured by the pharmaceutical sponsor, which also designed the study,” he said.

Nateglinide, an insulin secretagogue that lowers postprandial glucose, failed to prevent the development of diabetes and related cardiovascular events among high-risk patients in a large international clinical trial.

The angiotensin-receptor blocker valsartan also failed to prevent cardiovascular events in the same trial, but it did induce an unexpected relative reduction of 14% in the incidence of diabetes, according to the Nateglinide and Valsartan in Impaired Glucose Tolerance Outcomes Research (NAVIGATOR) Study Group.

The results were published online in the New England Journal of Medicine, simultaneously with the planned presentation at the annual meeting of the American College of Cardiology in Atlanta.

In an editorial comment accompanying the two reports, Dr. David M. Nathan of Massachusetts General Hospital, Boston, said, “The authors suggest that the prevention of diabetes with valsartan might make it a preferred drug as compared with antihypertensive drugs that potentially worsen glycemia.”

Instead, the study findings show that “for now we should steer away from these two drugs” when attempting to forestall diabetes and cardiovascular complications in high-risk patients, Dr. Nathan said (N. Engl. J. Med. 2010 March 14 [doi:10.1056/NEJMe1002322

In NAVIGATOR, 9,306 patients who had impaired glucose tolerance and either known cardiovascular disease or cardiovascular risk factors were randomly assigned to take 60 mg oral nateglinide before meals three times daily, a placebo, or in a 2-by-2 factorial design, oral valsartan or a placebo.

Nateglinide was studied to determine whether it would slow progression to diabetes by restoring a more physiologic insulin response to meals. However, during a mean follow-up of about 6 years, progression to diabetes occurred in 36% of the nateglinide group and 34% of the placebo group, a nonsignificant difference, said Dr. Rury R. Holman of Oxford (England) University's Centre for Diabetes, Endocrinology, and Metabolism, and his associates said (N. Engl. J. Med. 2010 March 14 [doi:10.1056/NEJMoa1001122

Similarly, a composite cardiovascular outcome event occurred in 14% of the nateglinide group and 15% of the placebo group, a nonsignificant difference. There also were no differences between the two groups in any of the individual components of the composite cardiovascular outcome, including mortality rates.

The valsartan results were reported in a separate article. Unexpectedly, the angiotensin-receptor blocker had no effect on combined cardiovascular outcomes. Also unexpectedly, it reduced the incidence of diabetes by 14% relative to placebo, said Dr. Robert M. Califf of the Duke Translational Medicine Institute in Durham, N.C., and his NAVIGATOR colleagues.

It is possible that valsartan did not improve cardiovascular outcomes as it should have because most risk factors were already well controlled, since study subjects were allowed to take nonstudy medications such as ACE inhibitors, they said (N. Engl. J. Med. 2010 March 14 [doi:10.1056/NEJMoa1001121

Also, a “substantial proportion” of study subjects discontinued valsartan during the trial, which may have further mitigated its beneficial effects, they said.

In his editorial comment, Dr. Nathan agreed that “the high rates of loss to follow-up (13%), use of off-study ACE inhibitors or ARBs among participants assigned to placebo (24%), and nonadherence to valsartan (34% by study end) could explain the absence of an effect on cardiovascular disease.”

He went on to question the use of valsartan in the study in the first place. “The rationale behind the choice of valsartan to inhibit the renin-angiotensin axis is less clear, other than the fact that both nateglinide and valsartan are manufactured by the pharmaceutical sponsor, which also designed the study,” he said.

Major Finding: Those who started using marijuana at 15 were twice as likely to receive a diagnosis of nonaffective psychosis in young adulthood as their counterparts who said they never used.

Data Source: Prospective, sibling pair analysis of 3,081 adults born between 1981 and 1984.

Disclosures: None of the investigators had any financial conflicts of interest to report.

The use of cannabis at a younger age is associated with psychosis symptoms in early adulthood, according to a study published online in the Archives of General Psychiatry.

Such a link has been reported before, but this is the first study to demonstrate the association in a subgroup of sibling pairs, “thus reducing the likelihood that the association was due to unmeasured shared genetic and/or environmental influences,” said Dr. John McGrath of the Queensland Brain Institute, Wacol, Australia, and his colleagues.

In addition, their study demonstrated a dose-response relationship between younger age at first marijuana use and higher risk of psychosis-related outcomes, they noted.

Earlier studies had found a link between early-onset cannabis use and later symptoms of psychosis, but there were “lingering concerns that the association may reflect methodological biases and unmeasured residual confounding” in those studies. Dr. McGrath and his associates examined the link using data from a birth cohort of more than 7,000 mother-infant pairs who were first studied in 1981–1984 and followed up 5, 14, and 21 years later.

A total of 3,801 of these infants and their close-in-age siblings comprised the subjects in this study. The latest follow-up occurred when they were aged 18–23 years. At that time, about 18% of the subjects said they had been using marijuana for 3 or fewer years; 16% said they had been using it for 4–5 years; and 14% said they had been using it for 6 or more years.

At this final follow-up, 65 of these subjects had received diagnoses of nonaffective psychosis because they met the criteria for schizophrenia (53 subjects), persistent delusional disorder (3), or acute transient psychotic disorders (9). An additional 233 subjects reported at least one visual or auditory hallucination on the Composite International Diagnostic Interview (CIDI).

Only subjects with the longest duration since first cannabis use–that is, those who started using marijuana at age 15 years or younger–were at significantly increased risk for developing symptoms of nonaffective psychosis in young adulthood.

Those who started using marijuana at that age were twice as likely to receive such a diagnosis than were subjects who said they had never used marijuana, the researchers said (Arch. Gen. Psychiatry 2010 [doi:10.1001/archgenpsychiatry.2010.6]).

Compared with subjects who did not use cannabis, those who used it at a younger age were 4 times more likely to score in the top quartile on the 21-item Peters et al. Delusional Inventory (PDI) and to report hallucinations on the CIDI.

Moreover, the longer the interval since first cannabis use, the higher the risk of these adverse psychosis-related outcomes.

In a subsample of 218 sibling pairs, there was a significant association between earlier first use of cannabis and higher scores on the PDI. For every additional year since first exposure to marijuana, the sibling with the younger age at first use scored one item higher than the other sibling, wrote the authors.

This study could delineate an association between early marijuana use and later symptoms of psychosis, and was not designed to determine causality. “We cannot confidently exclude the possibility that some of the cohort members may have developed psychosis as young adolescents, which may have contributed to subsequent [early] cannabis use,” they added.

The National Health and Medical Research Council of Australia funded the study.

Major Finding: Those who started using marijuana at 15 were twice as likely to receive a diagnosis of nonaffective psychosis in young adulthood as their counterparts who said they never used.

Data Source: Prospective, sibling pair analysis of 3,081 adults born between 1981 and 1984.

Disclosures: None of the investigators had any financial conflicts of interest to report.

The use of cannabis at a younger age is associated with psychosis symptoms in early adulthood, according to a study published online in the Archives of General Psychiatry.

Such a link has been reported before, but this is the first study to demonstrate the association in a subgroup of sibling pairs, “thus reducing the likelihood that the association was due to unmeasured shared genetic and/or environmental influences,” said Dr. John McGrath of the Queensland Brain Institute, Wacol, Australia, and his colleagues.

In addition, their study demonstrated a dose-response relationship between younger age at first marijuana use and higher risk of psychosis-related outcomes, they noted.

Earlier studies had found a link between early-onset cannabis use and later symptoms of psychosis, but there were “lingering concerns that the association may reflect methodological biases and unmeasured residual confounding” in those studies. Dr. McGrath and his associates examined the link using data from a birth cohort of more than 7,000 mother-infant pairs who were first studied in 1981–1984 and followed up 5, 14, and 21 years later.

A total of 3,801 of these infants and their close-in-age siblings comprised the subjects in this study. The latest follow-up occurred when they were aged 18–23 years. At that time, about 18% of the subjects said they had been using marijuana for 3 or fewer years; 16% said they had been using it for 4–5 years; and 14% said they had been using it for 6 or more years.

At this final follow-up, 65 of these subjects had received diagnoses of nonaffective psychosis because they met the criteria for schizophrenia (53 subjects), persistent delusional disorder (3), or acute transient psychotic disorders (9). An additional 233 subjects reported at least one visual or auditory hallucination on the Composite International Diagnostic Interview (CIDI).

Only subjects with the longest duration since first cannabis use–that is, those who started using marijuana at age 15 years or younger–were at significantly increased risk for developing symptoms of nonaffective psychosis in young adulthood.

Those who started using marijuana at that age were twice as likely to receive such a diagnosis than were subjects who said they had never used marijuana, the researchers said (Arch. Gen. Psychiatry 2010 [doi:10.1001/archgenpsychiatry.2010.6]).

Compared with subjects who did not use cannabis, those who used it at a younger age were 4 times more likely to score in the top quartile on the 21-item Peters et al. Delusional Inventory (PDI) and to report hallucinations on the CIDI.

Moreover, the longer the interval since first cannabis use, the higher the risk of these adverse psychosis-related outcomes.

In a subsample of 218 sibling pairs, there was a significant association between earlier first use of cannabis and higher scores on the PDI. For every additional year since first exposure to marijuana, the sibling with the younger age at first use scored one item higher than the other sibling, wrote the authors.

This study could delineate an association between early marijuana use and later symptoms of psychosis, and was not designed to determine causality. “We cannot confidently exclude the possibility that some of the cohort members may have developed psychosis as young adolescents, which may have contributed to subsequent [early] cannabis use,” they added.

The National Health and Medical Research Council of Australia funded the study.

Major Finding: Those who started using marijuana at 15 were twice as likely to receive a diagnosis of nonaffective psychosis in young adulthood as their counterparts who said they never used.

Data Source: Prospective, sibling pair analysis of 3,081 adults born between 1981 and 1984.

Disclosures: None of the investigators had any financial conflicts of interest to report.

The use of cannabis at a younger age is associated with psychosis symptoms in early adulthood, according to a study published online in the Archives of General Psychiatry.

Such a link has been reported before, but this is the first study to demonstrate the association in a subgroup of sibling pairs, “thus reducing the likelihood that the association was due to unmeasured shared genetic and/or environmental influences,” said Dr. John McGrath of the Queensland Brain Institute, Wacol, Australia, and his colleagues.

In addition, their study demonstrated a dose-response relationship between younger age at first marijuana use and higher risk of psychosis-related outcomes, they noted.

Earlier studies had found a link between early-onset cannabis use and later symptoms of psychosis, but there were “lingering concerns that the association may reflect methodological biases and unmeasured residual confounding” in those studies. Dr. McGrath and his associates examined the link using data from a birth cohort of more than 7,000 mother-infant pairs who were first studied in 1981–1984 and followed up 5, 14, and 21 years later.

A total of 3,801 of these infants and their close-in-age siblings comprised the subjects in this study. The latest follow-up occurred when they were aged 18–23 years. At that time, about 18% of the subjects said they had been using marijuana for 3 or fewer years; 16% said they had been using it for 4–5 years; and 14% said they had been using it for 6 or more years.

At this final follow-up, 65 of these subjects had received diagnoses of nonaffective psychosis because they met the criteria for schizophrenia (53 subjects), persistent delusional disorder (3), or acute transient psychotic disorders (9). An additional 233 subjects reported at least one visual or auditory hallucination on the Composite International Diagnostic Interview (CIDI).

Only subjects with the longest duration since first cannabis use–that is, those who started using marijuana at age 15 years or younger–were at significantly increased risk for developing symptoms of nonaffective psychosis in young adulthood.

Those who started using marijuana at that age were twice as likely to receive such a diagnosis than were subjects who said they had never used marijuana, the researchers said (Arch. Gen. Psychiatry 2010 [doi:10.1001/archgenpsychiatry.2010.6]).

Compared with subjects who did not use cannabis, those who used it at a younger age were 4 times more likely to score in the top quartile on the 21-item Peters et al. Delusional Inventory (PDI) and to report hallucinations on the CIDI.

Moreover, the longer the interval since first cannabis use, the higher the risk of these adverse psychosis-related outcomes.

In a subsample of 218 sibling pairs, there was a significant association between earlier first use of cannabis and higher scores on the PDI. For every additional year since first exposure to marijuana, the sibling with the younger age at first use scored one item higher than the other sibling, wrote the authors.

This study could delineate an association between early marijuana use and later symptoms of psychosis, and was not designed to determine causality. “We cannot confidently exclude the possibility that some of the cohort members may have developed psychosis as young adolescents, which may have contributed to subsequent [early] cannabis use,” they added.

The National Health and Medical Research Council of Australia funded the study.

A mandatory influenza vaccination program for all employees in a large Midwestern health care organization increased the rate of immunization to more than 98%, according to a study.

Previous, nonmandatory efforts in this organization as well as in other health care facilities across the country have had much more limited efficacy. The average influenza vaccination rate among U.S. health care workers was only 44% in recent years, said Dr. Hilary M. Babcock of Washington University, St. Louis, and her associates.

They reported the first study in the medical literature to describe outcomes of a multihospital health care organization's mandatory vaccination program. The program, an initiative focused on patient safety, required seasonal influenza vaccination as a condition of employment for clinical and nonclinical staff, contracted clinical personnel, and volunteers.

The program was implemented for the 2008-2009 flu season at BJC HealthCare, a network with approximately 26,000 employees working at 11 acute care hospitals and 3 extended care facilities, as well as day care centers, physician group practices, occupational medicine providers, home care providers, and behavioral health services in urban, suburban, and rural settings.

Employees were offered free immunizations, including thimerosal-free and intranasal formulations, at multiple locations and times in each facility. They were encouraged to review educational materials emphasizing patient safety and to consult with a medical director to discuss any concerns. They could request medical or religious exemptions using a standardized declination statement.

“Exemption requests often reflected misinformation about the vaccine…. Several requests cited chemotherapy or an immunosuppressed state as reasons not to get the vaccine, even though these groups are at high risk for complications from influenza and are specifically recommended to be vaccinated. Several requests cited pregnancy, although the vaccine is recommended during pregnancy,” the investigators noted.

Declination requests were reviewed by medical and human resources personnel, then accepted or denied. Employees who were neither vaccinated nor exempted by mid-December were suspended without pay. Those who were still not vaccinated nor exempted by mid-January were terminated for failing to meet the conditions of their employment.

A total of 25,561 employees were vaccinated (98.4%). All the physicians in the network were immunized, including all 907 residents and fellows, said Dr. Babcock, who is also medical director of occupational health at Barnes-Jewish and St. Louis Children's Hospitals, and her colleagues.

Another 321 employees (1.24%) received medical exemptions for reasons such as allergy to eggs, prior allergic reaction to a vaccine, or a history of Guillain-Barré syndrome. Ninety employees (0.35%) received religious exemptions. Eight employees (0.03%) were terminated.

This represents a 43% increase in the vaccination rate compared with the rate in 2006 and a 27% increase compared with 2007 at BJC HealthCare, Dr. Babcock and her associates said (Clin. Infect. Dis. 2010;50:459-64).

In an editorial, Dr. Andrew T. Pavia of the division of pediatric infectious diseases at the University of Utah, Salt Lake City, agreed that this program was “highly successful,” particularly when compared with the “very modest” success of efforts over the past 10 years to increase the “unacceptably low” rate of vaccination among U.S. health care workers.

“Mandatory vaccination policies have been endorsed by several organizations, including the New York State Department of Health, the Infectious Diseases Society of America, the American College of Physicians, the Association for Professionals in Infection Control and Epidemiology, and the National Foundation for Patient Safety,” he noted.

Although “some of the largest and most prestigious” health care organizations in the nation have adopted mandatory vaccination policies. “mandatory vaccination has also generated vigorous debate and opposition, including legal challenges,” Dr. Pavia said (Clin. Infect. Dis. 2010;50:465-7). “The debate should focus on results, not intentions or methods. Health care organizations should be expected to achieve influenza vaccine coverage that optimizes patient safety and to make data on coverage readily available t he noted.

“I propose 90% coverage as an appropriate target. Organizations can then choose to achieve the target with less coercive methods if they can or, if necessary, choose to mandate vaccination,” Dr. Pavia said.

Disclosures: Dr. Babcock and her associates reported no potential conflicts of interest. Dr. Pavia reported serving as a consultant to NexBio.

A mandatory influenza vaccination program for all employees in a large Midwestern health care organization increased the rate of immunization to more than 98%, according to a study.

Previous, nonmandatory efforts in this organization as well as in other health care facilities across the country have had much more limited efficacy. The average influenza vaccination rate among U.S. health care workers was only 44% in recent years, said Dr. Hilary M. Babcock of Washington University, St. Louis, and her associates.

They reported the first study in the medical literature to describe outcomes of a multihospital health care organization's mandatory vaccination program. The program, an initiative focused on patient safety, required seasonal influenza vaccination as a condition of employment for clinical and nonclinical staff, contracted clinical personnel, and volunteers.

The program was implemented for the 2008-2009 flu season at BJC HealthCare, a network with approximately 26,000 employees working at 11 acute care hospitals and 3 extended care facilities, as well as day care centers, physician group practices, occupational medicine providers, home care providers, and behavioral health services in urban, suburban, and rural settings.

Employees were offered free immunizations, including thimerosal-free and intranasal formulations, at multiple locations and times in each facility. They were encouraged to review educational materials emphasizing patient safety and to consult with a medical director to discuss any concerns. They could request medical or religious exemptions using a standardized declination statement.

“Exemption requests often reflected misinformation about the vaccine…. Several requests cited chemotherapy or an immunosuppressed state as reasons not to get the vaccine, even though these groups are at high risk for complications from influenza and are specifically recommended to be vaccinated. Several requests cited pregnancy, although the vaccine is recommended during pregnancy,” the investigators noted.

Declination requests were reviewed by medical and human resources personnel, then accepted or denied. Employees who were neither vaccinated nor exempted by mid-December were suspended without pay. Those who were still not vaccinated nor exempted by mid-January were terminated for failing to meet the conditions of their employment.

A total of 25,561 employees were vaccinated (98.4%). All the physicians in the network were immunized, including all 907 residents and fellows, said Dr. Babcock, who is also medical director of occupational health at Barnes-Jewish and St. Louis Children's Hospitals, and her colleagues.

Another 321 employees (1.24%) received medical exemptions for reasons such as allergy to eggs, prior allergic reaction to a vaccine, or a history of Guillain-Barré syndrome. Ninety employees (0.35%) received religious exemptions. Eight employees (0.03%) were terminated.

This represents a 43% increase in the vaccination rate compared with the rate in 2006 and a 27% increase compared with 2007 at BJC HealthCare, Dr. Babcock and her associates said (Clin. Infect. Dis. 2010;50:459-64).

In an editorial, Dr. Andrew T. Pavia of the division of pediatric infectious diseases at the University of Utah, Salt Lake City, agreed that this program was “highly successful,” particularly when compared with the “very modest” success of efforts over the past 10 years to increase the “unacceptably low” rate of vaccination among U.S. health care workers.

“Mandatory vaccination policies have been endorsed by several organizations, including the New York State Department of Health, the Infectious Diseases Society of America, the American College of Physicians, the Association for Professionals in Infection Control and Epidemiology, and the National Foundation for Patient Safety,” he noted.

Although “some of the largest and most prestigious” health care organizations in the nation have adopted mandatory vaccination policies. “mandatory vaccination has also generated vigorous debate and opposition, including legal challenges,” Dr. Pavia said (Clin. Infect. Dis. 2010;50:465-7). “The debate should focus on results, not intentions or methods. Health care organizations should be expected to achieve influenza vaccine coverage that optimizes patient safety and to make data on coverage readily available t he noted.

“I propose 90% coverage as an appropriate target. Organizations can then choose to achieve the target with less coercive methods if they can or, if necessary, choose to mandate vaccination,” Dr. Pavia said.

Disclosures: Dr. Babcock and her associates reported no potential conflicts of interest. Dr. Pavia reported serving as a consultant to NexBio.

A mandatory influenza vaccination program for all employees in a large Midwestern health care organization increased the rate of immunization to more than 98%, according to a study.

Previous, nonmandatory efforts in this organization as well as in other health care facilities across the country have had much more limited efficacy. The average influenza vaccination rate among U.S. health care workers was only 44% in recent years, said Dr. Hilary M. Babcock of Washington University, St. Louis, and her associates.

They reported the first study in the medical literature to describe outcomes of a multihospital health care organization's mandatory vaccination program. The program, an initiative focused on patient safety, required seasonal influenza vaccination as a condition of employment for clinical and nonclinical staff, contracted clinical personnel, and volunteers.

The program was implemented for the 2008-2009 flu season at BJC HealthCare, a network with approximately 26,000 employees working at 11 acute care hospitals and 3 extended care facilities, as well as day care centers, physician group practices, occupational medicine providers, home care providers, and behavioral health services in urban, suburban, and rural settings.

Employees were offered free immunizations, including thimerosal-free and intranasal formulations, at multiple locations and times in each facility. They were encouraged to review educational materials emphasizing patient safety and to consult with a medical director to discuss any concerns. They could request medical or religious exemptions using a standardized declination statement.

“Exemption requests often reflected misinformation about the vaccine…. Several requests cited chemotherapy or an immunosuppressed state as reasons not to get the vaccine, even though these groups are at high risk for complications from influenza and are specifically recommended to be vaccinated. Several requests cited pregnancy, although the vaccine is recommended during pregnancy,” the investigators noted.

Declination requests were reviewed by medical and human resources personnel, then accepted or denied. Employees who were neither vaccinated nor exempted by mid-December were suspended without pay. Those who were still not vaccinated nor exempted by mid-January were terminated for failing to meet the conditions of their employment.

A total of 25,561 employees were vaccinated (98.4%). All the physicians in the network were immunized, including all 907 residents and fellows, said Dr. Babcock, who is also medical director of occupational health at Barnes-Jewish and St. Louis Children's Hospitals, and her colleagues.

Another 321 employees (1.24%) received medical exemptions for reasons such as allergy to eggs, prior allergic reaction to a vaccine, or a history of Guillain-Barré syndrome. Ninety employees (0.35%) received religious exemptions. Eight employees (0.03%) were terminated.

This represents a 43% increase in the vaccination rate compared with the rate in 2006 and a 27% increase compared with 2007 at BJC HealthCare, Dr. Babcock and her associates said (Clin. Infect. Dis. 2010;50:459-64).

In an editorial, Dr. Andrew T. Pavia of the division of pediatric infectious diseases at the University of Utah, Salt Lake City, agreed that this program was “highly successful,” particularly when compared with the “very modest” success of efforts over the past 10 years to increase the “unacceptably low” rate of vaccination among U.S. health care workers.

“Mandatory vaccination policies have been endorsed by several organizations, including the New York State Department of Health, the Infectious Diseases Society of America, the American College of Physicians, the Association for Professionals in Infection Control and Epidemiology, and the National Foundation for Patient Safety,” he noted.

Although “some of the largest and most prestigious” health care organizations in the nation have adopted mandatory vaccination policies. “mandatory vaccination has also generated vigorous debate and opposition, including legal challenges,” Dr. Pavia said (Clin. Infect. Dis. 2010;50:465-7). “The debate should focus on results, not intentions or methods. Health care organizations should be expected to achieve influenza vaccine coverage that optimizes patient safety and to make data on coverage readily available t he noted.

“I propose 90% coverage as an appropriate target. Organizations can then choose to achieve the target with less coercive methods if they can or, if necessary, choose to mandate vaccination,” Dr. Pavia said.

Disclosures: Dr. Babcock and her associates reported no potential conflicts of interest. Dr. Pavia reported serving as a consultant to NexBio.

A single negative whole-leg compression ultrasound may safely identify which patients with suspected deep vein thrombosis can forego anticoagulation therapy because they are at low risk for venous thromboembolism, according to a meta-analysis.

But the authors of an editorial cautioned against drawing firm clinical conclusions from the meta-analysis.

For patients assessed for possible DVT, practice guidelines currently recommend serial compression ultrasound imaging of the proximal veins after an initial negative result. Such imaging may minimize the risk that distal DVT is present and could propagate into the proximal veins, putting the patient at risk for VTE. However, only 1%-2% of these repeat studies detect thrombus propagation, making many of the studies ultimately unnecessary.

“Because many distal thrombi appear to resolve without use of anticoagulant therapy, it may be argued that detection and treatment of distal DVT is unnecessary because it may place patients at undue risk for anticoagulant-related complications,” wrote Dr. Stacy A. Johnson of the University of Utah, Salt Lake City, and associates (JAMA 2010;303:438-45).

Whole-leg compression ultrasound (CUS) has been proposed as an alternative strategy to improve initial detection of distal DVT and obviate repeat compression ultrasound. But many clinicians are reluctant to rely on a single whole-leg CUS for that purpose, citing concerns about the technical feasibility and safety of such an approach, the investigators noted.

The researchers performed a meta-analysis “to address the safety of withholding anticoagulation after a negative whole-leg CUS by providing estimates of the incidence of symptomatic VTE during the 3 months after a single negative result.” They reviewed 156 studies and limited the meta-analysis to 6 prospective cohort studies and 1 randomized clinical trial. Outcomes for 4,731 patients were included.

The combined end point of confirmed VTE and mortality possibly related to VTE developed in 34 (0.7%) of the patients. There were 11 cases of distal VTE, 7 cases of proximal DVT, and 7 cases of nonfatal pulmonary embolism.

Nine deaths may have been related to VTE, but no necropsies were done to establish the causes of death. All of the deaths occurred in acutely ill hospitalized patients or patients with advanced cancer.

“Overall, the risk for symptomatic VTE was low, with a pooled VTE event rate of 0.57%,” the researchers said. “To our knowledge, these results represent the first reported pooled risk assessment of VTE following a negative lower extremity whole-leg CUS result.”

However, “summary statements from meta-analyses should not be used to guide patient care,” cautioned Robert A. McNutt, M.D., Ph.D., of Rush University Medical Center, Chicago, and Dr. Edward H. Livingston of the University of Texas Southwestern Medical Center, Dallas, in their editorial. “Such conclusions are not helpful when the clinical studies are combined and averaged in a way that reduces the complex world of medical care to overly simple and consequently not clinically useful statistical summaries,” they said (JAMA 2010;303:454-5).

“Generalizing the findings related to a diagnostic test or treatment regimen beyond the specific context from which a study was performed is fraught with danger,” Dr. McNutt and Dr. Livingston noted. “For instance, based on the meta-analysis by Johnson et al., clinicians may infer that not initiating anticoagulation treatment after a negative CUS result in some surgical or ambulatory patients at low risk of having VTE may be appropriate; however, that inference may not be true for hospitalized patients or those with cancer.

“Greater detail about individual patient scenarios is necessary to facilitate better application of the study results.”

Disclosures: Dr. Johnson's associates reported receiving consulting and speaker's fees from AGEN Biomedical, Janssen-Ortho, Boehringer Ingelheim, Sanofi-Aventis, AstraZeneca, Pfizer, and Leo Pharma. The editorialists reported that they had no financial disclosures.

A single negative whole-leg compression ultrasound may safely identify which patients with suspected deep vein thrombosis can forego anticoagulation therapy because they are at low risk for venous thromboembolism, according to a meta-analysis.

But the authors of an editorial cautioned against drawing firm clinical conclusions from the meta-analysis.

For patients assessed for possible DVT, practice guidelines currently recommend serial compression ultrasound imaging of the proximal veins after an initial negative result. Such imaging may minimize the risk that distal DVT is present and could propagate into the proximal veins, putting the patient at risk for VTE. However, only 1%-2% of these repeat studies detect thrombus propagation, making many of the studies ultimately unnecessary.

“Because many distal thrombi appear to resolve without use of anticoagulant therapy, it may be argued that detection and treatment of distal DVT is unnecessary because it may place patients at undue risk for anticoagulant-related complications,” wrote Dr. Stacy A. Johnson of the University of Utah, Salt Lake City, and associates (JAMA 2010;303:438-45).

Whole-leg compression ultrasound (CUS) has been proposed as an alternative strategy to improve initial detection of distal DVT and obviate repeat compression ultrasound. But many clinicians are reluctant to rely on a single whole-leg CUS for that purpose, citing concerns about the technical feasibility and safety of such an approach, the investigators noted.

The researchers performed a meta-analysis “to address the safety of withholding anticoagulation after a negative whole-leg CUS by providing estimates of the incidence of symptomatic VTE during the 3 months after a single negative result.” They reviewed 156 studies and limited the meta-analysis to 6 prospective cohort studies and 1 randomized clinical trial. Outcomes for 4,731 patients were included.

The combined end point of confirmed VTE and mortality possibly related to VTE developed in 34 (0.7%) of the patients. There were 11 cases of distal VTE, 7 cases of proximal DVT, and 7 cases of nonfatal pulmonary embolism.

Nine deaths may have been related to VTE, but no necropsies were done to establish the causes of death. All of the deaths occurred in acutely ill hospitalized patients or patients with advanced cancer.

“Overall, the risk for symptomatic VTE was low, with a pooled VTE event rate of 0.57%,” the researchers said. “To our knowledge, these results represent the first reported pooled risk assessment of VTE following a negative lower extremity whole-leg CUS result.”

However, “summary statements from meta-analyses should not be used to guide patient care,” cautioned Robert A. McNutt, M.D., Ph.D., of Rush University Medical Center, Chicago, and Dr. Edward H. Livingston of the University of Texas Southwestern Medical Center, Dallas, in their editorial. “Such conclusions are not helpful when the clinical studies are combined and averaged in a way that reduces the complex world of medical care to overly simple and consequently not clinically useful statistical summaries,” they said (JAMA 2010;303:454-5).

“Generalizing the findings related to a diagnostic test or treatment regimen beyond the specific context from which a study was performed is fraught with danger,” Dr. McNutt and Dr. Livingston noted. “For instance, based on the meta-analysis by Johnson et al., clinicians may infer that not initiating anticoagulation treatment after a negative CUS result in some surgical or ambulatory patients at low risk of having VTE may be appropriate; however, that inference may not be true for hospitalized patients or those with cancer.

“Greater detail about individual patient scenarios is necessary to facilitate better application of the study results.”

Disclosures: Dr. Johnson's associates reported receiving consulting and speaker's fees from AGEN Biomedical, Janssen-Ortho, Boehringer Ingelheim, Sanofi-Aventis, AstraZeneca, Pfizer, and Leo Pharma. The editorialists reported that they had no financial disclosures.

A single negative whole-leg compression ultrasound may safely identify which patients with suspected deep vein thrombosis can forego anticoagulation therapy because they are at low risk for venous thromboembolism, according to a meta-analysis.

But the authors of an editorial cautioned against drawing firm clinical conclusions from the meta-analysis.

For patients assessed for possible DVT, practice guidelines currently recommend serial compression ultrasound imaging of the proximal veins after an initial negative result. Such imaging may minimize the risk that distal DVT is present and could propagate into the proximal veins, putting the patient at risk for VTE. However, only 1%-2% of these repeat studies detect thrombus propagation, making many of the studies ultimately unnecessary.

“Because many distal thrombi appear to resolve without use of anticoagulant therapy, it may be argued that detection and treatment of distal DVT is unnecessary because it may place patients at undue risk for anticoagulant-related complications,” wrote Dr. Stacy A. Johnson of the University of Utah, Salt Lake City, and associates (JAMA 2010;303:438-45).

Whole-leg compression ultrasound (CUS) has been proposed as an alternative strategy to improve initial detection of distal DVT and obviate repeat compression ultrasound. But many clinicians are reluctant to rely on a single whole-leg CUS for that purpose, citing concerns about the technical feasibility and safety of such an approach, the investigators noted.

The researchers performed a meta-analysis “to address the safety of withholding anticoagulation after a negative whole-leg CUS by providing estimates of the incidence of symptomatic VTE during the 3 months after a single negative result.” They reviewed 156 studies and limited the meta-analysis to 6 prospective cohort studies and 1 randomized clinical trial. Outcomes for 4,731 patients were included.

The combined end point of confirmed VTE and mortality possibly related to VTE developed in 34 (0.7%) of the patients. There were 11 cases of distal VTE, 7 cases of proximal DVT, and 7 cases of nonfatal pulmonary embolism.

Nine deaths may have been related to VTE, but no necropsies were done to establish the causes of death. All of the deaths occurred in acutely ill hospitalized patients or patients with advanced cancer.

“Overall, the risk for symptomatic VTE was low, with a pooled VTE event rate of 0.57%,” the researchers said. “To our knowledge, these results represent the first reported pooled risk assessment of VTE following a negative lower extremity whole-leg CUS result.”

However, “summary statements from meta-analyses should not be used to guide patient care,” cautioned Robert A. McNutt, M.D., Ph.D., of Rush University Medical Center, Chicago, and Dr. Edward H. Livingston of the University of Texas Southwestern Medical Center, Dallas, in their editorial. “Such conclusions are not helpful when the clinical studies are combined and averaged in a way that reduces the complex world of medical care to overly simple and consequently not clinically useful statistical summaries,” they said (JAMA 2010;303:454-5).

“Generalizing the findings related to a diagnostic test or treatment regimen beyond the specific context from which a study was performed is fraught with danger,” Dr. McNutt and Dr. Livingston noted. “For instance, based on the meta-analysis by Johnson et al., clinicians may infer that not initiating anticoagulation treatment after a negative CUS result in some surgical or ambulatory patients at low risk of having VTE may be appropriate; however, that inference may not be true for hospitalized patients or those with cancer.

“Greater detail about individual patient scenarios is necessary to facilitate better application of the study results.”

Disclosures: Dr. Johnson's associates reported receiving consulting and speaker's fees from AGEN Biomedical, Janssen-Ortho, Boehringer Ingelheim, Sanofi-Aventis, AstraZeneca, Pfizer, and Leo Pharma. The editorialists reported that they had no financial disclosures.

Current methods for profiling individual physicians as to whether they provide low-cost or high-cost care are often inaccurate and produce misleading results, according to a report in the March 18 issue of the New England Journal of Medicine.

Health plans use cost profiling to limit how many physicians get in-network contracts and to allot bonuses to the physicians whose "resource use" is lower than average. In each case, there must be a method for determining physicians' costs, yet the accuracy of these methods has never been proven, according to John L. Adams, Ph.D., of Rand Corp., Santa Monica, Calif., and his associates.

"To our knowledge, the reliability of physician cost profiling has not been previously addressed," they noted.

Dr. Adams and his colleagues assessed the reliability of current methods of cost profiling using claims data from four Massachusetts insurance companies concerning 1.1 million adult patients treated during 2004-2005. A total of 12,789 physicians were included in the study. They were predominantly men who were board certified, had been trained in the United States, and had been in practice for more than 10 years.

The physicians worked in 28 specialties, including cardiology, endocrinology, gastroenterology, and obstetrics and gynecology. Family physicians, general physicians, and internal medicine physicians comprised approximately one-third of the sample.

The investigators estimated the reliability of cost profiles on a scale of 0-1, with 0 representing completely unreliable profiles and 1 representing completely reliable profiles. They then estimated the proportion of physicians in each specialty whose cost performance would be calculated inaccurately.

Overall, only 41% of physicians across all specialties had cost profile scores of 0.70 or greater, a commonly used threshold of acceptable accuracy. Only 47% of internists, 30% of cardiologists, 41% of family or general physicians, 57% of ob.gyns., 59% of gastroenterologists, and 22% of endocrinologists received scores of 0.70.

Overall, only 9% of physicians in the study had scores of 0.90 or greater, indicating optimal accuracy.

The proportion of physicians who were classified as "lower cost" but who were not in fact lower cost ranged from 29% to 67%, depending on the specialty. Fully 50% of internists, 40% of cardiologists, 39% of family or general physicians, 36% of ob.gyns., 32% of gastroenterologists, and 50% of endocrinologists were misclassified as "lower-cost" providers when they were not.

In addition, 22% of internists were misclassified as "higher cost" when they were not in fact higher cost. This same misclassification occurred as well for 14% of cardiologists, 16% of family or general physicians, 10% of ob.gyns., 11% of gastroenterologists, and 19% of endocrinologists.

These findings indicate that standard methods of cost profiling are highly unreliable, and that many individuals and groups are basing important decisions on inaccuracies. "Consumers, physicians, and purchasers are all at risk of being misled by the results produced by these tools," the investigators concluded (N. Engl. J. Med. 2010;362:1014-21).

The study findings also suggest that using cost profiles that are based on these unreliable methods will not reduce health care spending. "There are serious threats to insurance plans' abilities to achieve cost-control objectives and to patients' expectations of receiving lower-cost care when they change physicians for that purpose," they added.

This study was supported by the Department of Labor, the National Institutes of Health, and the Robert Wood Johnson Foundation. The investigators' conflicts of interest include support from the Integrated Healthcare Association, American Medical Association, American Board of Medical Specialties, Arkansas Medical Society, American Board of Internal Medicine Foundation, Massachusetts Medical Society, Physicians Advocacy Institute, Commonwealth Fund, and Ingenix Inc.

Current methods for profiling individual physicians as to whether they provide low-cost or high-cost care are often inaccurate and produce misleading results, according to a report in the March 18 issue of the New England Journal of Medicine.

Health plans use cost profiling to limit how many physicians get in-network contracts and to allot bonuses to the physicians whose "resource use" is lower than average. In each case, there must be a method for determining physicians' costs, yet the accuracy of these methods has never been proven, according to John L. Adams, Ph.D., of Rand Corp., Santa Monica, Calif., and his associates.

"To our knowledge, the reliability of physician cost profiling has not been previously addressed," they noted.

Dr. Adams and his colleagues assessed the reliability of current methods of cost profiling using claims data from four Massachusetts insurance companies concerning 1.1 million adult patients treated during 2004-2005. A total of 12,789 physicians were included in the study. They were predominantly men who were board certified, had been trained in the United States, and had been in practice for more than 10 years.

The physicians worked in 28 specialties, including cardiology, endocrinology, gastroenterology, and obstetrics and gynecology. Family physicians, general physicians, and internal medicine physicians comprised approximately one-third of the sample.

The investigators estimated the reliability of cost profiles on a scale of 0-1, with 0 representing completely unreliable profiles and 1 representing completely reliable profiles. They then estimated the proportion of physicians in each specialty whose cost performance would be calculated inaccurately.

Overall, only 41% of physicians across all specialties had cost profile scores of 0.70 or greater, a commonly used threshold of acceptable accuracy. Only 47% of internists, 30% of cardiologists, 41% of family or general physicians, 57% of ob.gyns., 59% of gastroenterologists, and 22% of endocrinologists received scores of 0.70.

Overall, only 9% of physicians in the study had scores of 0.90 or greater, indicating optimal accuracy.

The proportion of physicians who were classified as "lower cost" but who were not in fact lower cost ranged from 29% to 67%, depending on the specialty. Fully 50% of internists, 40% of cardiologists, 39% of family or general physicians, 36% of ob.gyns., 32% of gastroenterologists, and 50% of endocrinologists were misclassified as "lower-cost" providers when they were not.

In addition, 22% of internists were misclassified as "higher cost" when they were not in fact higher cost. This same misclassification occurred as well for 14% of cardiologists, 16% of family or general physicians, 10% of ob.gyns., 11% of gastroenterologists, and 19% of endocrinologists.

These findings indicate that standard methods of cost profiling are highly unreliable, and that many individuals and groups are basing important decisions on inaccuracies. "Consumers, physicians, and purchasers are all at risk of being misled by the results produced by these tools," the investigators concluded (N. Engl. J. Med. 2010;362:1014-21).

The study findings also suggest that using cost profiles that are based on these unreliable methods will not reduce health care spending. "There are serious threats to insurance plans' abilities to achieve cost-control objectives and to patients' expectations of receiving lower-cost care when they change physicians for that purpose," they added.

This study was supported by the Department of Labor, the National Institutes of Health, and the Robert Wood Johnson Foundation. The investigators' conflicts of interest include support from the Integrated Healthcare Association, American Medical Association, American Board of Medical Specialties, Arkansas Medical Society, American Board of Internal Medicine Foundation, Massachusetts Medical Society, Physicians Advocacy Institute, Commonwealth Fund, and Ingenix Inc.

Current methods for profiling individual physicians as to whether they provide low-cost or high-cost care are often inaccurate and produce misleading results, according to a report in the March 18 issue of the New England Journal of Medicine.

Health plans use cost profiling to limit how many physicians get in-network contracts and to allot bonuses to the physicians whose "resource use" is lower than average. In each case, there must be a method for determining physicians' costs, yet the accuracy of these methods has never been proven, according to John L. Adams, Ph.D., of Rand Corp., Santa Monica, Calif., and his associates.

"To our knowledge, the reliability of physician cost profiling has not been previously addressed," they noted.

Dr. Adams and his colleagues assessed the reliability of current methods of cost profiling using claims data from four Massachusetts insurance companies concerning 1.1 million adult patients treated during 2004-2005. A total of 12,789 physicians were included in the study. They were predominantly men who were board certified, had been trained in the United States, and had been in practice for more than 10 years.

The physicians worked in 28 specialties, including cardiology, endocrinology, gastroenterology, and obstetrics and gynecology. Family physicians, general physicians, and internal medicine physicians comprised approximately one-third of the sample.

The investigators estimated the reliability of cost profiles on a scale of 0-1, with 0 representing completely unreliable profiles and 1 representing completely reliable profiles. They then estimated the proportion of physicians in each specialty whose cost performance would be calculated inaccurately.

Overall, only 41% of physicians across all specialties had cost profile scores of 0.70 or greater, a commonly used threshold of acceptable accuracy. Only 47% of internists, 30% of cardiologists, 41% of family or general physicians, 57% of ob.gyns., 59% of gastroenterologists, and 22% of endocrinologists received scores of 0.70.

Overall, only 9% of physicians in the study had scores of 0.90 or greater, indicating optimal accuracy.

The proportion of physicians who were classified as "lower cost" but who were not in fact lower cost ranged from 29% to 67%, depending on the specialty. Fully 50% of internists, 40% of cardiologists, 39% of family or general physicians, 36% of ob.gyns., 32% of gastroenterologists, and 50% of endocrinologists were misclassified as "lower-cost" providers when they were not.

In addition, 22% of internists were misclassified as "higher cost" when they were not in fact higher cost. This same misclassification occurred as well for 14% of cardiologists, 16% of family or general physicians, 10% of ob.gyns., 11% of gastroenterologists, and 19% of endocrinologists.

These findings indicate that standard methods of cost profiling are highly unreliable, and that many individuals and groups are basing important decisions on inaccuracies. "Consumers, physicians, and purchasers are all at risk of being misled by the results produced by these tools," the investigators concluded (N. Engl. J. Med. 2010;362:1014-21).

The study findings also suggest that using cost profiles that are based on these unreliable methods will not reduce health care spending. "There are serious threats to insurance plans' abilities to achieve cost-control objectives and to patients' expectations of receiving lower-cost care when they change physicians for that purpose," they added.

This study was supported by the Department of Labor, the National Institutes of Health, and the Robert Wood Johnson Foundation. The investigators' conflicts of interest include support from the Integrated Healthcare Association, American Medical Association, American Board of Medical Specialties, Arkansas Medical Society, American Board of Internal Medicine Foundation, Massachusetts Medical Society, Physicians Advocacy Institute, Commonwealth Fund, and Ingenix Inc.

Twelve to 15 million white patients living in the United States have had at least one nonmelanoma skin cancer in their lifetimes, according to estimates based on a new mathematical model reported in the March issue of the Archives of Dermatology.

This figure is approximately twice that of previous estimates based on patient surveys, such as the estimate calculated in the U.S. National Health Interview Study, according to Dr. Robert S. Stern of Beth Israel Deaconess Medical Center and Harvard Medical School, Boston.

This difference can be attributed in part to people often incorrectly reporting their skin cancer histories when surveyed, falsely believing that basal cell and squamous cell lesions are not cancerous or that all skin cancers can be considered melanoma.

Dr. Stern devised his mathematical model using the same basic data available from national samples, such as the Surveillance, Epidemiology, and End Results (SEER) studies and information from the National Cancer Institute. His model, however, took into consideration several factors that had not been accounted for in previous estimates, such as the likelihood that patients develop numerous nonmelanoma skin cancers over the course of several years and that “a substantial proportion” of patients with melanoma also have nonmalignant skin cancers.

According to his model, “about 13 million white non-Hispanic U.S. residents (6%) have had more than 22 million nonmelanoma skin cancers” (Arch. Dermatol. 2010;146:279-82).

This puts the prevalence of a skin cancer history at a level far higher than that of any other cancer – prevalence that “exceeds that of all other cancers diagnosed since 1975,” he added. “Recent population-based data concerning skin cancer incidence, morbidity, and cost of care are lacking for the most common types of skin cancer, [basal cell carcinoma] and [squamous cell carcinoma].”

Regarding patients’ inaccurate reporting of skin cancer histories in surveys and interviews, Dr. Stern noted that many patients equate skin cancer with melanoma. “Hundreds of thousands of patients may be unnecessarily burdened with the belief that they have had a potentially lethal cancer (melanoma) when in fact they have had a skin tumor that is very unlikely to be lethal,he wrote.

In contrast, patients with breast or prostate cancer are much more likely to accurately report their cancer histories, so incidence and prevalence estimates for these tumors are much more accurate than those for skin cancer, Dr. Stern noted.

This study was funded in part by the National Institutes of Health. Dr. Stern reported being a consultant for Johnson & Johnson, Vertex, and Takeda and an expert witness for Alphapharm, Mutual Pharm, and Johnson & Johnson.

Twelve to 15 million white patients living in the United States have had at least one nonmelanoma skin cancer in their lifetimes, according to estimates based on a new mathematical model reported in the March issue of the Archives of Dermatology.

This figure is approximately twice that of previous estimates based on patient surveys, such as the estimate calculated in the U.S. National Health Interview Study, according to Dr. Robert S. Stern of Beth Israel Deaconess Medical Center and Harvard Medical School, Boston.

This difference can be attributed in part to people often incorrectly reporting their skin cancer histories when surveyed, falsely believing that basal cell and squamous cell lesions are not cancerous or that all skin cancers can be considered melanoma.

Dr. Stern devised his mathematical model using the same basic data available from national samples, such as the Surveillance, Epidemiology, and End Results (SEER) studies and information from the National Cancer Institute. His model, however, took into consideration several factors that had not been accounted for in previous estimates, such as the likelihood that patients develop numerous nonmelanoma skin cancers over the course of several years and that “a substantial proportion” of patients with melanoma also have nonmalignant skin cancers.

According to his model, “about 13 million white non-Hispanic U.S. residents (6%) have had more than 22 million nonmelanoma skin cancers” (Arch. Dermatol. 2010;146:279-82).

This puts the prevalence of a skin cancer history at a level far higher than that of any other cancer – prevalence that “exceeds that of all other cancers diagnosed since 1975,” he added. “Recent population-based data concerning skin cancer incidence, morbidity, and cost of care are lacking for the most common types of skin cancer, [basal cell carcinoma] and [squamous cell carcinoma].”

Regarding patients’ inaccurate reporting of skin cancer histories in surveys and interviews, Dr. Stern noted that many patients equate skin cancer with melanoma. “Hundreds of thousands of patients may be unnecessarily burdened with the belief that they have had a potentially lethal cancer (melanoma) when in fact they have had a skin tumor that is very unlikely to be lethal,he wrote.

In contrast, patients with breast or prostate cancer are much more likely to accurately report their cancer histories, so incidence and prevalence estimates for these tumors are much more accurate than those for skin cancer, Dr. Stern noted.

This study was funded in part by the National Institutes of Health. Dr. Stern reported being a consultant for Johnson & Johnson, Vertex, and Takeda and an expert witness for Alphapharm, Mutual Pharm, and Johnson & Johnson.

Twelve to 15 million white patients living in the United States have had at least one nonmelanoma skin cancer in their lifetimes, according to estimates based on a new mathematical model reported in the March issue of the Archives of Dermatology.

This figure is approximately twice that of previous estimates based on patient surveys, such as the estimate calculated in the U.S. National Health Interview Study, according to Dr. Robert S. Stern of Beth Israel Deaconess Medical Center and Harvard Medical School, Boston.

This difference can be attributed in part to people often incorrectly reporting their skin cancer histories when surveyed, falsely believing that basal cell and squamous cell lesions are not cancerous or that all skin cancers can be considered melanoma.

Dr. Stern devised his mathematical model using the same basic data available from national samples, such as the Surveillance, Epidemiology, and End Results (SEER) studies and information from the National Cancer Institute. His model, however, took into consideration several factors that had not been accounted for in previous estimates, such as the likelihood that patients develop numerous nonmelanoma skin cancers over the course of several years and that “a substantial proportion” of patients with melanoma also have nonmalignant skin cancers.

According to his model, “about 13 million white non-Hispanic U.S. residents (6%) have had more than 22 million nonmelanoma skin cancers” (Arch. Dermatol. 2010;146:279-82).

This puts the prevalence of a skin cancer history at a level far higher than that of any other cancer – prevalence that “exceeds that of all other cancers diagnosed since 1975,” he added. “Recent population-based data concerning skin cancer incidence, morbidity, and cost of care are lacking for the most common types of skin cancer, [basal cell carcinoma] and [squamous cell carcinoma].”

Regarding patients’ inaccurate reporting of skin cancer histories in surveys and interviews, Dr. Stern noted that many patients equate skin cancer with melanoma. “Hundreds of thousands of patients may be unnecessarily burdened with the belief that they have had a potentially lethal cancer (melanoma) when in fact they have had a skin tumor that is very unlikely to be lethal,he wrote.

In contrast, patients with breast or prostate cancer are much more likely to accurately report their cancer histories, so incidence and prevalence estimates for these tumors are much more accurate than those for skin cancer, Dr. Stern noted.

This study was funded in part by the National Institutes of Health. Dr. Stern reported being a consultant for Johnson & Johnson, Vertex, and Takeda and an expert witness for Alphapharm, Mutual Pharm, and Johnson & Johnson.