Stanley Feld M.D., FACP, MACE
Dr. Aaron Carroll and Dr. Ronald Ackerman, the principal authors of the 2003 survey and 2008 letter in the Annals of Internal Medicine, states very clearly that physicians support a national health insurance plan. They does not state whether physician support a government run single party payer plan.
"Many claim to speak for physicians and reflect their views. We asked doctors directly and found that, contrary to conventional wisdom, most doctors support the government creating national health insurance," study author Dr. Aaron E. Carroll, director of Indiana University's Center for Health Policy and Professionalism Research, said in a prepared statement.”
However they seem to be saying that physicians want the government to take over.
"Across the board, more physicians feel that our fragmented and for-profit insurance system is obstructing good patient care, and a majority now support national insurance as the remedy," Ackermann said in a prepared statement.”
The source data is not available in their 2008 letter published in the Annals of Internal Medicine. I have evaluated the source data in their 2003 survey. The survey consists of two questions. The questions have misleading implications. Both questions start by asking one “to assume the principal goal of any national health insurance proposal is to arrange health care financing for all U.S. citizens”. In 2008 I believe most physicians would agree there should be universal healthcare coverage. The real question is do physicians want the government as the single party payer?
Does that mean the government creates a plan that should cover everyone or does that question mean the government establishes a National Health Insurance that it administers as a single party payer?
Question 2 implies the former. “Do you support or oppose a National Health Insurance plan where the entire healthcare is paid for by the government?”
The survey design is flawed. Therefore it cannot yield valid conclusions. However, let us assume we could draw conclusions from the data in the survey.
Figure 2 describes the characteristics of the respondents in the survey.1650 physicians completed the survey sent to 3250 physicians. Only 1263 physicians had their characteristics compared to the AMA Physician Master file of 733,183 U.S. Physicians. What happened to the characteristics of the other 387 included in the survey? Does this 30% drop in number of physicians invalidate any power to arrive at conclusions for the data? Does the number of physicians in the survey represent a large enough number of physicians to represent the 733,180 physicians in the U.S.?
None of the mean insurance types, primary practice settings and primary practice locations was available in the AMA Physician Master file. This is one of several flaws in the study
The most important characteristics of the 1263 physician respondents was missing. How many physicians were in private practice? How many physicians in the survey group were employed and salaried by a university or hospital? What is the percentage of survey group physicians in private practice who received an institutional salary? These characteristics would have a effect on the physician response to the questions.
Additionally, any study whose published results present percentages without actual numbers or statistics has conclusions that are suspect.
Physician attitudes about National Health Insurance Financing are the most important table in the article.(Figure 3). The results are presented in the most misleading way. If a physician assumed that a national health insurance plan was universal coverage without a single party payer they might choose one answer to a quick survey. They might have thought the question meant the government would be the single party payer. 19% generally oppose, 21% strongly oppose and11% were neutral. The 11% neutral respondents might have been uncertain of the meaning of the question. If opposed to a single party payer they might have chosen to remain neutral. Therefore 51% of physicians might have opposed a national health plan as opposed to the 40% in the survey.
The answer to question 2 was important. It nullifies the principle authors’ message. 33% of physician strongly opposed, 27% generally opposed while 14% were neutral to a government paid for National Health Insurance plan. A neutral answer to this question could mean to some that the government would give me less grief than the healthcare industry has in recent years. Only 9% of physicians strongly supported the government as a single party payer. Again, we do not know the actual number of these physicians in private practice with monthly overhead. We do not know if the sample is valid or the results statistically significant. The number of physicians opposed to a single party payer was not discussed in either the 2003 or 2008 press release.
Only three predictors of physician support for governmental legislation to establish national health insurance plan were statistically significant. (Figure 4) The statistically significant predictors of support for national health insurance plan in reality support opposition to a single party payer according to the data presented. The statically positive predictors were inner city physicians vs. non inner city physicians, Medicaid vs. non Medicaid providers, and primary care vs. specialists. The predictor chart did not separate out the demographics of the physicians in each category.
In the last table the percentages of physicians in favor of the federal government as the sole payer was soundly defeated by all specialties. All of the specialties represented voted below 50% for the government to be the single party payer.
Here we see the data telling us one thing and the authors, through the use of the media providing a sensational but false conclusion from the data. The survey was poorly designed with little statistical significance.
Quoting Dr. Carroll, a primary author "Many claim to speak for physicians and reflect their views. We asked doctors directly and found that, contrary to conventional wisdom, most doctors support the government creating national health insurance,"
The moral of my story is to not believe anything until you see the data, the source of the data, and the methods used to evaluate its validity. Unfortunately, this is very hard to do. Refereed journals have been delegated as our surrogate evaluators. To my disappointment it seems the American College of Physicians has not fulfilled its obligation to practicing physicians of internal medicine.
It looks like the few governing physicians of the American College of Physicians and the Annals of Internal Medicine have chosen instead to pursue their own agenda and not the agenda of science at the expense of their member physicians and their patients.
The opinions expressed in the blog “Repairing The Healthcare System” are, mine and mine alone.