This article is more than 5 years old. Information may no longer be current.
BLOG: Salary, call pay and other surveys can be misleading
Whenever you enter a financial relationship with a hospital, the hospital is likely to rely on survey data to verify that the relationship is at fair market value. While the desire of the hospital to use such data is completely understandable, the odds are that the survey data is imperfect. How can I feel comfortable making such a sweeping generalization? Because if you carefully examine most of these surveys, the limitations are immediately apparent.
Let’s start with call pay. I was recently asked to review some survey data for call pay for physicians at a rural hospital. The survey data did not include any urban/rural distinction; it was presumably data from across the United States. By itself, that would lower the applicability of the data to my client, but a more significant flaw was the number of physicians surveyed: 12. Yes, this nationwide call survey had an N of a dozen physicians. The portion of the survey provided to the client yielded no more information. It is possible that the dozen physicians were all contained within a single group, although it is also possible the survey was done at the hospital level and at least represented a dozen hospitals. The excerpt didn’t provide sufficient detail to determine the answer, but I will say with confidence that no nationwide survey in the United States should be considered sufficient with an N of 12.
While most of the salary surveys I have seen have a greater number of respondents, that does not mean the surveys are representative of the entire marketplace. One of the more common surveys is conducted by the Medical Group Managers Association (MGMA). Tim Smith of Ankura Consulting includes the following observation from Meghan Wong at the MGMA in the BVR/AHLA Guide to Valuing Physician Compensation and Health Care Service Arrangements: Wong explains that “the data are not intended to be used as an academic data set for extrapolating to the U.S. population of physicians,” and are not a “one-to-one representation of the universe of medical practices that are in the country.” Despite that the survey explicitly disclaimed it should not be used to extrapolate to a population of physicians, that is exactly how these surveys are routinely used. Hospitals routinely use the surveys in a way the survey designers say is incorrect.
There are only a handful of national surveys of physician compensation. The disparate results between the two surveys provide compelling evidence of their imperfection as a tool to analyze nationwide physician compensation. For example, here is the 90th percentile conversion factor paid to cardiologists for a recent year as reported by two of the most common surveys:
- AMGA $102.06
- MGMA $86.47
Obviously, it is impossible for both of those figures to truly constitute the 90th percentile of the marketplace. I don’t know which is more accurate, but the fact that these vary by 18% is telling.
Looking within one of the surveys over time reveals the wild variability of the data. According to one of the surveys, here is the 90th percentile rvu production for interventional cardiologists during a series of years:
- 2009 $16,758
- 2010 $18,316
- 2011 $16,136
- 2012 $15,208
If that survey is correct, a physician working at the 90th percentile worked 20% fewer RVUs in 2012 than he or she did in 2010. That is simply not believable.
Despite these readily apparent flaws, there is a cultish obsession with salary surveys. Perhaps the draconian Stark penalties render this obsession understandable, but if you are entering any financial relationship with a hospital, you either need to take the time to find the flaws in the data or work with an expert who will. I would note most consultants and lawyers seem oblivious to this problem. It is another reason to be selective when you hire consultants and lawyers. If you are considering working with someone as you enter a relationship with a hospital, ask them about the use of survey data. If they do not raise the issues in this article, hire someone else. If you are looking for suggestions of consultants to work with, feel free to contact me at dglaser@fredlaw.com.
Reference: