Physician Practice Administrative Costs

Health Affairs has an online article (HA article) reporting on a survey of physicians regarding the time they spend on administrative interactions with health plans.  The researchers also extrapolated a cost associated with these activities.  The survey relied solely on self-reported data from physicians in smaller practices and nurses and administrative staff in larger practices.  The conclusions were that physicians spent on average 3 hours a week interacting with health plans; nurses about 19.1 hours and clerical staff about 36 hours.  The researchers inferred that over $68,000 per physician per year was spent on these activities.

There is good reason to be skeptical of the survey’s results.  Although the researchers went to some lengths to ensure good results, there is no getting around the fundamental weakness of self-reported results on a topic the respondents likely feel very strongly about.  There was bound to be some selection bias in which physicians even chose to respond and one has to suspect that overestimation resulted.  Generally speaking, individuals tend to be very poor estimators of how much time they spent on any activity.  Another warning sign is the very wide divergence between the mean and median.  Although the average was 3 hours per week, the median was less than two.  This suggests that either a few physicians account for a lot of the total health plan interaction or those physicians who most dislike working with health plans put down very high time spent on these interactions.  Similarly, there is not a detailed explanation of how the costs were arrived at; what the specific formula was, particularly in regard to actual physician costs.  Confusing income with costs seems possible.

Also of note is that the largest single interaction with health plans was around formularies.  Since most formularies are online or easily available, physicians could avoid much of the interaction either with pharmacies calling to tell them they wrote a prescription which was off-formulary or with health plans or PBMs to ask for exceptions to formulary guidelines.  Some of this burden is obviously self-inflicted and some is undoubtedly due to drug company marketing that has physicians insisting on prescribing a certain drug.

It is also unfortunate, and somewhat hard to understand, that the researchers failed to ask about interactions with Medicare and Medicaid, because at least anecdotally, both programs are even more frustrating than private plans to deal with.  That is relevant information for those who are considering whether there should be a Medicare-like public option for health insurance coverage.  Certainly those public programs impose much more burdensome and intrusive regulation on providers than do private plans and one would expect correspondingly higher commitment of time and cost to result.

Certainly those who believe that private insurance imposes high administrative costs on both the health plan and provider side, and that those expenses are largely wasted, will like the results of this study, which makes its shaky foundation dangerous.  A much more rigorous methodology should be applied to the issue, and a better comparison with costs imposed by public programs, before any conclusions are drawn which might be used in reform discussions.