Thinking about productivity: Survey data 2017

What are ‘reasonable expectations’ for compensation and productivity?



The 2017 MGMA survey data on compensation and productivity were released last June. While the numbers aren’t surprising, reviewing them always gets me thinking about factors that influence reasonable expectations for compensation and productivity in any individual hospitalist group.

The data were collected in early 2017, reflecting work done in 2016, and show a national median hospitalist compensation for internal medicine physicians of $284,000, up from $278,500 the year before. Since MGMA added a hospitalist category to the survey, compensation has been growing significantly faster than inflation, even though productivity has been essentially flat. I’ve always thought that the high demand for hospitalists, which isn’t letting up much, in the face of a limited supply is probably the most significant force causing hospitalist compensation to rise faster than in most other specialties.

The survey shows a median of 2,114 billed encounters and 4,159 wRVUs (work relative value units) generated per internal medicine hospitalist annually (family medicine hospitalists are reported separately). These numbers have been pretty stable for many years.

Whether it is reasonable to expect hospitalists in your group to produce at this level is a question that can unspool into a lengthy conversation. Below are several assertions I regularly hear others make about productivity, and following each is my commentary.

Dr. John Nelson, cofounder and past president of SHM, principal in Nelson Flores Hospital Medicine Consultants, and codirector for SHM’s practice management courses.
Dr. John Nelson

“Surveys show only what is most typical, not what is optimal. Our field suffers from concerning levels of burnout, essentially proving that median levels of productivity shown in surveys is too high.”

I share this concern, but this is a complicated issue. You’ll have to make up your own mind regarding how significantly workload influences hospitalist burnout. But the modest amount of published research on this topic suggests that workload itself isn’t as strongly associated with burnout as you might think. I’m certain workload does play a role, but other factors such as “occupational solidarity” seem to matter more. Lowering workload in some settings might be appropriate, but without other interventions may not influence work-related stress and burnout as much as might be hoped.

“Surveys don’t capture unbillable activities (‘unbillable wRVUs’), so are a poor frame of reference when thinking about productivity expectations in our own group.”

It’s true that hospitalists do a lot of work that isn’t captured in wRVUs. My work with many groups around the country suggests the amount and difficulty of this unbillable work is reasonably similar across most groups. We all spend time with handoffs, managing paperwork such as charge capture and completing forms, responding to a rapid response call that doesn’t lead to a billable charge, etc. The average amount of this sort of work is built into the survey. Clearly some groups are outliers with meaningfully more unbillable work than elsewhere, but that can be a difficult or impossible thing to prove.

“My hospital has unique barriers to efficiency/productivity, so it’s more difficult to achieve levels of productivity shown in surveys.”

This is another way of expressing the previous issue. To support this assertion hospitalists will mention that it is tougher to be productive at their hospital because they’re a referral center with unusually sick and complicated patients; they teach trainees in addition to clinical care; and/or their patients and families are unusually demanding, so they take much more time than at other places.

Yet for each of these issues I also hear the reverse argument regularly. Hospitalists point out that because they’re a small hospital (not a referral center) they lack the support of other specialties so must manage all aspects of care themselves; they don’t have residents to help do some of the work; and their patients are unsophisticated and lack social support. For these reasons, the argument goes, they shouldn’t be expected to achieve levels of productivity shown in surveys.

I have worked with hospitalist groups that I am convinced do face unusual barriers to efficiency that are meaningful enough that unless the barriers can be addressed, I think productivity expectations should be lower than survey benchmarks. For example, in most academic medical centers and a very small number of nonacademic hospitals, only the attending physician writes orders; consulting doctors don’t. This means that the attending hospitalist must check a patient’s chart repeatedly through the day just to see if the consultant proposed even small things like ordering a routine lab test, advancing the diet, etc., that the hospitalist must order.

A separate daytime admitter shift is a modest barrier to efficiency that is so common it is clearly factored into survey results. Most hospitalist groups with more than about five doctors working daily have one doctor (or more than one in large groups) manage admissions while the rest round and are protected from admissions. While this may have a number of benefits, overall hospitalist efficiency isn’t one of them. It means that all patients, not just those admitted at night, will have a handoff from the admitting provider to a new attending for the first rounding visit. This new attending will spend additional time becoming familiar with the patient – time that wouldn’t be necessary had that doctor performed the admission visit herself.

   Comments ()

Next Article: