“Forty percent of physicians reported that their typical inpatient census exceeded safe levels at least monthly.”1
This quote is taken from an article by Henry Michtalik and colleagues that appeared at the end of January this year in JAMA Internal Medicine. In 2010 the authors conducted an on-line survey asking hospitalists their perceptions of their workload. Respondents indicated that with concerning frequency a high workload prevented them from adequately discussing with patients treatment options or answering questions, delay admitting or discharging patients until the next day or shift, or in some other way risk patient safety or the overall quality of their work.
This alarming finding matches my anecdotal experience working with many different hospitalist groups around the country. I think few hospitalists were surprised by the survey’s findings. Excess hospitalist workloads are indeed a problem in some settings, and those who bear them are typically not shy about speaking out.
The demand for hospitalists has exceeded the supply of doctors available to do the work throughout the history of the field. Under the weight of stunningly rapid growth in referral volume, from about 1995 to 2002, it was reasonably common for the original doctors in a hospitalist practice to become overwhelmed and leave for other work after a year or two, sometimes resulting in the collapse of the practice. Most practices are no longer in such a rapid-growth phase, but for many of them, staffing has not yet caught up with workload. The result can be chronic excess work, and even if daily patient volume is not seen as being unsafe, the number of days or shifts worked might be excessive and lead to fatigue and poor performance.
Other Workload Data
The respondents to the Michtalik survey reported that regardless of any assistance, “they could safely see 15 patients per shift, if their effort was 100% clinical.” What we don’t know is how long their shifts were, whether they included things like ICU coverage, and how many shifts they work consecutively or in a year.
SHM’s 2012 State of Hospital Medicine report, which is based on 2011 data, provides additional context. It shows that hospitalists serving adult patients report a median 2,092 billed encounters annually (mean 2,245, standard deviation 1,161). They spread this work over a median 185 shifts (“work periods”) annually (mean 192). While there are lots of methodological problems in manipulating those numbers further, 2,092 encounters divided by 185 shifts yields 11.3 encounters per shift. These numbers exclude academicians who typically spend significant time in activities other than direct patient care, and I’m intentionally ignoring such issues as the night-shift doctor, who typically has low productivity, bringing down the average per full-time doctor in a practice.
The numbers from both surveys are sort of fuzzy because they aren’t audited or verified, but the 2012 State of Hospital Medicine data suggest that typical workloads aren’t too high in most practices, yet 40% of respondents in the Michtalik survey said they were high enough—unreasonably high—to risk quality and safety at least once a month.
One way to reconcile these findings is to take into account the standard deviation in daily volume in a single practice of about 30% to 40% on above or below the mean. If a hospitalist averages 14 encounters each day shift, then he should expect that the daily number might vary between about eight and 20. The Michtalik survey responses were likely reflecting the shifts on the high end.