Bob Wachter, MD, created buzz in March 2016 when, at the SHM annual meeting in San Diego, he displayed a slide titled “What did we get wrong?” The slide contained the copy, “Hospitalist shifts run 7 a.m.-7 p.m.; 10 a.m.-10 p.m. 7on/7off” circled in bold red.
Over the last several years, thought leaders in the hospital medicine field have expressed concern that this one-size-fits-all schedule model is a threat to the well-being of many physicians and, by extension, the sustainability of their hospital medicine groups. Despite this, the 2016 State of Hospital Medicine Report reveals relatively little change in the way hospital medicine groups schedule their physicians.
Dr. Kimberly Eisenstock
Most groups (69.2%), report the duration of scheduled day shifts to be between 12 and 13.9 hours, similar to the 65.4% reported in the 2014 survey for this same metric. Likely, most of these shifts are the traditional 12-hour shift displayed on Dr. Wachter’s slide. Groups reporting shorter shifts tended to be either very large, with the number of bodies needed to develop flexible scheduling, or in academic settings where they could utilize house-staff coverage.
Night shifts echo this trend. There is an even greater number of groups utilizing the 12- to 13.9-hour shift length (79%), which has also varied less at just approximately 5% in either direction over the last two surveys. It is likely very hard to be creative with the shift length for your night physicians when the group is structured predominately around a 12-hour day position.
The 12-hour shift scheduled in long blocks is straightforward to employ for the scheduler, limits hand offs of care, and maximizes number of days off. So, why are Wachter et al. calling for change? Seven day stretches off may seem attractive when you are just starting out, but, as physicians mature, the very long day competes with family time that cannot be made up on weekday mornings when others are at school and work. Furthermore, the very long hours for 7 days straight lead to burn out and eventually retention issues as well. Some argue that this design promotes disengagement. It sets the expectation that, during “off” weeks, physicians might be unavailable for email responses, committee meetings, or participation in quality improvement initiatives, which disrupts integration into the larger hospital community and perhaps even our own career advancement.
Some groups are trying to address these concerns with innovative approaches to block scheduling. While the hallmark hospital medicine schedule of 7on/7off blocks remains the predominant model – 38.1% of all groups – this represents a drop of approximately 15%, compared with the prior survey. A new large contingent of groups entering the survey this year utilize a Monday-Friday model with rotating moonlighter/weekend coverage. This lifestyle and family-friendly model predominates in the Midwest. It is also found more in smaller groups, which may employ this model to keep the most system-knowledgeable worker around during high volume times, as well as to preserve the well-being and retention of their limited physician work force.
Of note, reconfiguring the 7on/7off model does not necessarily translate into more time off. The median number of shifts per year is also relatively stable at 182 which is the exact number of shifts per year in a strict 7on/7off schedule. This number does not vary by region of the country, group size, or teaching status. Some might argue that working 182 annual shifts is ideal, giving hospitalists a “vacation” every other week. However, this line of thought does not take into account the very long workdays, nor the 52 weekend days spent in the hospital – far more than most specialty peers who serve fewer weekend calls often with more limited in-house hours. In addition, one might argue that defining ourselves as available only during our 182 clinical “on” days is not in our own best interest, as it is the important nonclinical quality and committee service activities that are likely to lead to professional recognition and advancement.
Our hospital medicine group has deviated from this scheduling mainstay and requires only 160 shifts per year. We have set this number based on removal of the number of shifts equivalent to the vacation hours received by our medical group peers. The model poses a challenge in terms of matching our productivity up to benchmarks when talking to system leaders. This challenge pales in comparison to the increased buy-in from our physicians, as they feel equitable vacation time signifies respect from the medical group leadership.
In addition, our group has had success in being flexible around the number of days worked in a continuity stretch. We utilize everything from a 3-day block over holidays to a 7-day block. In general, we allow physicians to select their desired block length. The scheduler then works to accommodate that stretch as much as is feasible. The upfront work in this system is significant, but the downstream effect is decreased turnover costs. Even our own entrenched standard of 7on/7off schedules for house staff services (designed to protect continuity for the learner) have been the target of change. A pilot of alternating 4 and 5 day runs in a 4-week stretch has been implemented over the last few months. The number of days the residents are exposed to a given attending is the same in this model, but there is one additional switch day. The additional switch day puts the residents at risk of managing a change in care plan related to change in attending, but this was mitigated by paring attendings with very similar teaching and patient management styles. For our group, the extra administrative effort needed to work around the 7on/7off model has always paid off in terms of provider satisfaction and retention.
On the other hand, although I lead a large academic group, we have not yet developed flexibility around the shift length. Only one of the 29 roles our providers fill each 24-hour period is not a 12-hour shift. Over the years, I have tried to offer alternate models with shorter shifts to improve flow, reduce burn out, and increase family time. No matter how eloquent the reasoning, the response from the group was always the same: a resounding “no.” Most providers felt that they would wind up with a very similar work load and not actually leave the hospital earlier. Other reasons included not wanting to come in more days per month and concerns about increased handoffs/cross coverage.
There is some reason to think change may actually come. For one, burnout is high and may lead physicians to try a new model even with fear of the unknown. Our practice may be reconsidering this one-size-fits-all shift length in the very near future as an increasing percentage of candidates seeking to join our group express a strong interest in finding more accommodating hours.
Overall, I am hopeful that, in the coming years, my hospital medicine group, as well as many others, will heed the thoughts expressed by Dr. Wachter. Finding the flexibility to break out of these rigid scheduling models will be a first step in promoting both physician and system well being.