Estimating a Sustainable Workload in Primary Care: Implications on the Physician Shortage

Many primary care physicians trudge into clinic every morning, not yet recovered from the day before, with a mounting sense of dread. This is compounded as we work through lunch and write notes at the end of the day in our pajamas. We are called on to meet an incessant need, and when we can't, the guilt rushes in. This is not a moral failing of the individual; it is the system failing us. A majority of physicians name workload and powerlessness as the major contributors to burnout.¹ Research from the American Medical Association shows that three of the “big four” burnout factors are related to demanding schedules (see Table 1).

As the U.S. population ages and the health care system moves to quality-based care, which takes longer to provide, primary care physicians will have additional demands on our time. Within this context, how can we set more reasonable boundaries for ourselves without exacerbating the primary care shortage?

The goal of this editorial is threefold:

1. Examine what many of us have come to accept as a “normal” primary care workload and show why it is unsustainable,
2. Estimate a more reasonable and sustainable primary care workload using a business mathematical model,
3. Explore how this stark difference is contributing to the primary care physician shortage by hampering new physician recruitment efforts and pushing current physicians into early retirement and burnout.

Clinic A is an established outpatient practice operating with an eight-hour workday and 15-minute visits for office E/M services, which make up the majority of the workday; other services, such as annual preventive visits, are allotted more time. However, based on conservative estimates using Medicare billing data, we can reasonably deduce that the weighted average total time required for an office E/M visit — including time spent on non-face-to-face patient care activities — is 26 minutes. (For calculations, see the Appendix.) This means that, on average, each E/M patient encounter in Clinic A takes 11 minutes (or 73%) more than its allotted time, and we can assume that other visit types similarly require more work than their scheduled time. Essentially, this stretches a 4.5-day workweek with 36 patient-facing hours into 62 actual hours of work per week.

Many family physicians may find themselves nodding and thinking, “Yep, that's about right.” However, we can further examine the traditional workload by using the “M/M/1” line queuing model, which is common in the service industry.² The model uses two inputs: 1) appointment length, or how often patients arrive, which Clinic A has defined as 15-minute intervals for E/M services, and 2) service time, which we previously calculated at 26 minutes. The model assumes the system will eventually reach a steady state; however, this cannot happen if service time is longer than appointment length. Under these conditions, the model reads as “error” because the physician's time is utilized at 173%, well above the 100% possible during clinic hours. Thus, the remaining work spills into the physician's evenings and weekends. If we tried to do that extra work during the workday, patient wait times would skyrocket. (See the Appendix for model calculations.)