The Fake Math of Primary Care

It's a design failure built on outdated numbers.

Primary Care Is a Design Failure Based on Outdated Numbers

I want to start with an assumption: Primary care is the crux of good medicine. A primary care doctor knows you, your medical history, and your values. A primary care doctor is your first point of contact for any changes in your health. A primary care doctor looks for the big picture. I built my own primary care-based cancer survivorship practice because of a firm belief that primary care can be all of the above. My views on its importance have only cemented as I’ve been in practice for the last few years.

But as I’ve written extensively before — here and here, for example — primary care has strayed so far from that ideal it’s almost unrecognizable. Modern primary care is a lose-lose situation where doctors are working harder than ever — and doing the wrong kind of work — all while patients can access us less.

To understand why, we have to dig into the numbers. Primary care is built on a host of numerical assumptions. How many minutes does a patient need per appointment? What percent of the job happens in between appointments versus during them? How many patients should a single doctor be responsible for? The problem is that many of these assumptions do not remotely match modern realities. Primary care is a design failure built on distorted, outdated math.

This post goes into one source of that math: How many hours per week do primary care doctors need to do the job? And how does that square with the hours they’re actually spending?

---

How Many Hours Per Week Do Primary Care Doctors Need?

Last month, Dr. Lisa Rotenstein and colleagues published a paper in Annals of Internal Medicine showing that full-time primary care doctors spend a median of 61.8 hours per week caring for their patients. They also found that part-time primary care doctors spend more time on each patient than full-time physicians: 2 hours per patient per year, compared to 1.7 hours. Of note, the median total practice size of full-time doctors in this study was 1,668 patients — significantly below the typical 2,000 to 2,500 range nationwide.

Where does all that time go? Using the estimate of 5.5 hours of electronic chart work for every 8 hours of patient-facing care from a separate JAMA paper by Dr. A. Jay Holmgren and colleagues, that comes out to 22 hours per week in electronic charts on top of 32 hours of face-to-face appointments: documenting patient encounters, clicking orders, reviewing test results, refilling prescriptions, messaging other team members, and answering patient messages. The remaining ~8 hours involve all the non-electronic administrative work that a primary care doctor must do to move care forward, such as peer-to-peer phone calls with insurance companies and filling and signing paper forms.

Remarkably, working 62 hours per week leaves work undone. Another paper by Dr. Justin Porter and colleagues in the Journal of General Internal Medicine mapped out that if primary care doctors did everything expected of them for a standard 2,500-patient practice, they’d be working 26.7 hours per day. That totals 133.5 hours in a five-day workweek. If we use the 1.7 hours per patient per year figure from Rotenstein’s paper and apply it to 2,500 patients, primary care doctors would work about 92 hours per week (over 46 weeks in a year). That leaves a difference of about 41 hours each week — work that should be done but simply cannot.

In short, even when working over 90 hours per week, primary care doctors can only accomplish less than 70% of the work asked of them. Using the part-time data of 2 hours per patient per year, they’d be able to meet about 81% of the job description.

---

What This Means for Primary Care

These findings back with data what primary care doctors already know: Going part-time is the only way to meet our patients’ needs. Full-time primary care is neither doable nor sustainable in its current form. When I began as a full-time attending, while I didn’t track how many hours I worked per week, it was far higher than 40; work always spilled into nights and weekends, and I still was never caught up. This also jibes with nationwide data showing that primary care doctors are going part-time more now than in previous years. The Wall Street Journal has cited older doctors bemoaning this generation’s desire for work-life balance as the reason, but the data don’t fully support that. Primary care doctors aren’t necessarily going part-time to spend time with their families or pursue other hobbies, although these are obviously important. They’re going part-time to be able to do the job.

This calls for a radical redistribution of labor. In the Porter paper, with team support from nurses, nurse practitioners, and others, the 26.7-hour day decreased to 9.3 hours. The other option is reducing total practice sizes, as these numbers were developed when there was less chronic illness, fewer administrative tasks, and no electronic portals allowing constant asynchronous communication. I don’t have the perfect number for how many patients a primary care doctor should care for today. But 2,500 or even 1,668 isn’t it.

---

Coming soon: Dr. Rotenstein joins me on the Hard Medicine podcast (airing January 2026) for a deep dive into the question: Can primary care be saved?