More PCPs Pays Off -- So Pay Them More?

F. Perry Wilson, MD, MSCE


February 20, 2019

Welcome to Impact Factor, your weekly dose of commentary on a breaking medical study. I'm Perry Wilson.

I'm going to open this week with an opinion: Primary care doctors are the best doctors. There, I said it. Where would we all be without these guardians of the medical universe, these polymath experts on the front line fighting against disease, both chronic and acute?

Well, dead, it turns out, according to this study appearing in JAMA Internal Medicine.[1]


Lately, talented residents have shied away from general practice in favor of the more lucrative procedural specialties. It's hard to blame them; take a look at the results of Medscape's salary survey in 2018.


But should we be working harder to increase the pool of primary care providers?

The main finding of this study was that the more primary care providers there were within a county, the longer people in that county lived. For every 10 additional primary care physicians (PCPs) per 100,000 population, the median life expectancy increased by 52 days. Is that a lot?


Well, 10 additional specialty physicians only bought you 19 days of life expectancy. In fact, that gain of 52 days for 10 additional primary care providers was on a par with reducing the rate of smokers from 21% to 18% and greater than the effect of reducing obesity rates from 31% to 28.5%. In other words, if you were, say, a government trying to improve public health, increasing primary care docs may not be the worst place to spend your money.

OK, I know what you're thinking. Maybe healthier, wealthier counties attract more primary care docs. The researchers addressed this confounding in multiple ways. First, they did a fairly extensive multivariable adjustment, accounting for a bunch of county-level factors including median home price, number of hospital beds, and specialist physician density, as you can see here.


Next, they did an instrumental variable analysis. The trick to this type of analysis is finding a variable associated with your exposure—in this case, primary care provider density—but not your outcome (life expectancy).


If you can find such a variable—and that's a big if—you're in great shape because you can use it to analyze the study almost like a randomized trial. In this case, they used the Public Service Loan Forgiveness program as the instrument, arguing that the institution of the program would drive people into primary care, but not directly impact lifespan over all.


This analysis redemonstrated the primary results. So far, so good.

The authors even included a negative control, often sorely missing from studies of this type. They asked whether PCP density was associated with lower deaths from interpersonal violence. Since there isn't such a mechanism for PCPs to decrease deaths due to, say, murder, if you saw a relationship it would suggest that some third factor—a confounder—was driving all these results. But there was no relationship between PCP density and deaths due to interpersonal violence.

Taken together, this is a fairly compelling story telling us something we probably expected: Investments in primary care pay substantial dividends. Perhaps it's time to start paying these docs what they deserve.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: