• Physician workforce composition, spending, and quality

    Katherine Baicker and Amitabh Chandra explored the relationships among physician workforce composition, Medicare spending, and quality in a 2004 Health Affairs article (ungated). Broadly, they found that higher spending is associated with lower quality. More interestingly, however, they suggest why: too many specialists relative to general practitioners.

    This suggests that specialists are clustered in areas where costly intensive care crowds out high-quality care and that one mechanism for this is a lesser presence of general practitioners. Encouraging greater access to general practitioners, or involving specialists in the provision of effective care, could improve the overall quality of care received by elderly Americans.

    Their charts and tables tell the story. Below are four that support the quote above. First, let’s consider how state quality rank relates to physician workforce composition. And by quality rank I (they) mean the average of the ranks for the 24 quality measures developed by the Medicare Quality Improvement Organization. The more specialists per capita, the lower a state’s quality rank. (Note: for all charts, total physicians per capita is held constant.)

    But the more general practitioners per capita, the higher the state’s quality rank.

    What about spending? Well, specialists cost a lot more than general practitioners. Here are the charts, first specialists and then general practitioners:

    Maybe this particular quality rank is biased against specialists? The authors write,

    It is possible that although areas with more specialists do not provide higher quality care along these dimensions, they may be better at the treatment of more acute conditions. It is also possible that areas “specialize” in different types of care: Some areas specialize in primary care, while others may specialize in the delivery of technologically aggressive care for heart attacks. We do not find evidence of this here: States with more specialists have neither lower mortality rates from all causes nor reduced post-AMI mortality.

    Maybe the states vary so much in patient health that high spending areas need more specialists. Could be, but it likely doesn’t account for all the variation. There is even evidence that areas with greater need have fewer specialists. (See, for example, the work of David Goodman et al.) Moreover, there is no excuse for high spending areas not to deliver quality care. The quality metrics that underlie the ranks are process measures that correspond to evidence-based, effective care. Everybody should be receiving such care, and especially those with greater health risk. They should not be underprovided just because a region has more specialists.


    • What does it mean that “total physicians are held constant”? I looked at the original paper, but didn’t find a more detailed description. It seems to me that would be relevant, though: maybe the problem is less too many specialists than too many (or not enough) physicians overall??

    • The problem with this set of indicators is that they are not outcome based. They are tilted towards follow up testing and initiating therapy within a certain amount of time. We know that there are also at least a few studies which seem to indicate that when more is spent, patients do better. I think this implies that we need either more comprehensive studies, looking at both these indicators and outcome, or we need coned down studies looking at specific disease entities.


    • It’s a bit awkward, statistically, to graph ranks against some continuous variable. And not really good form to draw what might be considered a regression line to summarize the relationship. BUT, the relationship seems to hold up when you take a continuous quality score, like the Jencks state level quality indicators, or an updated Commonwealth quality Index and compare to primary care MD supply. I’d be happy to send the data and graphs.

      However, if you start to look at the relationship between costs and components of supply, you find that percentage of IMGs is a predictor of much higher costs and that the percentage of supply and per capita supply of surgeons is associated with lower costs. Again, happy to send the graphs and data. see: Ricketts TC, Belsky D. 2012 Medicare Costs and Surgeon Supply in Hospital Service Areas. Annals of Surgery, 255(3):474-7

      • I’m likely to look at papers, but not likely to play with data. It’s not that I don’t like data (I work with it every day), but that I won’t have time. With your permission, I will email you my email address. Is that OK?

    • Well, having pulled up the data just now I found a negative relationship between primary care docs per 100k (AMA Masterfile) and costs per beneficiary (CMS) for 2004–that is, higher costs where there are more primary care docs per 100k (by state). I wonder what the definitional differences are. I note that the Health Affairs doc data are labeled “general practitioners”. Hmmm. what’s happening?

      • Did you control for total physician supply and do it all per capita? That would be what Baicker and Chandra did. If you’re still getting opposite results, please share in a document. I’d love to see it.

    • Having done a pretty exhaustive search on this subject recently, what is striking (or maybe not) is the deceptive label of “primary care” vs the superior outcomes achieved with practice organization over nature of training. Nephrologists in a coordinated care model for example, will likely trump FP’s in a small <5 person group.

      Additionally, primary care is ineffective as a value enhancer–even when concentrated– when in the presence of concurrently clustered specialists. Cath jockeys are kryptonite.

      The best review I found on this subject::

      What I would like to see is a 2×2 table and plots: high and low concentration of primary care vs high and low concentration specialists


    • How exactly would one take the conclusions of this article and use them to improve quality and lower costs?

      Does anyone really believe that you could take a cohort of persons who have medical conditions that are typically treated by specialist physicians and instead have them treated by GP’s, and see no difference in the outcomes? I think a thought experiment will suffice to answer that question.

      Are the specialist physicians themselves responsible for poor quality care? Has anyone demonstrated that this is the case?

      If someone can only think in terms of statistical aggregates and regression coefficients, then I suppose it’s possible for the said person to believe that there’s a causal connection between having lots of medical specialists in a geographic area and poor quality care (such persons presumably avoid Mayo, Dana Farber, Fred Hutch, the Cleveland Clinic, etc like the plague. Must be glowing red “hot-zones” of poor quality care if this regression is illuminating a causal relationship).

      Other authors have noted that there’s a relationship between the composition of the physician workforce and the demographic characteristics of a population they serve (places that are not impoverished, non-urban tend to have a higher percentage of PCP’s per capita) and conclude that inasmuch as such relationships exist they’re best viewed as incidental correlations rather than causal factors that will automatically reduce the quality of care and increase the price for all patients under all circumstances.

      • @JayB
        Here is the best take on your question:

        I do believe that while a subspecialist might address the CHF more tightly for example, if left to the generalist in a head to head, findings might surprise you. Given attn to say DM, depression, etc., overall outcomes might be better, despite less state of the art CHF care. That is a head to head.

        If one is to assume team care, and the incremental costs and outcomes due to too many chefs, there is a danger of unintended consequences.. The right mix of services at the right price to produce best outcome will vary. We have not mastered it yet, and in 2012, 4 subspecialists vs one GP may produce counterintuitive results.

        However, on theoretical grounds, if we desegrated a patient into individual parts–a theoretical discussion–free from costs and realities of practice, sure, subspecialists win. That is only an academic exercise however.