Geographic variation in per capita health spending has been a prominent research topic for several years, primarily because some of the researchers suggest that the variation is unwarranted and that reducing expenses in the high-cost regions to that of lower-cost areas could help control total spending. The New England Journal of Medicine carries a study on spending variation in Medicare’s Part D benefit, which covers prescription drugs. (NEJM Article) The researchers looked at any correlation between variation in the medical benefits and the drug benefits.
The researchers attempted to remove all sources of variation other than geography, including the health status of the beneficiary. What they found is that there is variation in drug spending, but less so than for medical spending. The variation is due to both the number of prescriptions per beneficiary and the specific drugs prescribed, brand names versus generics, for example. It also appears that drug and medical spending variation are weakly correlated at best. This may suggest that use of drugs is a substitute for other medical care, so that higher drug spending may mean lower medical expenses. On the other hand it also suggests that areas with high medical spending also often have high drug spending.
This study is another useful addition to the growing understanding of variation in spending. The causal explanation for drug or medical spending variation, however, is still not clear. Is it reflective of physician practice preferences, potentially driven by revenue generation incentives, or is it due to patient preferences in certain areas. Answering this key question will allow formulating policies to reduce any inappropriate variation in health care delivery.