Stephen J.LurieMD, PhD, Senior EditorIndividualAuthorJody W.ZylkeMD, Contributing EditorIndividualAuthor
In Reply: Dr Hsu and colleagues correctly point out that differential coding practices theoretically could explain the observed differences between groups. However, there is little reason to believe that physicians caring for African Americans are consistently less likely to code follow-up visits after hospitalization, for example, because they face similar financial incentives within the Medicare fee-for-service system to do so. This same argument applies to a lesser extent to residents residing in poverty ZIP codes or health provider shortage areas. Moreover, to explain our results, the differential coding practices would have to apply uniformly over a wide variety of diagnostic tests and procedures. Nonetheless, we agree that comparison with data from a medical record review would be useful.
Asch SM, Sloss EM, Brook RH, Kravitz RL. Overuse of Administrative Data to Measure Underuse of Care—Reply. JAMA. 2001;285(6):735-737. doi:10-1001/pubs.JAMA-ISSN-0098-7484-285-6-jlt0214