September 18th, 2013
Disparities in Rate of Inappropriate Use of PCI
CardioExchange’s editor-in-chief Harlan Krumholz asks Paul Chan some questions about his new study in JACC that found a surprising pattern of both overuse and underuse of PCI among different populations in the U.S. For a summary of the findings, see CardioExchange’s news coverage of the paper.
Harlan Krumholz: The paper would be consistent with a hypothesis that patients for whom physicians are paid more are more likely to undergo inappropriate procedures? Is that true?
Paul Chan: Our findings by insurance status would seem to suggest this. We found a gradation by insurance status in the likelihood of undergoing an elective PCI categorized as inappropriate (i.e., with a low likelihood of clinical benefit). Whereas publicly insured patients were only 15% to 20% less likely to undergo an inappropriate PCI compared with privately insured patients, uninsured patients were half as likely. It is not clear, though, whether this was mediated at the physician level, or, more likely, at the hospitals at which predominantly uninsured and publicly insured patients received care.
Harlan Krumholz: Do you think that the lower rate of procedures in non-white populations might be due to overuse in white populations?
Paul Chan: There are 2 types of potential disparities in care. I do not think that a racial difference in a “hard” clinical end point would be disputable. For instance, we have previously found that non-whites have a substantially lower rate of survival compared with whites for cardiac arrests within hospitals, and I doubt anyone would dispute that this difference is unacceptable. Where there is some uncertainty is when there is a racial difference in a treatment or an intervention. There are numerous studies which have documented underuse in a treatment in non-white populations. What our study intimates is that some of the racial differences in treatment may also be due to potential overuse. Therefore, efforts to end racial disparities may be more complex– eradicating underuse among vulnerable populations in circumstances where there is clear clinical benefit of treatment for patients, but also addressing potential overuse which may contribute to the measured racial differences.
Harlan Krumholz: Why would white men be more likely to have inappropriate procedures?
Paul Chan: We found this to be a head-scratcher, too. Unlike differences by race and insurance status, I would imagine the representation of women across Cath PCI sites was pretty homogeneous, so this difference would be unlikely to be mediated by the hospital at which men and women received care. If anything, since women are more likely to present with atypical symptoms which could be coded as having no or minimal symptoms, I would not have been surprised were women found to have higher rates of PCIs categorized as inappropriate. It could be that women’s propensity to have more common presentations of diffuse coronary artery disease may have led to less aggressive treatment in clinical settings where symptoms, angiographic findings, and stress test results were not high-risk (i.e., less frequent PCIs for inappropriate indications).
Harlan Krumholz: What are the implications of these findings for clinicians and policymakers?
Paul Chan: We have come a long way in ensuring that care is delivered equitably and thoughtfully in the U.S., and there is no doubt that underuse in certain populations remains a persistent and huge problem. For policymakers, as mentioned in my response to the second question, it highlights the importance of thinking about differences in treatment in a more complex way– as due to underuse and also potential overuse. Therefore, the goal may be to narrow the gap in vulnerable populations in instances where treatment has clearly established benefit rather than assuming that the measured difference is entirely due to a disparity in care.