Stephen J.LurieMD, PhD, Senior EditorIndividualAuthor
In Reply: Among a national sample of Medicare
beneficiaries enrolled in managed care plans, we found significantly lower
quality of care for blacks than whites. With the exception of breast cancer
screening, these disparities in quality were not fully explained by other
individual factors that were available to us or by concentration of blacks
in plans that provided lower quality of care.
The Health Plan Employer Data and Information Set (HEDIS) data we used
represent an important advance over analyses based on "utilization," "claims,"
or "administrative" data. As Drs Barnhart and Wassertheil-Smoller suggest,
studies based only on claims or utilization data frequently lack important
clinical details that might explain observed racial disparities. We agree
that in studies of racial disparity it is important to account carefully for
clinical status, and we believe the HEDIS protocol for data collection does
this. For example, the quality measure of β-blocker use after myocardial
infarction excludes patients with clinical contraindications to β-blocker
use. While some HEDIS measures can be calculated using claims data alone,
most health plans also use specified protocols to review medical records for
evidence of such contraindications. Also, recent studies show unexplained
racial disparities in influenza vaccination,1
and renal transplantation4 despite controlling
for detailed clinical characteristics, as well as patients' beliefs and preferences
about health care.
Schneider EC, Epstein AM, Zaslavsky AM. Racial Disparities in Health Care—Reply. JAMA. 2002;287(22):2942-2943. doi:10.1001/jama.287.22.2942