Background: There is a need for meaningful and accurate ways of tracking preventive service delivery among different sectors of the US population.
Objectives: To compare methodologies of and clinical preventive service use estimates obtained from 2 data sets: the Health Plan Employer Data and Information Set (HEDIS 3.0) and the Behavioral Risk Factor Surveillance System (BRFSS).
Methods: HEDIS used a combination of mailed-survey, administrative, and medical-record data to measure preventive service use among commercial enrollees of 320 HMOs in 42 states during 1996. BRFSS data are from insured respondents (excluding those reporting Medicare or Medicaid coverage) to a random-digit-dialed telephone survey conducted in the same 42 states during 1996.
Results: The median state-specific mammography, Papanicolaou smear, and retinal examination rates reported by HEDIS were consistently and substantially lower than those reported by BRFSS. For mammography, the median HEDIS rate was 72.4%, compared with 81.1% for BRFSS. For Papanicolaou smear and retinal examinations, HEDIS rates were 72.7% and 40.8%, respectively, compared with 91.2% and 61.6% for BRFSS. The median state rates of advice to quit smoking reported by HEDIS were similar to those for BRFSS: 62.1% versus 62.2%, respectively. For each measure, the absolute difference between HEDIS and BRFSS rates varied substantially both within a state and between states.
Conclusions: It does not appear that the BRFSS and HEDIS data can be compared directly to accurately track progress toward national preventive health objectives. This study highlights some of the problems with comparing these data and possible means for addressing the discrepancies.