Variable Chlamydia Screening Coverage Estimates Derived Through Healthcare Effectiveness Data and Information System (HEDIS) Procedures and Indirect Estimation

Background: The best method for estimating chlamydial screening coverage is unknown. 

Objectives: To compare chlamydia screening coverage estimation techniques.

Methods: We used HEDIS procedures to estimate screening coverage among women age 15-25 who utilized services in Group Health (GH), WA State, 2009, and estimated screening coverage for GH enrollees by assuming that non-utilizers did not CT test and that 61% were sexually active (National Survey of Family Growth [NSFG] estimate).  We also indirectly estimated screening coverage.  Indirect estimates used census data and NSFG sexual activity data to define the population size, and estimated the number of tests performed by dividing the number of reported CT cases by CT test positivity in 5 reporting laboratories (RL).  We derived a combined direct/indirect estimate using the same population size estimate and estimating the number of tests performed as follows:                                                                                                                                                                                                                                                                                          Total Tests=Test_RL+((Total Cases – Cases_RL)/Chlamydia positivity in all private labs)                                                                                                                                                                                                                                                                                          Where Test_RL is tests performed in RL and Cases_RL is cases identified in RL.  We adjusted indirect and direct/indirect estimates for multiple tests performed on individual women.

Results: HEDIS-estimated screening coverage was higher among GH-utilizers (43.6%) than among all enrollees (34.2%).  CT test positivity in RL varied from 4.5-6.3%.  Indirect screening coverage estimates ranged from 49.1-68.7% depending on the source of CT positivity data.  4916 (5.5%) of 83,513 tests performed in RL were positive; 44% of all CT cases were identified in RL.  The direct/indirect estimate of statewide screening coverage was 54.6%.

Conclusions: HEDIS methods may overestimate chlamydial screening coverage due to their focus on service utilizers.  Indirect methods of calculating screening coverage are sensitive to CT positivity estimates, and are limited by the uncertain representativeness of RL positivity data.

Implications for Programs, Policy, and Research: Current methods for estimating CT screening coverage provide imprecise estimates of unknown validity.  Collecting CT positivity data from large laboratories may provide an improved means for estimating screening coverage.