Article Text


Epidemiology poster session 4: Methodological aspects
P1-S4.09 Validation of the HEDIS measure of sexual activity in adolescents
  1. S Wiehe,
  2. J D Fortenberry,
  3. M Rosenman,
  4. W Tu
  1. School of Medicine, Indiana University, Indianapolis, USA


Background The proportion of sexually active young women who are screened annually for chlamydia is a Health Plan Employer Data and Information Set (HEDIS) performance measure. The accuracy of this measure, however, depends critically on how the denominator is calculated (ie, how many sexually active young women for whom screening is indicated). Few studies have assessed whether the measure is valid. Validating is difficult because sexual activity may not be routinely captured in the chart or other administrative data sources. Our objective was to determine the sensitivity and specificity of the HEDIS measure of sexual activity, by comparing it to a self-reported measure, and assess whether these vary by age, age of first sex, and prior sexually transmitted infection (STI) or pregnancy.

Methods Participants were 14−17-year old, non-pregnant, English-speaking females from three low-income urban clinics enrolled in a longitudinal study of STI epidemiology among young women (N=387). Self-reported age of sexual debut was collected during a baseline and quarterly face-to-face interviews. Electronic medical record (EMR) data from the participants' clinics were linked to study data in order to compare the calculated number of sexually active participants using EMR data (as defined by HEDIS) to self-reported sexual debut using study data. We calculated the sensitivity and specificity of HEDIS criteria of sexual activity, using self-reported sexual activity as the gold standard, accounting for repeated visits per individual. We also calculated the sensitivity and specificity stratified by age, age of first sex, prior STI, and prior pregnancy.

Results Overall, the sensitivity and specificity of the HEDIS criteria were 91.4% and 46.0%, respectively. These measures varied significantly by prior STI and age of presentation, and less so by prior pregnancy and age of sexual debut (Abstract P1-S4.09 table 1).

Abstract P1-S4.09 Table 1

Sensitivity and specificity of HEDIS criteria of sexual activity using administrative data (gold standard: self-reported sexual debut)

Conclusions These data indicate that the measure has good overall sensitivity but that sensitivity and specificity vary significantly based on age and sexual history. The sensitivity and specificity of the HEDIS denominator indicating sexual activity directly influences the accuracy of reported chlamydia screening rates. Thus, depending on the population, this indicator may be an inaccurate indication of sexual activity and poor measure of healthcare performance.

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.