1252Agreement between self-reported chlamydia infection and chlamydia infection ascertained using testing and medication administrative data

Background Chlamydia trachomatis is the most frequently notified sexually transmitted infection (STI) in several countries, including Australia. We compared self-report of infection by women with the use of linked testing and medication administrative data to identify chlamydia infection. Methods We used data from 8474 women aged 22-27 years from the Australian Longitudinal Study on Women’s Health in 2017. We compared self-report of a chlamydia infection in the previous 12 months with chlamydia infection estimated by having a prescription filled for azithromycin (the most commonly recommended treatment for chlamydia) within 28 days of having a chlamydia test through data linkage with the Medicare and Pharmaceutical Benefits database over the same 12-month period. Results Chlamydia infection was identified by self-report by 238 women. In contrast, 151 women had a prescription filled for azithromycin within 28 days of having a chlamydia test. Of the 151 women identified with chlamydia infection through data linkage, 101 also self-reported chlamydia infection (positive percent agreement = 67%). Of the 8323 women who did not have a chlamydia infection identified through data linkage, 8186 women also did not self-report a chlamydia infection (negative percent agreement = 98%). Conclusions Our study is consistent with previous studies indicating only moderate agreement between self-report of chlamydia infection and infection identified through data linkage. Further triangulation with medical records is needed to ascertain which of the two methods is more accurate. Key messages It remains unclear whether linked administrative data is preferable to self-report in estimating chlamydia infection in population studies.