Background In the US military, chlamydia is the mostly commonly diagnosed bacterial sexually transmitted infection and the rates of pelvic inflammatory disease (PID) have remained high since the early 2000s.

Methods The relationship between the number of chlamydia diagnoses and hazard of PID was investigated in a retrospective cohort analysis among US Army women from 2006 to 2012. Cox regression model was used to estimate hazard ratios for associations between the number of repeat chlamydia diagnoses and PID.

Results The study population comprised 33,176 women with chlamydia diagnosis. Of these, 25,098 (75%) were diagnosed only once (“nonrepeaters”). By comparison, 6282 (19%), 1435 (4%), and 361 (1%) women had one, two and three repeat chlamydia diagnoses, respectively. Among these 4 groups, 1111, 325, 72, and 25 PID diagnoses were noted. According to the Cox regression analysis, for every additional diagnosis of chlamydia, the hazard of PID increased by 28% (95% confidence interval, 19%–38%) compared with women with a single diagnosis or nonrepeaters. Moreover, the corresponding adjusted hazard ratio of 1.28, 1.35, and 1.97 represented a significantly greater risk for PID among the three “repeater” groups compared with nonrepeaters.

Conclusions We found an increased hazard of PID among US Army women with repeat chlamydia diagnoses and the characterization of a dose-response relationship. These findings reinforce the notion that early diagnosis and treatment of chlamydia is necessary to avoid subsequent PID and associated morbidity.