Introduction Mycoplasma genitalium is increasingly recognised as an important cause of urethritis, cervicitis, PID, infertility, and increased HIV risk. A better understanding of its natural history is crucial to informing control policy. Two cohort studies (students in London, UK, and sex workers in Uganda) suggest very different clearance rates. We aimed to gain insight into the reasons and to obtain improved estimates by making maximal use of the data from those studies.

Methods To estimate incidence and clearance rates, we developed a model for time-to-event analysis incorporating the processes of individuals becoming infected after enrolment into the study, clearing infection, becoming reinfected after clearance, and fitted it to data from the two cohort studies. As the studies collected limited data on sexual partnership dynamics, we tested the sensitivity of the model to different assumptions that were consistent with the available information. Additionally, we modelled study-design differences, including sample handling conditions affecting testing sensitivity.

Results In the London students, the estimated clearance rate was 0.80p.a. (mean duration 15 months), with incidence 1.31% p.a. and 3.93% p.a. (in low- and high-risk groups, respectively). Without adjusting for study design, corresponding estimates from the Ugandan data were 3.35p.a. (mean duration 3.6 months) and 44% p.a. Clearance-rate differences could be explained by differences in testing sensitivity, with ‘true’ rates being similar, and adjusted incidence in the Uganda study being 21% p.a.

Conclusion Analysis needs to account for study design, and we recommend cohort studies collect more information on partnership dynamics to inform more-accurate estimates of natural-history parameters. The cohorts’ clearance rates were probably similar, with the apparent difference due mostly to differences in sample handling in the studies, and perhaps partly due to the sex workers having more-frequent antibiotic treatment (for other infections), and in the London students some reinfection in stable partnerships causing some of the apparently-persistent infection.

Disclosure of interest statement TS and PJW thank the UK National Institute for Health Research Health Protection Research Unit (NIHR HPRU) in Modelling Methodology at Imperial College London in partnership with Public Health England (PHE) for funding (grant HPRU-2012–10080). PJW also thanks the UK Medical Research Council for Centre funding (grant MR/K010174/1). The views expressed are those of the authors and not necessarily those of the NHS, the NIHR, the Department of Health, or Public Health England. We have no conflicts of interest.