OBJECTIVETo study the loss to follow-up (LTF) rate, HIV emerging incidence and influencing factors in the longitudinal study of female sex workers (FSWs) in Kaiyuan, Yunnan, and their influence on HIV emerging incidence estimate.

METHODSThe open cohort study on FSWs was launched in March 2006 and ended in June 2013 in Kaiyuan, Yunnan. Investigations were made every six months and lasted for 7 years. 661 FSWs found as HIV negative in the baseline study in March 2006 were chosen to study their LTF in the 7 year follow-up investigation. The Cox regression model was used to explore risk factors for HIV emerging infection and those for LTF. In June 2013, a survey was also conducted to explore the detailed reasons for loss to follow-up by contacting FSWs themselves.

RESULTSDuring 1 238.5 person-years of follow-up among 661 HIV negative FSWs, the HIV incidence rate was 1.29 (95% CI: 0.74-2.10)/100 person-year, and the LFT incidence rate was 48.68 (95% CI: 44.88-52.73) /100 person-year. The multivariate analysis showed drug abuse as an independent risk factor for FSWs' infection of HIV (adjusted risk ratio = 4.15, 95% CI: 1.43-12.02); FSWs over 25 years old (adjusted risk ratio = 0.68, 95% CI: 0.57-0.81), and drug abuse (adjusted risk ratio = 0.52, 95% CI: 0.35-0.79) were found with lower LFT rate to remain in the cohort.

CONCLUSIONHigh LFT rate was found in FSW cohort study in Kaiyuan, Yunnan, while the HIV infection risk exposure of the LFT group was lower than the groups of HIV follow-up. HIV prevalence of FSWs in the city might be overestimated. Causes of LFT of FSWs group required further study in the future, and the cohort follow-up retention strategy for FSWs needs to be developed.