OBJECTIVETo examine the proportion and reasons of drop-out from antiretroviral therapy (ART) among 8 367 adult HIV-infected individuals in Dehong prefecture, Yunnan province.

METHODSAll adult HIV-infected patients receiving ART before September 30 of 2014 were examined for the situation of drop-out from ART.

RESULTSThe proportion of drop-out from ART among adult HIV-infected patients in Dehong prefecture was 14.4% (1 202/8 367). Results from the univariate logistic regression analyses indicated that drop-out from ART was significantly correlated with factors as: living area, gender, age, marital status, HIV transmission route, baseline CD4⁺ T cell counts and initial treatment regimen of the patients. After adjusted for potential confounding variables by multiple logistic regression model, drop-out from ART was significantly correlated with residential area, marital status, HIV transmission route, baseline CD4⁺ T cell count and initial treatment regimen of the patients. HIV-infected patients who were living in Mangshi city, Lianghe county or Yingjiang County, being married or living with partner, HIV infection through sexual contact, with baseline CD4⁺ T cell counts ≤ 200 cells/mm³, and ART included in the initial treatment regimen etc., were less likely to drop out from ART. The proportion of drop out from ART was significantly decreasing along with the increasing time of ART. Data from specific investigation revealed that among the 1 202 patients who dropped out from ART, 704 (58.6%) were lost to follow-up, 303 (25.2%) did not adhere to treatment, 74 (6.2%) moved out the region, 64 (5.3%) were Burmese that had returned to Burma, 29 (2.4%) stopped the treatment according to doctors' advice, 18 (1.5%) were incarcerated and 10 (0.8%) were under other reasons. Reasons for the drop-out varied, according to the situation of patients.

CONCLUSIONThe proportion of drop-out from ART varied significantly according to the characteristics of HIV-infected patients in Dehong prefecture that underscoring the needs for tailored responses to reduce drop-out of ART. Focus should be targeted on reducing the loss to follow-up and improving the treatment adherence.