OBJECTIVETo determine whether non-targeted testing strategy (screening all patients with blood sample withdrawn) could identify more patients with newly diagnosed HIV infection than symptom and risk behavior based targeted testing strategy or not.

METHODSControlled trial design was applied in this study.From July to November 2011, outpatient department of L and J county hospital in Guangxi province were selected to conduct the targeted strategy and non-targeted strategy respectively. The two counties had similar population, outpatient volume, previous testing rate and number of newly identified HIV cases.Outpatients older than 15 years were recruited as study subjects, with 62 106 person time in L hospital and 58 257 in J hospital. Data about visit number, persons receiving HIV testing and HIV positive cases were collected by outpatient department. Chi-square test was used to compare the percentage of newly identified HIV cases, HIV positive detection rate and proportion of cases in early AIDS phase between two strategies.

RESULTSDuring the study period, 9.69% (5627/58 057) of all outpatients in J hospital with non-targeted strategy and 1.38% (859/62 106) of all outpatients in L hospital with targeted strategy received HIV test. The average age of patients receiving HIV testing was 46.23 ± 16.81 and 40.75 ± 15.48 respectively, which was statistically different (t = 8.81, P < 0.05). The percentage of newly identified HIV cases was significantly greater in J hospital (0.03% (19/58 057)) than that in L hospital (0.02% (10/62106)) (P < 0.05) while the HIV positive detection rate was lower in J hospital (0.34% (19/5627)) than that in L hospital (1.16% (10/859)) (χ(2) = 9.66, P < 0.05).

CONCLUSIONIn a concentrated epidemic, a hospital based non-targeted strategy could detect more unidentified HIV cases than targeted strategy.