Study participants
Participants were recruited through a multi-stage cluster sampling strategy. The county with the most severe HIV epidemic was selected from all 17 counties in Liangshan Prefecture. According to the basic information system for Acquired Immune Deficiency Syndrome (AIDS) prevention and treatment, all PLWH in this county were included in our study. We have tried to identify as many PLWH as possible through all three possible channels: (1) Provider-Initiated HIV Testing & Counseling (PITC) service for HIV, (2) Voluntary Counseling and Testing (VCT) service through fixed VCT sites and non-government organizations (NGOs), (3) HIV sentinel surveillance (HSS) for unmarried youths, pregnant women, community population, and drug users. The inclusion criteria for PLWH were: (1) receiving confirmatory HIV diagnosis, (2) being permanent residents or had stayed in Liangshan area for more than five years, (3) receiving ART in the local township health center for more than six months. The exclusion criteria for participants were having have major psychiatric illness (e.g., schizophrenia and bipolar disorder.
According to the inclusion criteria and exclusion criteria, medical staff phoned candidate participants about the study and confirmed their eligibility to join the study. All participants signed informed consent forms before enrollment. Finally, a total of 5,157 PLWH were included into the study.
Data collection
Questionnaire was used to ask all participants in private rooms to collected the socio-demographic information, possible route of HIV transmission, HIV-related behaviors and disease-related characteristics. The socio-demographic information including gender, age, occupation, current marital status, ethnicity, education level. The route of HIV transmission included drug injection, heterosexual behaviors and drug injection, casual sexual behavior and risk behaviors related to mother-to-child transmission treatment time. The HIV-related behaviors included information about needle sharing, number of casual sexual partners and history of sexually transmitted disease except for HIV. The disease related characteristics were extracted from their medical record, including viral load, stage of the disease, the initial regimen and time of ART and change of ART regimen. The above information was collected from National AIDS reporting system. When a newly diagnosed with HIV-1 infection was found or received ART, the detailed information (the initial regimen and time of ART) would be collected and inputted into system.
Laboratory tests
Each participant was asked to provide 5 mL of venous blood for detecting the viral load of HIV-1. The plasma samples were isolated from each participant, and preserved in a -80℃ freezer until analysis. HIV viral load was measured in the Sichuan Center for Disease Control and Prevention. Virological failure in ART was defined as HIV RNA level ≥ 1000 copies/ml[20, 21] after receiving ART for more than six months.
Nucleic acid extraction, amplification, sequencing and drug-resistance analyses
Total viral nucleic acid was extracted from 200-μl plasma using an automatic extraction machine (MagNA Pure LC 2.0 system, Roche, Branchburg, NJ, USA). The full-length protease gene in Pol region and the first 300 codons of the reverse transcriptase gene was amplified by using reverse transcription polymerase chain reaction (RT-PCR). The amplified products were purified in and sequenced at Beijing Genomics Research Center Ltd., in China.
The HIV-1 pol sequences obtained in the study, together with reference sequences of different subtypes and CRFs, were edited and aligned using ChromasProl.33, and the sequence alignments were manually performed by BIOEDIT Sequence Alignment Editor software (Ibis Biosciences, Carlsbad, CA, USA). HIV-1 genotype subtypes were identified by phylogenetic analysis, and determined by DNAStar (Madison, Wisconsin, USA) software. Drug resistance mutations were determined using the Stanford University HIV Drug Resistance Database (HIVDB): HIV db Program (http://hivdb.stanford.edu/). The detailed amplification, sequencing and drug resistance analyses were detailed described previously [22, 23].
Statistical analysis
Data were analyzed using SPSS version 21.0 for Windows (SPSS, Inc, Chicago, IL, USA). Categorical variables were expressed as frequencies and percentages and compared using the chi-squared (χ2) test. Using the presence of virological failure in ART as the dependent variable, univariate odds ratios (ORu) and 95% confidence interval (CI) for socio-demographic information, possible route of HIV transmission, HIV-related behaviors and disease-related characteristics were estimated. A summary model was obtained by fitting a multiple regression model that considered all variables that were significant in the respective univariate analysis as candidates. Multivariate odds ratios (ORm) and respective 95%CI were obtained. was used for data analysis, with p<0.05 taken as statistically significant.