Evaluating the agreement between different substance use recall periods in multiple HIV cohorts.

BACKGROUND: This study aims to evaluate the agreement in substance use on both binary and ordinal scales between 3-month and 6-month recall periods with samples from different communities, demographic backgrounds, and HIV status. METHODS: We administered the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST) to 799 participants from three different North American cohorts focused on substance use and HIV. We conducted a within-person agreement analysis by calculating the agreement levels and Kappa statistic between data collected using the 3-month recall ASSIST and 6-month custom substance use surveys as well as different terminology for each substance in multiple cohorts. RESULTS: For all drugs studied, the agreement on the binary use or ordinal frequency of use metrics showed a high agreement level between 80.4% and 97.9% and an adequate adjusted kappa value between 0.61 and 0.96, suggesting substantial agreement. According to the agreement criteria we proposed, substance use data collected using different recall periods and with variation in drug names can be harmonized across cohorts. CONCLUSIONS: This study is the first to evaluate the feasibility of data harmonization of substance use by demonstrating high level of agreement between different recall periods in different cohorts. The results can inform data harmonization efforts in consortia where data are collected from cohorts using different questions and recall periods.

Patterns of Antiretroviral Therapy Use and Immunologic Profiles at Enrollment in the REPRIEVE Trial.

BACKGROUND: Patterns of antiretroviral therapy (ART) use and immunologic correlates vary globally, and contemporary trends are not well described. METHODS: The REPRIEVE trial (Randomized Trial to Prevent Vascular Events in HIV) enrolled persons with human immunodeficiency virus (HIV) who were aged 40-75 years, receiving ART, and had low-to-moderate cardiovascular disease risk. ART use was summarized within Global Burden of Disease (GBD) super-regions, with adjusted linear and logistic regression analyses examining associations with immune parameters and key demographics. RESULTS: A total of 7770 participants were enrolled, with a median age of 50 years (interquartile range, 45-55 years); 31% were female, 43% were black or African American, 15% were Asian, 56% had a body mass index >25 (calculated as weight in kilograms divided by height in meters squared), and 49% were current or former smokers. The median CD4 T-cell count was 620/µL (interquartile range, 447-826/ µ L), and the median duration of prior ART use, 9.5 years (5.3-14.8) years. The most common ART regimens were nucleoside/nucleotide reverse-transcriptase inhibitor (NRTI) plus nonnucleoside reverse-transcriptase inhibitor (43%), NRTI plus integrase strand transfer inhibitor (25%), and NRTI plus protease inhibitor (19%). Entry ART varied by GBD region, with shifts during the trial enrollment period. In adjusted analyses, entry CD4 cell count and CD4/CD8 ratio were associated with GBD region, sex, entry regimen, duration of ART, and nadir CD4 cell count; CD4 and CD8 cell counts were also associated with body mass index and smoking status. CONCLUSIONS: There were substantial variations in ART use by geographic region and over time, likely reflecting the local availability of specific medications, changes in treatment guidelines and provider/patient preferences. The analyses of CD4 cell counts and CD4/CD8 ratios may provide valuable insights regarding immune correlates and outcomes in people living with HIV. CLINICAL TRIALS REGISTRATION: NCT02344290.