This retrospective cohort study used electronic health record data from TriNetX representing 110 US healthcare organizations covering 2010–2026. The investigators identified adults with documented HIV/HBV coinfection and classified social vulnerability (SV) using a four-domain framework: material vulnerability, healthcare access and engagement vulnerability, interpersonal vulnerability, and psychosocial vulnerability. Adults with and without documented SV were matched 1:1 by propensity score, yielding 2,024 participants in each group. Total observed follow-up exceeded 15,900 person-years. The dataset consisted of deidentified records; TriNetX obtained IRB exemption from the Western Institutional Review Board and no additional IRB review was required for this analysis.
Social vulnerability was associated with a substantially increased risk of death among adults with HIV/HBV. The reported hazard ratio for mortality was 2.06 (95% confidence interval [CI], 1.72–2.47), indicating roughly a twofold higher hazard of death in the group with documented SV compared with the matched group without SV. The authors highlight that the association with mortality was one of the most robust findings across outcomes assessed.
SV was associated with multiple measures of liver disease. The study reported increased hazards for a composite liver events outcome (hazard ratio [HR], 1.37; 95% CI, 1.07–1.76). More severe hepatic endpoints showed larger effect sizes: hepatic decompensation had an HR of 1.94 (95% CI, 1.39–2.70) and hepatic failure had an HR of 2.39 (95% CI, 1.53–3.73). These results indicate that documented social vulnerability correlated not only with more frequent liver-related events but also with higher likelihood of severe hepatic complications.
The investigators evaluated virologic outcomes for both hepatitis B virus and HIV. SV was associated with higher risk of HBV viremia (HR, 1.69; 95% CI, 1.32–2.16) and HIV viremia (HR, 2.05; 95% CI, 1.71–2.46). These findings suggest that individuals with documented SV had poorer control of both infections as measured by detectable or uncontrolled viremia in the clinical records used for the analysis.
Beyond hepatic and virologic outcomes, SV was associated with higher rates of several extrahepatic conditions. The study reported an association with major adverse cardiovascular events (HR, 1.47), chronic kidney disease (HR, 1.49), and diabetes (HR, 1.25). The abstract also indicates increased acute care utilization among those with SV, though specific utilization metrics and effect estimates beyond those listed were not detailed in the abstract.
The authors compared multidomain SV (combining vulnerabilities across the four domains) with single-domain measures. Multidomain SV generally showed stronger associations for most hepatic and virologic outcomes compared with single-domain SV. Reported hazard ratio ranges for multidomain SV were approximately 1.76–2.62 for certain outcomes versus 1.35–1.76 for single-domain SV, per the abstract. Among the domains examined, healthcare access and engagement vulnerability was most consistently associated with mortality and hepatic outcomes.
Based on the observed associations, the authors propose that assessment of social vulnerability during HIV/HBV care could improve risk stratification and help identify targets for interventions intended to reduce hepatic and extrahepatic morbidity and mortality. The study highlights healthcare access and engagement as a potentially actionable domain given its consistent link to mortality and liver outcomes. The manuscript emphasizes that SV assessment may identify patients at higher risk for virologic failure and organ complications.
Key design features are: retrospective cohort analysis of TriNetX electronic health record data from 110 US healthcare organizations; propensity score matching 1:1 to compare adults with HIV/HBV with and without documented SV; follow-up totaling more than 15,900 person-years. The investigators used a predefined four-domain SV framework (material, healthcare access and engagement, interpersonal, psychosocial). The authors report institutional review board exemption for the use of deidentified TriNetX data and declare no competing interests. The manuscript is a medRxiv preprint and has not been peer reviewed; the full article contains additional methods and results not reproduced in the abstract.
Note: All numerical effect estimates and study details above are drawn from the preprint abstract. The preprint states that full data and additional analyses are contained in the manuscript; readers should consult the full text for complete methods, sensitivity analyses, and secondary outcomes.