People living with HIV (PLWH) and substance use disorder (SUD) are particularly vulnerable to harmful health consequences of the global COVID-19 pandemic. The health and social consequences of the pandemic may exacerbate substance misuse and poor management of HIV among this population. This study compares substance use and HIV care before and during the pandemic using data collected weekly through an opioid relapse prevention and HIV management mobile-health intervention. We found that during the pandemic, PLWH and SUD have increased illicit substance use and contact with other substance-using individuals and decreased their confidence to stay sober and attend recovery meetings. The proportion of people missing their HIV medications also increased, and confidence to attend HIV follow-up appointments decreased. Optimal support for PLWH and SUD is critical during pandemics like COVID-19, as drug-related and HIV antiretroviral therapy (ART) non-adherence risks such as overdose, unsafe sexual behaviors, and transmission of infectious diseases may unfold.

The social mediation role of mobile technology is typified by mHealth apps designed to connect individuals to others and support substance use disorder (SUD) recovery. In this study, we examined the use and utility of one such app designed to support people living with HIV (PLWH) and SUD. Drawing on Ling’s emphasis on reciprocity and micro-coordination in mobile telephony as a social mediation technology, we gathered digital trace data from app logs to construct two metrics, initiation (i.e. whether a particular feature is engaged on a given day) and intensity (i.e. degree of involvement in the activity when engaged on that day), at three levels of communication—networked (one-to-many), dyadic (one-to-one), and intraindividual (self-to-self). We consider these system features alongside use of information resources, games and relaxation links, a meeting and events calendar, and support tools to address use urges. We found few differences in patterns of use by race, sex, and age, though African Americans were less likely to engage in intraindividual expression, whereas women and older users were more likely to make use of this feature. The initiation and intensity of network and dyadic reception, as well as the intensity of network expression, predicts recovery outcomes as measured on a weekly “check-in” survey, suggesting the utility of mobile log data for digital phenotyping in mHealth. By implementing this app during the COVID-19 pandemic, the study also found the disruption caused by national lockdown was negatively related to the app use.