A mobile health-based alcohol reduction intervention and its usability among persons with Human Immunodeficiency Virus in Uganda.

Background: Human Immunodeficiency Virus (HIV) and alcohol use are widespread in sub-Saharan Africa (sSA), negatively impacting the continuum of HIV care, from prevention and testing to treatment outcomes. Reducing alcohol consumption among people with HIV (PWH) is therefore a public health priority. Most alcohol reduction strategies require in-person interactions and can be challenging during pandemics and social unrest, which are common in sSA. Mobile Health (mHealth) interventions may offer a more accessible reduction counselling, an accessible alternative to, or may augment, in-person interventions. However, the usability of these mHealth interventions in the target population should be assessed before implementation. This analysis evaluates usability of an mHealth intervention (automated booster sessions) to augment in-person alcohol use reduction counselling and examines its association with booster intervention uptake among PWH who self-reported unhealthy alcohol use. Methods: We analyzed data from PWH who participated in an alcohol reduction counselling intervention. Participants were PWH who self-reported unhealthy alcohol use in the prior 3 months by the Alcohol Use Disorders Identification Test-Consumption (AUDIT-C positive; ≥3 women, ≥4 men). We analyzed data from one of the intervention arms, called the technology booster study arm. In this arm, participants received an in-person counselling intervention that was augmented by technology-based boosters delivered over 3 months. These boosters were delivered via interactive voice response (IVR) or short message service (SMS) text messages, based on participant preference. Usability of the IVR and SMS sessions was assessed at 6 months using the Systems Usability Scale (SUS). Booster intervention uptake, the outcome variable for this analysis, was defined as the percentage of booster sessions during the intervention period in which the participant successfully entered a personal identification number (PIN) needed to start the session. Multivariable linear regression (adjusting for sex, age, being literate, social desirability, and baseline high-risk alcohol use) was conducted to examine the relationship between usability and booster intervention uptake. Results: Of the 86 participants randomized to the technology arm, 35% were female, with a median age of 40 years [interquartile range (IQR): 32-47]. Fifty-six (65%) chose to receive IVR calls, while 30 (35%) chose SMS; the median booster intervention uptake was 73% (IQR: 53-83). The median SUS score (out of 100) was 90 (IQR: 85-90). SUS score and being literate were associated with booster intervention uptake [adjusted β =0.65, 95% confidence interval (CI): 0.02-1.29, P=0.04, and β =21.33, 95% CI: 7.65-35.01, reduction counselling, an 0.01, respectively]. Conclusions: In this analysis of usability and booster intervention uptake among PWH using an mHealth-augmented alcohol reduction intervention, usability was associated with increased booster intervention uptake, as was literacy. Understanding the usability aspects of an mHealth intervention and the level of literacy in each unique population may be predictive of intervention uptake rates.