Adolescent girls and young women (AGYW) have a disproportionately high incidence of HIV compared to males of the same age in Uganda. AGYW are a priority sub-group for daily oral Pre-Exposure Prophylaxis (PrEP), but their adherence has consistently remained low. Short Message Service (SMS) reminders could improve adherence to PrEP in AGYW. However, there is a paucity of literature about the acceptability of SMS reminders among AGYW using PrEP. We assessed the level of acceptability of SMS reminders as a PrEP adherence support tool and the associated factors, among AGYW in Mukono district, Central Uganda.

We consecutively enrolled AGYW using PrEP in Mukono district in a cross-sectional study. A structured pre-tested questionnaire was administered to participants by three trained research assistants. Data were analyzed in STATA 17.0; continuous variables were summarized using median and interquartile range (IQR) while categorical variables were summarized using frequencies and percentages. Acceptability of SMS was defined as willingness to accept SMS reminders to support PrEP adherence and was assessed using the seven constructs of the theoretical framework of acceptability. The relationship between the outcome and independent variables was assessed using a modified Poisson regression with robust standard errors.

During 2022, 142 AGYW with median age 22 years (IQR; 18, 24) of whom 80.3% owned a personal mobile phone were assessed. SMS reminders were highly acceptable [90.9%, 95% Confidence Interval (CI) [84.9, 95.0]]. Rural residence (aPR: 0.92, 95% CI (0.84, 0.99)) and having belief that SMS cannot breach individual’s privacy (aPR: 1.40, 95% CI (1.07, 1.84)) were significantly associated with acceptability of SMS reminders.

Having a personal mobile phone and acceptability of SMS reminders were high in this sub-population. SMS reminder can be leveraged to support AGYW to adhere to PrEP but should be designed in a way that maintains confidentiality, and supports AGYW living in rural settings.

The data set was collected through a reseacher administered questionnaire. 

The main dependent variable was acceptability of SMS reminders. This was measured using the seven constructs derived from the Theoretical Framework of Acceptability (TFA)(1). These include; affective attitude, burden, perceived effectiveness, ethicality, intervention coherence, opportunity costs, and self-efficacy. A 5-point Likert item question per construct was used and each level of a Likert scale was given a weight ranging from one to five. The summated scores from the weights assigned to each response were computed. The obtained summated acceptability score was then dichotomized using the 50th percentile of the possible summated scores which ranges from 7 to 35 (the 50th percentile is 21). Therefore “Acceptability of SMS reminders" was defined as a value greater than 21. 

The independent variables were captured as described in the data dictionary attached

Data analysis was performed in STATA version 17.0 (Texas, USA).  At Univariate analysis, categorical variables were summarized using frequencies and percentages; and continous variables were summarized with median and inter-quartile range (IQR).

The level of acceptability was calculated as a proportion of total participants that accepted the use of SMS reminders and its exact confidence interval (CI) was calculated.

In the evaluation of factors associated with acceptability of SMS reminders, the primary outcome was the acceptability of SMS reminders as a PrEP adherence support tool recorded as a binary outcome (“Acceptable” versus “Not acceptable”). The relationship between the outcome and the independent variables was assessed using a modified Poisson regression model with robust standard errors. At bivariate analysis, the crude prevalence ratio (cPR) was computed together with the CI and P-value for each independent variable. The independent variables with a P value of less than 0.2 were included in the multivariate model.

At multivariate analysis, Interaction was assessed using the chunk test (-2loglikelihood ratio test). Confounding was assessed and an independent variable was considered a confounder if they changed the magnitude of the adjusted prevalence ratio of another independent variable in the multivariate model by more than 10%. The adjusted prevalence ratio (aPR) was computed and reported together with the CI and P-value of each independent variable in the final model. statistical significancy was set at 0.05.