Study sampling and participants
The dataset used was from the 2019-20 Rwanda Demographic Survey (RDHS), which was a cross-sectional study and employed a two-stage sample design. The first stage involved cluster selection consisting of enumeration areas (EAs), while the second stage involved systematic sampling of households in all the selected EAs leading to a total of 13,005 households [24]. In particular, the data used in this analysis were from the household and the woman’s questionnaires.
During this survey (RDHS), the data collection period was from November 2019 to July 2020, which took longer than expected due to the COVID-19 pandemic restrictions [1]. Women aged 15-49 years who were either permanent residents of the selected households or visitors who stayed in the household the night before the survey were eligible to be interviewed. Of the 13,005 households that were selected for the survey, 12,951 were occupied and 12,949 were successfully interviewed giving a response rate of 99.9% [24]. The eligible sample was 14,675 women aged 15-49, but 14,634 women were successfully interviewed leading to a 99.7% response rate [24]. For this analysis, we considered only adolescent girls aged 15-19 years interviewed during the survey, and of the selected households, whose sample was 3,258.
Variables
Dependent variables
The study outcome variable was comprehensive knowledge about HIV/AIDS, which was a composite variable scored from six yes/no questions; 1) Always using condoms during sex can reduce the risk of getting HIV, 2) Having only one sexual partner, who has no other partners can reduce risk of getting HIV, 3) A healthy looking person can have HIV, 4) Can get HIV from mosquito bites, 5) Can get HIV by sharing food with a person who has AIDS, and 6) Can get HIV by witchcraft or supernatural means [25,26]. Comprehensive knowledge was considered if an adolescent girl answered all six questions correctly; that is, “Yes” for questions 1,2, and 3, and “No” for questions 4,5, and 6).
Explanatory variables
Based on the available literature and data, we included possible determinants of comprehensive knowledge about HIV/AIDS [25-29]. Eighteen (18) variables were considered and of these, two were community-level factors that included; place of residence (categorized as rural and urban), and region of residence (Kigali, South, West, East and North). Four household-level factors included; household size (classified into “less than six” and “six and above”), sex of household head (female and male), wealth index (categorized into five quintiles that ranged from the poorest to the richest quintile), and health insurance (yes and no). Wealth index was calculated by RDHS from information on household asset ownership using Principal Component Analysis [24]. Twelve (12) individual-level factors were also considered in the analysis, including; age (categorized as 15, 16, 17, 18 and 19 years), educational level (no education, primary, secondary and tertiary), working status (working and not working), marital status (married and unmarried), religion (Catholic, Protestant, and others), history of having an STI in last 12 months (yes and no), and exposure to news, radio and television (yes and no), contraceptive use (yes and no), having done an HIV test before (yes and no) and owning a mobile phone (yes and no).
Statistical analysis
In this analysis, we applied the DHS sample weights to account for the unequal probability sampling in different strata and ensure the representativeness of the study results [30,31]. We used Statistical Package for Social Science’s (SPSS) (version 25.0) complex samples package incorporating the following variables in the analysis plan to account for the multistage sample design inherent in the RDHS dataset: individual sample weight, sample strata for sampling errors/design, and cluster number [25,30]. Initially, we did descriptive statistics for both dependent and independent variables, where frequencies and proportions/percentages for categorical dependent and independent variables have been presented. We, then, conducted bivariable logistic regression to assess the association of each independent variable (i.e selected socio-demographic factors) with comprehensive HIV knowledge and crude odds ratio (COR), 95% confidence interval (CI) and p-values are presented. Independent variables found significant at the bivariable level with p-values less than 0.25 were then included in the multivariable model. Moreover, independent variables that were non-significant at the bivariable analysis level but were associated with comprehensive knowledge in previous studies were also included in the multivariable logistic regression model. The final model controlled for all the included variables/factors where we calculated and presented their respective adjusted odds ratios (AOR), 95% confidence intervals (CI) and p-values, at a statistical significance level of 0.05. All selected variables in the model were assessed for multi-collinearity, which was considered present if the variables had a variance inflation factor (VIF) greater than 10 [32]. However, none of the variables had a VIF above 3.