Study site and design
This analysis uses data collected as part of a national survey conducted among AGYW aged 10-24 years in 14 purposely selected districts between September 3 and 7, 2020. The survey was conducted as part of repeated cross-sectional surveys conducted among AGYW in selected priority districts in Uganda with funding from the Global Fund. A detailed description of the main survey methodology is documented elsewhere [28]. In brief, we used a multi-stage sampling approach to select villages within each district and households within each village. Using a list of villages obtained from the Uganda Bureau of Statistics, the study team randomly sampled up to twenty-two villages per district. Upon obtaining administrative clearance to conduct the survey in a district, teams visited selected villages and, working with the local leadership in each village, generated a list of households that were presumed to have AGYW aged 10-14 years. Teams visited the selected households to screen AGYW for study eligibility and eligible girls were invited to participate in the study after providing written informed consent.
Study population
This analysis uses data for 2109 sexually-active AGYW who were interviewed as part of the above-mentioned survey. AGYW were considered to be sexually active if they reported that they had had penetrative vaginal sex with a male partner in the 12 months preceding the survey.
Data extraction
The questionnaire for the national survey included questions on socio-demographic and behavioral characteristics, contraception, menstruation hygiene management, and intimate partner violence, among other sections. For this analysis, we extracted data on socio-demographic characteristics (e.g. age in completed years, highest level of education, current marital status, readability and household possessions, among other characteristics), behavioral characteristics (e.g. ever had sex, sexual intercourse in the past 12 months, among others), contraception history (ever and current use), discussion with husband/partner prior to contraceptive use, who made the decision about the use of contraception, who made the final decision on the specific FP method that the AGYW were currently using, and interest in self-care-oriented family planning methods.
Measures of variables
We assessed two primary outcomes: a) contraception decision-making autonomy (defined as the proportion of AGYW who made a decision to use contraception on their own) among sexually-active AGYW, and b) interest in self care-oriented FP methods among AGYW. Contraception decision-making was assessed among sexually active AGYW who were current contraceptive users while interest in self care-oriented FP was assessed among all sexually active AGYW. We defined the term ‘contraception’ in the generic sense to refer to any modern or traditional FP methods used by AGYW to avoid or delay getting pregnant. Contraception decision-making autonomy was assessed with the question: ‘Would you say that using contraception is mainly your decision, mainly your husband/partner’s decision or did you both decide together?’ which was coded ‘1=Mainly my decision’, ‘2=Mainly my partner’s decision’, or ‘3=Joint decision’. If a girl had more than one partner at the time of interview, it was up to them to decide if, in their opinion, using contraception was mainly their own decision, any of their partner’s decision, or jointly decided by them and any of their partners.
Independent variables included the factors that are likely to affect contraception decision-making autonomy including contraceptive discussion with partner prior to contraceptive use, schooling status (i.e. in- or out-of-school), highest level of education attained (coded as ‘no education’, ‘primary education’ or ‘post-primary education [secondary school education or higher]’), marital status (coded as ‘never married’, ‘in a relationship but not married’, ‘currently married’ or ‘divorced/widowed/separated’), ability to read health risk messages in the local language (as a measure of literacy), phone ownership, age-group (categorized as 10-16; 17-19, 20-24) and socio-economic status. A respondent was presumed to have discussed with their partner prior to using contraception if they answered the question, ‘Before you started using [most recent / current method], had you discussed the decision to delay or avoid pregnancy with your husband/sexual partner?’, in the affirmative. Current contraceptive users were asked about who made the final decision about the specific FP method that they were currently using, which was coded as: ‘1=respondent alone’, ‘2=provider or partner’, ‘3=respondent and provider’, or ‘4=respondent and partner’. Age-group was categorized as 10-16 years; 17-19 and 20-24 based on available data. There were very few girls that were sexually active and had ever used contraception in the category 10-14 years to allow for meaningful statistical analysis; thus, a special category, 10-16 years, was created.
We used household possessions (e.g. radio, television, bicycle, motorcycle, cell phone, running water inside the house or inside the compound, among other aspects) to construct a socio-economic status (SES) index. Each household item was assigned a weight ascertained through principal components analysis. Then, the scores were standardized in relation to a standard normal distribution with a mean of zero and a standard deviation of one. For each individual, the scores on household possessions were then summed up and individuals were ranked and sub-divided into wealth tertiles (i.e. lowest, middle, and highest), depending on their scores, with each tertile containing approximately one-third (33.3%) of the respondents.
To assess interest in self care-oriented FP methods, all sexually-active AGYW (irrespective of current contraceptive use status) were asked: ‘Would you be interested in ways to access information on the range of ways or methods that you can use to delay or avoid pregnancy without seeing a provider or a member of the village health team (VHT)?’ coded as ‘1=Yes’ and ‘2=No’. AGYW that responded in the affirmative were asked if they were interested in accessing information on ways to obtain or use specific FP methods without seeing a provider or a VHT. Responses to these questions were used to measure interest in ‘self-care-oriented FP methods’, defined as those FP methods that the respondents could use on their own without the assistance of a health provider or a VHT. Respondents were also asked questions about their interest in self-management of FP side effects and bleeding changes as a result of FP use.
Data analysis
To assess contraception decision-making autonomy, we computed descriptive statistics to determine the percentage of current contraceptive users stratified by who made the decision to use contraception (i.e. respondent, partner, or both partner and respondent). We conducted bivariate analyses to determine the association between contraception decision-making autonomy and each independent variable; all variables (discussion with husband/partner about the need to delay or avoid pregnancy prior to contraception, marital status, age-group, wealth tertile, schooling status, highest level of education attained, one’s ability to read text prepared in the local language, and phone ownership) were considered for the final multivariable analysis. We used a modified Poisson regression model to identify the factors that were independently associated with contraception decision-making autonomy, after adjusting for potential confounders, and report the adjusted prevalence ratios (adj. PR) and 95% confidence intervals (95%CI) associated with these factors, using a p-value of 0.05 as a measure of statistical significance.
To assess interest in self care-oriented FP methods, we computed descriptive statistics to determine the percentage of all sexually-active AGYW that were interested in receiving information about self care-oriented FP methods in general and specifically for each method. We assumed that interest in self care-oriented FP methods would vary by age (e.g. the very young adolescent girls might want to use a method that they can manage on their own, for fear that their parents might find out that they are already sexually active), and therefore conducted cross-tabulations of interest in self care-oriented FP methods by age-group. We also computed the percentage of those that were interested in receiving additional information on how to manage FP side-effects and bleeding changes due to contraceptive use. No inferential statistics were computed for this outcome.
Ethical considerations
The 2020 survey from which the data used in this analysis were derived was approved by the Makerere University School of Public Health Higher Degrees Research and Ethics Committee (Protocol#: 749) and registered with the Uganda National Council for Science and Technology (Protocol#: SS411 ES). All study procedures were performed in accordance with the ethical standards of the institutional and/or national research committees and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. In summary, we obtained written informed consent (or assent, as the case may be) from all the respondents. Adolescent girls aged 10-17 years who were still living with their parents or guardians were interviewed only after we obtained parental/guardian consent and the adolescent girls’ assent to the interview. However, those aged 10-17 years who were married, or living on their own, were interviewed after obtaining informed consent from them since these were considered as emancipated minors, as per the research guidelines from the Uganda National Council for Science and Technology.