Factors associated with contraceptive uptake among HIV-positive women on dolutegravir-based antiretroviral therapy – a cross-sectional survey in urban Uganda | BMC Women’s Health
This was a cross-sectional survey conducted from April to July 2019 across the five ART clinics of the Kampala Capital City Authority (KCCA) in Uganda, supported by the Infectious Diseases Institute (IDI).
We included HIV-positive women aged 15–49 years receiving DTG-based regimens; pregnant women were excluded. We estimated the study sample size to be 393 using Kish’s formula ; at a 95% confidence interval, a Z value of 1.96, a margin of error of 0.05 and a proportion of FP use among HIV-positive women (P) of 36% . Stratified random sampling with proportional allocation of participants to the five institutions was used to select participants eligible for study enrollment.
Eligible participants were invited to study sites for data collection by telephone. We conducted data collection using semi-structured questionnaires administered by an interviewer and pretested on five participants who were then not included in the study. All data collection documents were translated into the local language (Luganda). The questionnaires were validated using previous cross-sectional studies that assessed contraceptive use among HIV-positive women . Questionnaires were cross-checked daily for completeness to ensure quality.
We collected data on the following variables: age in years (stratified as 15–24, 25–29, 30–34, 35–39, 40–44, 45–49), health facility coded as 1, 2, 3, 4 and 5 respectively. Level of education (primary, secondary, tertiary and none), marital status (married, single and in a couple), employment (salaried, unemployed and self-employed) parity in number of children (1–2, 3–4, 5– 6 and none), religion (Anglican, Catholic, Muslim and other categories including Pentecostals, Jehovah’s Witnesses, Adventists, Buddhists), desire to have children, sexual activity in the past month, knowledge of any method contraceptive, knowledge of DTG side effects, FP discussion with partner, and FP counseling at facility.
We also collected data on contraceptive use and the type of contraceptives used at the time. We have defined modern contraceptive products or medical procedures used to prevent pregnancy, such as oral contraceptives, intrauterine device (IUD), condoms, progestogen-only injections, implant under -cutaneous, vaginal barrier methods, spermicides, emergency contraception, female and male sterilization. We defined traditional birth control methods as withdrawal, abstinence, lactational amenorrhea, fertility awareness methods, the rhythm method, and moon beads.
For data analysis, we used version 14 of STATA. We used descriptive statistics to describe participant characteristics and then modified the Poisson regression model to identify predictors of contraceptive use. Since the prevalence of contraceptive use in this study was common (more than 10%), the Poisson model was adapted as an alternative to a logistic model. We examined the goodness of fit of the Poisson model, modified it to a generalized linear model with Poisson family, logarithmic link, and reported robust standard errors for unadjusted and adjusted ratios. All independent variables with a p value