Comparing Sociodemographic, Reproductive and Clinicopathological Characteristics Associated With Hormonal Contraceptive Use Among Women With Uncomplicated Diabetes Mellitus at Kenyatta National Hospital

Abstract

Introduction: Planned pregnancy is recommended for women with diabetes mellitus due to increased maternal and perinatal risks and adverse outcomes in pregnancy. The need for effective reversible and safe contraception is therefore essential to prevent pregnancy while the maternal metabolic conditions are optimized for pregnancy. Historically, all women with diabetes mellitus were restricted to non-hormonal methods due to potential adverse events from hormonal methods. Newer studies have, however, documented the safety of these methods in women with uncomplicated diabetes mellitus. Uncomplicated diabetes mellitus has no vascular disease while complicated diabetes mellitus is attended by microvascular or macrovascular disease. Hormonal contraception is, thus, safe, and effective for women with uncomplicated diabetes mellitus following the WHO Medical Eligibility Criteria for contraceptive use. Data on the use of hormonal methods by women with uncomplicated diabetes mellitus is lacking. Objective: To compare sociodemographic, reproductive and clinicopathological characteristics between women with uncomplicated diabetes mellitus who are using versus those not using hormonal contraception at Kenyatta National Hospital. Methods: A comparative cross-sectional study of 171 women with diabetes mellitus was undertaken at Kenyatta National Hospital. Sexually active, non-sterilized women with uncomplicated diabetes mellitus between 18 – 49 years of age were recruited from the Family planning, Gynecology and Diabetic outpatient clinics from October 2020 to December 2020. Fiftyfive of the participants recruited were using hormonal contraception while one hundred and sixteen lacked contraception or were on non-hormonal contraception. The Medical Eligibility Criteria was used as the theoretical framework and eligibility for hormonal contraception xi

determined at Category 1 and 2. Data was collected using an interviewer-administered structured pre-tested questionnaire. Quantitative data obtained was analyzed using STATA version 14.3 software. Bivariate analysis was used to determine factors associated with hormonal contraceptive use at 95% confidence interval and a statistical significance level of p<0.05. Multivariable logistic regression was done on significant factors identified from bivariate analysis according to their individual p value at 95% confidence interval and a statistical significance level of p<0.05. Results: Between October 2020 and December 2020, 300 women were screened and 60% found to be eligible for hormonal contraceptive use. Of those who were eligible, 8 were excluded because of failure to sign consent for the study. The sociodemographic characteristics were comparable between the two populations. Hormonal contraceptive use was less likely among those who had received contraceptive counselling (p=0.011). However, it was more likely with those who were satisfied with the counselling received (95% CI 4.2 [1.93-8.81]; p<0.01) and those who used the information received to choose the contraceptive method (95% CI 7.00 [3.06-16.3]; p<0.01). Women on hormonal contraception were also 2.24-fold (95% CI 1.11-4.47; P=0.026) more likely to make the decision in consultation with a partner. Presence of a complicated previous pregnancy was more likely with hormonal contraceptive users (95% CI 2.38 [1.17-4.79]; p=0.018). Having an unplanned previous pregnancy was also more among hormonal contraceptive users (25.5% vs 14%). It was found that hormonal contraception users were 3.76-fold (95% CI 1.32-10.43; p<0.01) more likely to have the method being used recommended by a doctor than other health care providers. Those who had been screened for neuropathy were 3.84-fold (95% CI 1.73-8.29; p<0.01) and screened for thrombosis 3.57-fold (95% CI 1.68-7.53; p<0.05) more likely to be using hormonal contraception. However, after adjusting for confounders, the two populations were found to be comparable in sociodemographic, reproductive and clinicopathological characteristics. Conclusions: Hormonal contraceptive use is not associated with poorer sociodemographic, reproductive and clinicopathological characteristics in women with uncomplicated diabetes mellitus.