Background: Family planning is a cost effective strategy for prevention of mother to child transmission of HIV and reduction of maternal/infant morbidity and mortality. Contraceptive implants are a safe, effective, long term and reversible family planning method whose use remains low in Kenya. We therefore set out to determine and compare the uptake, and factors influencing uptake of immediate postpartum contraceptive implants among HIV infected and uninfected women at two hospitals in Kenya. Methods: This cross sectional study targeted postpartum mothers at two Kenyan district hospitals (one urban and one rural). All participants received general family planning and method specific (Implant) counseling followed by immediate insertion of contraceptive implants to those who consented. The data was analyzed by descriptive analysis, T-test, Chi square tests and logistic regression. Results: One hundred eighty-five participants were enrolled (91 HIV positive and 94 HIV negative) with a mean age of 26 years. HIV positive mothers were significantly older (27.5 years) than their HIV negative counterparts (24.5 years), P = 0.001. The two groups were comparable in education, employment, marital status and religious affiliation. Overall, the uptake of contraceptive implants in the immediate postpartum period was 50.3 % and higher among HIV negative than HIV positive participants (57 % vs. 43 %, P = 0.046). Multivariate analysis revealed that a negative HIV status (P = 0.017) and prior knowledge of contraceptive implants (P = 0.001) were independently associated with increased uptake of contraceptive implants. Conclusion: There was a high uptake of immediate postpartum contraceptive implants among both HIV infected and un-infected women; efforts therefore need to be made in promoting this method of family planning in Kenya and providing this method to women in the immediate postpartum period so as to utilize this critical opportunity to increase uptake and reduce the high unmet need for family planning.
We conducted a cross sectional study targeting HIV infected and uninfected postpartum women admitted to the postnatal wards (PNW) of two Kenyan district hospitals, Naivasha District Hospital (a rural hospital) and Mbagathi District Hospital (an urban hospital) from the months of July to October 2012. Naivasha and Mbagathi district hospitals are public hospitals that provide both general and specialized care including regular FP service. A room within the postnatal units was used to provide FP counseling and insertion of implants to those who consented. Study participants were postpartum mothers in the postnatal wards of the two hospitals after delivery and before discharge. One HIV positive mother was recruited for every one HIV negative mother. The HIV status was obtained from the patients’ ANC card during admission into the postnatal wards and then confirmed from the patients’ hospital files. Those with unknown HIV status, were tested in the maternity unit and results documented in the patients’ files. The enrollment criteria included; consenting adult postpartum women with known HIV status, normotensive and with no known serious medical conditions such as active liver disease, deep venous thrombosis, migraine with aura, renal failure or breast cancer. Ethical approval was obtained from the Kenyatta National Hospital/University of Nairobi (KNH/UON) Ethics and Research committee. Consent was obtained in writing from all the study participants after adequate explanation for enrollment in this study. The study was conducted using an anonymous survey with no name identifying information provided in the questionnaire. All measures to maintain anonymity and confidentiality were strictly followed. Participant enrollment into the study was carried out by convenience sampling based the admission of patients in the PNW until the numbers of HIV positive and negative balanced. Pretesting of the questionnaire was done at the study sites using the same protocol and no revision was required. Trained research nurses administered the structured questionnaire and the information collected included; socio-demographic characteristics, parity, contraceptive knowledge and use, spouse approval of contraceptive use, future fertility intentions and knowledge and opinions about contraceptive implants. This was then followed by standardized contraceptive counseling about all available FP methods and specifically on the study method (implants). Those who further consented to insertion of contraceptive implants were inserted immediately after signing part B of the consent form and were encouraged to continue using barrier methods. A card that indicated the method given, date inserted and expiry date was given together with post insertion instructions on wound care. They were then advised to go to FP clinic of their choice for follow up. Those who did not consent to implant insertion were counseled on importance of FP use and referred to FP clinics of their choice for routine postpartum care (Fig. 1). Algorithm for enrolment of study participants Descriptive statistics were used to describe baseline maternal characteristics such as age, parity, education levels and FP knowledge. Mean and median was used for continuous variables and frequencies, and proportions were used for categorical variables. T-test was applied to compare means, while a two-sided chi-square test was used to compare proportions between the two groups. Fisher’s exact test was also utilized when at least one cell had a value of zero. Logistic regression was then applied to test the strength of association between demographic and reproductive variables and the primary outcome – uptake of contraceptive implants. The significant factors were then subjected to multivariate analysis to analyse for independent association. The data was entered into password protected Microsoft Access Database and subsequently transferred to SPSS statistical package for analysis. SPSS 16.0, Inc., 2007. One of the study limitations was selection bias, due to the non-probability sampling method but this was mitigated by the large sample size.
N/A