Study design and setting
We used qualitative and quantitative methods of collection and both were implemented concurrently. For the quantitative component, we designed a retrospective cohort study using available data within the EMTCT program in which exposure to breastfeeding had occurred at some time in the past. For the qualitative component, we held interviews with healthcare providers as key informants and focus group discussions with HIV infected postpartum mothers to enrich and triangulate the quantitative results. The study design was largely quantitative and data convergence was established at results interpretation. The cohort consisted of all HIV infected postpartum mothers enrolled on EMTCT program between June 2014 and June 2018 at Ndejje Health Center (HC) IV, a large peri-urban health facility in Makindye Division, Wakiso district, Uganda.
Ndejje HC IV is a county level health facility according to Uganda’s health system accreditation system and provides curative, preventive, rehabilitative, and promotive health services to an estimated 100,000 population (8). Ndejje HC IV serves a greater proportion of HIV infected mothers largely from the eight divisions of Busabala, Masajja, Bunamwaya, Kigo, Mutungo, Ndejje, Sseguku, and Mutundwe. The health facility serves approximately 133,068 patients per year which is a high patient load compared to its threshold. Currently, the health facility has a total of 61 staffs, nine of which are dedicated to the MCH clinic.
The health facility implements the Uganda national and World Health Organization’s Option B Plus policy (9), which we have fully described elsewhere.(10, 11) Under this policy, all HIV infected pregnant mothers are started on Anti-retroviral Therapy (ART) for life regardless of their immunological and clinical status and HEIs receive Nevirapine syrup as prophylaxis from birth until six weeks adjusted according to weight and age bands.
At six weeks, Nevirapine prophylactic syrup is stopped and Cotrimoxazole prophylaxis is introduced and a dry blood spot obtained for HIV test using Deoxyribonucleic Acid-Polymerase Chain Reaction (DNA-PCR). At one year, breastfeeding ceases and a second dry blood spot is obtained at exactly six weeks after cessation of breastfeeding for DNA-PCR test. A final diagnosis of HIV is performed at 18 months using a rapid antibody test. However, when HEIs seroconvert (test positive for HIV) at any of the testing time points, ART is started and Cotrimoxazole prophylaxis continued for life. HEIs receive exclusive breastfeeding for the first six months of life followed by complementary feeding.
To minimize loss of mother-baby pairs and improve retention, Uganda introduced a concept of Mother-Baby Care Point (MBCP) which is a service delivery model in the maternal and child health (MCH) clinic. The MBCP delivery model ensures HIV infected mothers and HEIs are paired and cared for together until the HEI reaches 18 months postpartum, a time when the final HIV infection status is determined (12). Entry into the MBCP starts immediately after delivery because the mother and HEI have to receive HIV care as a single pair. The MBCP at Ndejje HC IV has four healthcare providers (two midwives and two nursing officers), with one of them as the team leader. It operates five days a week, Monday to Friday, from 8.00 am to 5.00 pm, in accordance to the Uganda labor laws. Mothers routinely receive health education and counseling on various topics like PMTCT, nutrition, newborn care practices, maternal and newborn hygiene and breastfeeding among others.
Data are collected through paper-based system using HIV/ART care cards and EID and EMTCT registers, and later entered into an electronic database, the Open Medical Records System (Open-MRS). To ensure data quality, weekly data reviews are conducted alongside continuous quality improvement initiatives to address gaps in quality of HIV care. Notable quality improvement initiatives at the MBCP have targeted improved recording and reporting of data in the EMTCT and EID registers, uptake of HIV testing at six weeks, and maternal nutritional assessment using Mid-upper arm circumference (MUAC) among others.
Study population and sample size
The study population consisted of all HIV infected postpartum mothers enrolled to EMTCT program for at least one year, and all of them were still receiving HIV care at the time of data abstraction. We excluded HIV infected postpartum mothers with infants below one year of age because cessation of breastfeeding is at one year. So it would be erroneous to measure cessation of breastfeeding in such mother-baby pairs. Mother-baby pairs transferred to other health facilities were equally excluded because it was not possible to obtain data on cessation of breastfeeding. We further excluded records for mother-baby pairs where the HEI had died or seroconverted and those gross missing data. We did not calculate a sample size but used census sampling as a retrospective cohort study that consisted of records review was conducted. Four focus group discussions (FGDS) of eight to 12 HIV infected postpartum mothers each were randomly selected among those attending the EMTCT clinic. Four healthcare providers directly involved in providing care at the MBCP were purposively selected as key informants and interviewed.
Study variables
Our outcome variable was cessation of breastfeeding at one year, measured on a binary scale (yes or no), defined as the number of HIV infected postpartum mothers documented in the early infant diagnosis of HIV (EID) register who had stopped breastfeeding at one year. The independent variables included: maternal variables such as age in years but later dichotomized as below 25 or 25 years and beyond, antenatal care (ANC) attendance at last pregnancy measured as yes and no, number of ANC visits at last pregnancy, ART regimen categorized as Tenofovir (TDF) or Zidovudine (AZT) containing regimens, parity measured as nulliparous, secundiparous, and multiparous, and place of delivery measured as health facility or home. The infant variables included age in months, sex (male or female), time of initiation of cotrimoxazole prophylaxis (before six works or six weeks and beyond), and same-day initiation of breastfeeding (yes or no).
The qualitative variables explored included modes of mother to child transmission of HIV and likely preventive measures, importance of breastfeeding, when to stop breastfeeding in an HEI, and reasons for stopping to breastfeed before or after one year of age.
Data collection
Qualitative data
To enrich and triangulate the quantitative results, we conducted qualitative interviews with HIV infected postpartum mothers and healthcare providers. In particular, we held four focus group discussions (FGDs), each consisting of eight to 12 HIV infected postpartum mothers selected randomly from amongst those attending the EMTCT clinic. The group consisted of mothers who had ceased breastfeeding and those due for cessation of breastfeeding at one year. The FGDs were held within the premises of the health facility in the local language, “Luganda”, by two research assistants (JGP and MN, both female MPH postgraduate students trained in qualitative research methods). One research assistant (JGP) moderated all the FGDs while the other (MN) audio-recorded the responses and probed where necessary. Each FGD lasted for about 40-60 minutes on average.
The moderator encouraged all the group members to ask questions and to provide comments as much as possible, irrespective of the correctness. Group dominance by some members of the FGD was minimized by directing some probing questions and comments to other members of that group who seemed to engage less in the discussions. For key informant interviews (KII), four healthcare providers namely, two Midwives and two nursing officers directly engaged in the provision of EMTCT services at the MBCP were purposively selected and interviewed to elicit their expert opinions on practices of cessation of breastfeeding among HIV infected mothers. The KII lasted for 30-45 minutes, also conducted within the health facility premises in English language, but in a quiet and convenient room. Both FGDs and KIIs were held until saturation was reached.(13)
Quantitative data
We reviewed the PMTCT and EID registers for all eligible participants and abstracted data using a standardized checklist between April and May 2019. Whenever data were missing in the EID and EMTCT registers, the HIV care cards and the electronic database (Open MRS) were used to retrieve such data where possible.
Data processing and analysis
Qualitative data
We audio-recorded all interviews and transcribed them verbatim. To ensure accuracy in transcription, we correlated the audio-recordings with the transcripts by replaying it while reading through the transcripts. We exported the transcripts to Nvivo, a qualitative data analysis software, for thematic analysis. Three reviewers JGP (MPH postgraduate student), JI (Public health specialist and mixed-methods research fellow), and SO (Public health specialist and research supervisor) read the transcripts thoroughly and independently, and identified emergent categories and themes through coding using inductive data analysis approach. Independent coding was used to prevent selective perception and interpretive biases in the coding process. JGP, JI and SO discussed the categories and themes in a group, and then developed explicit summaries describing each category and theme. Discrepancies in the emergent categories and themes were resolved by consensus and a final codebook was thereafter developed. The verbatim quotations were then used to enrich and triangulate the quantitative results.
Quantitative data
We single-entered quantitative data using Epi-Data version 3.1 (14) concurrent with quality control measures namely skip patterns, alerts, range and legal values, and then exported the data to Stata version 15 for analysis.(15) We analyzed numeric data using descriptive statistics of means and standard deviations, and categorical data using frequencies and percentages. The outcome variable was computed as the proportion of HIV infected postpartum mothers who had stopped breastfeeding at one year.
We tested differences in proportions of cessation of breastfeeding at one year with categorical variables using the Chi-squared test for large cell counts (five and more counts) and the Fisher’s exact test for smaller cell counts (less than five counts). To test mean differences in cessation of breastfeeding at one year with numerical variables such as age, we used the student’s t-test. We performed sensitivity analysis to examine the effect of missing data on cessation of breastfeeding feeding for infant sex and maternal ART regimen at bivariate analysis.
We considered variables with probability values (p values) less than five percent at bivariate analysis as statistically significant for univariable and multivariable analyses. We also considered variables such as maternal age, parity, and same-day initiation of breastfeeding as birth as biological plausible factors for cessation of breastfeeding at one year at univariable and multivariable analyses. Our data showed that the outcome variable was frequent (more than 10%). Accordingly, the odds ratio (OR) was not an appropriate measure of association because of overestimation.(16, 17) We hence used risk ratios (RR) for both unadjusted and adjusted analysis computed using the modified Poisson regression analysis with robust error variance to control for mild violations of the assumptions of Poisson regression analysis. We reported each RR with subsequent 95% confidence intervals (CI). We noted that five (2.1%) observations were missing for the variable infant sex but did not impute them at multivariate analysis because they were fewer than 10%.
Human subject issues
Clarke International University Research Ethics Committee, CIU-REC (reference #CIU-REC/0136), approved this study. We received administrative approval from the Health Department of Wakiso district (reference # CR: MSMC 220/1). Key informants and focus group participants provided a written informed consent after explaining the purpose and benefits of the research as well as the risks involved. Participation in the study was voluntary and participants were free to withdraw at any stage if they so wished. All participant data were handled with confidentiality and privacy, and individual identifiers were anonymized.