Study design and setting
Tsepamo study overview
We performed an analysis of infant feeding and HIV RNA data collected in the Tsepamo Study, an ongoing non-interventional birth outcomes surveillance study that collects data from obstetric records at large public maternity wards throughout Botswana . The primary study aims of Tsepamo are to evaluate adverse birth outcomes and congenital abnormalities by HIV status and ART regimen. The study occurs at geographically distributed sites in major population areas of Botswana (Fig. 1), where > 95% of women deliver in a healthcare facility. Although the study expanded from eight to 18 sites in 2018, only data from the original eight sites were included in this analysis; these sites were located in Gaborone and Francistown (tertiary referral centres) and Maun, Serowe, Selebi-Phikwe, Mahalapye, Molepolole and Ghanzi (district and primary-level hospitals).
Data were abstracted from obstetric record cards into an electronic database by trained research assistants at the time of maternal discharge from the postnatal ward. The obstetric record card is a standardized government booklet to record the entirety of medical care during pregnancy and delivery, that is started at the first antenatal clinic (ANC) visit, and brought by the mother to each subsequent ANC visit and to the delivery site. Information extracted from the obstetric record included maternal demographic characteristics, medical history, medications prescribed at the time of conception and during pregnancy, maternal diagnoses during pregnancy, infant birth record, type of delivery, APGAR scores, gestational age, birthweight, congenital abnormalities, and vital status of the infant(s) at time of discharge. For WLHIV, the date of HIV diagnosis, most recent CD4 cell count, and antiretroviral history (including start date, regimen, and any switch or discontinuation during pregnancy) were also extracted. When available in the obstetric record, HIV RNA results were recorded. Births that occurred before arrival at the hospital and < 24 weeks gestation were excluded.
For this analysis, we used data collected between September 2016 and March 2019 and women were included if they were living with HIV, if their baby was alive at the time of discharge, and if they had a validated feeding method recorded at discharge. Validation started in September 2016 and occurred through direct feeding observation or by checking the prevention of mother-to-child transmission of HIV (PMTCT) counsellor report or formula dispensing log.
All HIV RNA results documented in the obstetric record were abstracted. We verified a subset of randomly selected records from our sample (150 with and 150 without a documented HIV RNA result in the obstetric record) to determine whether an HIV RNA test had occurred during pregnancy. This verification was performed using the Integrated Patient Management System (IPMS), a nationwide electronic system that includes laboratory records. After February 2018, we also evaluated participants’ Infectious Disease Control Centre (IDCC) cards (the medical record for outpatient HIV care), when available, to further identify HIV RNA results missing from the available obstetric record. However, it should be noted that the obstetric record was the only source of information routinely available through the Botswana PMTCT programme to help midwives in counselling for appropriate feeding recommendations at the time of delivery. While the IPMS captures laboratory records nationwide, not all maternity facilities have access to these electronic laboratory records. Laboratory results of participants are routinely sent non-electronically to antenatal facilities for the nursing staff to transcribe them into participants’ obstetric records.
Data were extracted from the electronic database in an excel format and analysed in Stata (Version 16, StataCorp, College Station, Texas). Descriptive statistics were used to describe the infant feeding choices of WLHIV (proportions of women in each feeding group). To identify factors associated with infant feeding choices of WLHIV, we performed univariable and multivariable logistic regression analyses. Infant feeding choice was categorized as breastfeeding versus formula feeding where breastfeeding was used as the reference category in logistic regression models. All independent variables that were significant or nearly significant in univariable analysis (P < 0.1) were included in the multivariable model. Statistical significance was inferred at a P-value of < 0.05. The outcome variable used was feeding choice at discharge. Independent variables of interest included age, marital status, education, occupation, nationality, delivery site, received antenatal care, documented viral load during pregnancy on the obstetric record, ART status during pregnancy, and gravida.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.