HIV-free survival among breastfed infants born to HIV-positive women in post-conflict northern Uganda: a perspective from two tertiary health facilities

Background Without interventions, the risk of mother-to-child transmission of HIV is up to 25% during pregnancy, labour and delivery with an additional risk of 5-20% during breastfeeding period, leading to an overall rate of up to 45%. Giving anti-retroviral therapy to the mother and anti-retroviral prophylaxis to the infant has been shown to significantly reduce the risk of HIV transmission through breastfeeding to less than 5%. According to the World Health Organization standard, the effectiveness of interventions towards prevention of mother-to-child transmission of HIV in any setting is measured by its HIV-free survival rate. We therefore carried out a study aimed at determining the HIV-free survival among breastfed infants in a resource poor setting in relation breastfeeding duration. Methods A cross-sectional retrospective study using routine clinical data and involving 365 HIV exposed infants followed up in two tertiary facilities in northern Uganda between 2014 and 2016. Data was analysed using Statistical Package for Social Scientists version 16 software package. Results Of the three hundred and sixty five (365) infants sampled for this study, 86.8% (317/365) were enrolled within the first 2 months of life, 12.1% (44/365) between 3-12 months of age and 1.1% (4/365) enrolled after 12 months of age. Almost all the infants (98.4%) were initiated on Nevirapine prophylaxis, 97.5% (356/365) of whom were initiated within 72 hours. The overall HIV-free survival rate in the current study was 93.7% (342/365), while 6.3% (23/365) were either HIV-infected (2.7%) or died (3.6%). The infants’ age at enrolment was the single most important factor significantly associated with HIV-free survival. The overall duration of breastfeeding did not significantly affect the HIV-free survival: 98.1% (304/310) for children breastfed > 12 months versus 100% (15/15) for those breastfed ≤12 months, though exclusive breastfeeding for at least 6 months was protective. Conclusions Adherence to current approaches to prevention of mother-to-child transmission of HIV with support to breastfeeding in low income countries can greatly enhance HIV-free survival for breastfed infants, and supports the current infant and young child feeding recommendations.


Introduction
It is estimated that about 1.8 million (1.3-2.4 million) children were living with HIV in 2017 who were infected by their HIV positive mothers during pregnancy, childbirth and breastfeeding (1), majority of who were in sub-Saharan African countries (2). Despite the fact that the incidence of paediatric HIV is falling as a result of an increase in effective methods to prevent mother-to-child transmission of HIV, an estimated 180,000 (110,000-260,000) new paediatric HIV-1 infections occurred in 2017, primarily through mother-to-child transmission, mainly in sub-Saharan Africa (1). In Uganda, a recent survey estimates that 95,000 children aged 0-14 years were living with HIV in 2017 (3), about 90% of them acquiring the infection through mother-to-child transmission (MTCT). An estimated 7,600 new paediatric HIV infections were registered in the country in 2017, with 3,800 AIDS-related deaths (1).
Without interventions for prevention of mother-to-child transmission of HIV (PMTCT), the rate of maternal-to-child transmission (MTCT) of HIV during pregnancy and delivery is estimated at 15-25%, and the additional risk through breastfeeding is estimated at 5-20% (4,5). With interventions for PMTCT, transmission reduces to <5% in the breastfeeding population and to <2% in the nonbreastfeeding population, making the prevention of mother to child transmission of HIV (PMTCT) a major public health approach of reducing the scourge (6). Infant feeding recommendations in the context of HIV have evolved over the last decades (7)(8)(9)(10)(11)(12) in response to emerging scientific evidence regarding the risk of postnatal transmission through breastfeeding (13)(14)(15)(16)(17)(18), and the long-term benefits of longer duration of breastfeeding for both maternal and child health outcomes, highlighting the relevance of supporting breastfeeding in high-and low-income settings alike (12). Earlier reports from Kenya and Botswana demonstrated efficacy of continued ART during breastfeeding in reducing the risk of transmission to 1.1% and 4.2% at 1 and 6 months, respectively (19,20). However, these studies showed some remaining risk of HIV transmission during breastfeeding, though small (20), and were based on shorter durations of breastfeeding (6-7 months).
In 2010, WHO recommended exclusive breastfeeding (EBF) for six months followed by complementary feeding (CF) and continued breastfeeding up to one year of age for infants of HIV positive mothers, under the cover of antiretroviral treatment (ART) to either the mother or the infant (10). Following this guidance, the Uganda ministry of health released guidelines on infant feeding recommending HIV positive mothers to exclusively breastfeed (EBF) for the first 6 months, introduce appropriate complementary feeds (CF) thereafter while being supported to adhere to ART and continue breastfeeding their infants for the first 12 months, breastfeeding should then only stop once a nutritionally adequate and safe diet can be provided (21,22). This was in an effort to balance the risk of acquiring HIV through breastfeeding and the higher risk of death from diarrhoea (23), malnutrition and pneumonia among non-breastfed infants in developing countries, where replacement feeding does not usually meet the WHO recommended criteria of being Acceptable, Feasible, Affordable, Sustainable and Safe (AFASS), thus having implications on child survival. In 2016, WHO updated the infant feeding guidelines stating that mothers living with HIV should breastfeed for at least 12 months and may continue breastfeeding for up to 24 months or beyond (similar to the general population) while being fully supported for ART adherence (12), albeit very low quality of evidence. However, these recommendations were based on limited evidence in terms of final HIV-free survival.
Uganda rolled out nationwide implementation of PMTCT option B+ approach in 2013, a policy whereby all HIV-infected pregnant and breastfeeding women are started on combination ART for life, regardless of their WHO clinical stage and/or CD4 cell count, following the World Health Organization (WHO) guidance (11). This was later expanded in a 2016 update of the national guidelines (24) following release of new WHO guidelines in 2015 (25), with recommendation for lifelong ART for all HIV-positive individuals, including pregnant and breastfeeding women, regardless of clinical stage and CD4 count.
Although there are promising results on program effectiveness for the ever evolving interventions for elimination of mother-to-child transmission of HIV, data on the long-term effectiveness of these interventions through the duration of breastfeeding and HIV-free survival in the study setting is scarce. Furthermore, most previous studies in Uganda were mainly confined to centres of excellence, involving urban dwellers. This study therefore aimed to determine the HIV-free survival rate and associated factors among babies born to HIV positive mothers in a post-conflict setting of northern Uganda, in light of the current efforts to elimination of mother-to-child transmission of HIV (eMTCT).

Study setting
This study was conducted at two tertiary referral hospitals in northern Uganda -Gulu regional referral hospital (GRRH), a public health facility and St. Mary's hospital Lacor, a private not-for-profit facility.
These are the two largest hospitals in the district of Gulu and the sub-region at large, serving the population from the entire northern region and beyond. Gulu district is located approximately 340 kilometers North of Uganda's capital city, Kampala, and is bordered by Pader district in the East, Amuru district in the west, Kitgum district in the North East, and Omoro district in the south. The population of Gulu district was estimated at 443,733 (26). Both hospitals run antiretroviral therapy (ART) clinics which provide HIV services to the people of Gulu district and beyond. Northern Uganda is a region recovering from over two decades of insurgency with a high burden of HIV (7.2%) compared to the national prevalence (6.2%) (3) and access to health services still remains a challenge in the district and the region as a whole. Over 37% of the population moves a distance of more than 5 km to reach health services. High levels of poverty and illiteracy, especially among women, is exacerbated by high prevalence of preventable diseases, including HIV/AIDS.

Study design
This was a cross-sectional retrospective study using routine clinical data carried out at two tertiary referral hospitals in Gulu, northern Uganda. The study population comprised all HIV exposed infants who were enrolled and followed up at the two ART facilities between 2014 and 2016. Infants who had complete data and documented final outcome were included in the study, while those who did not complete the recommended period of follow up and whose outcome were not known were excluded.
The minimum sample size was estimated using the formula for cross-sectional studies by Leslie Kish (1965) (27) and a standard error of 5%.

Determination of infants' HIV status
The infants final status was based on test results routinely performed using national procedures and algorithms for early infant diagnosis (EID) of HIV. The then Uganda national ART guidelines recommend that all infants born to HIV positive mothers be initiated on Nevirapine prophylaxis at birth according to their risk classifications. Infants were then routinely followed up at 6, 10, and 14 weeks, then monthly until 6 months of age and every 3 months until 18 months of age (24), with a final visit at 24 months. HIV status of infants were determined using national algorithms for early infant HIV diagnosis, and tested by polymerase chain reaction (PCR) for the detection of viral deoxyribonucleic acid (DNA) using dried blood samples (DBS) collected on filter paper at 4-6 weeks of age or at the earliest time thereafter (24 Sampling procedure and data collection Infants' medical records in the Early Infant Diagnosis (EID) register were retrospectively reviewed and a sampling frame for all the eligible infants was written down using codes. A simple random sampling method was then used. Simple random sampling is a probability sampling in which each individual has an equal chance of being selected. The code numbers were written down on pieces of paper which were folded and put in a container. These were properly mixed and then randomly picked with replacement until the calculated sample size was reached.Data was collected using checklists designed in English and contained variables which were selected based on the study objectives. These were filled in by the researchers in relation to the available recorded data in the source documents of both hospitals. Data was double checked for completeness and accuracy at the end of each day of data collection, kept in a safe place out of reach of unauthorized people.
HIV-free survival in this study was defined according to WHO as "an infant or young child born to a mother living with HIV who remains both HIV uninfected (confirmed negative HIV status) and also alive over a defined follow-up period. It is commonly reported at 18 months (as used in the current study) or 24 months of age" (12).

Data analysis
Data were entered, cleaned and analysed using Statistical Package for Social Scientists (SPSS) version 16 software package. In univariate analysis, categorical variables were summarized as proportions, while continuous variables as means and standard deviations (SD). HIV-free survival rate was calculated as the proportion of study infants who were alive with final HIV-negative status at 18 months, the denominator being all infants enrolled in the study. In the bi-variate analysis, the chisquare test (for categorical variables) and student t-test (for continuous variables) were used to test if the factors among infants with negative HIV status at 18 months were different from those among infants with outcomes other than negative HIV status at 18 months. Odds ratios with 95% confidence interval (CI) were used to measure the strength of association between the outcome and predictor variables. Multivariate analysis using logistic regression was used to determine the factors that were independently associated with HIV-free survival. P-value <0.05 was considered for statistical significance. Results were summarized in bar graphs, tables, and texts. Of the infants whose breastfeeding status could be ascertained, 94.5% (345/365) were exclusively breastfed. Overall, majority of the infants 95.4% (310/325) were breastfed for at least 12 months and more, and only 4.6% (15/325) were breastfed for a total duration of less than 12 months (Table 1).

HIV-free survival
The HIV-free survival rate, depicted by the proportion of infants discharged HIV-negative was 93.7% Factors associated with HIV-free survival The mean age of enrolment of the infants discharged negative (1.97, SD 1.66) was lower than that of the infants who either became HIV infected or died (2.99, SD 4.11) and this was statistically significant, P-value 0.014, χ2 = 6.057 (  There was essentially no difference in HIV-free survival between infants whose mothers had viral loads ˂1000cps/ml 94.9% (37/39) and those whose mothers had viral loads ≥1000cps/ml 100% (5/5), P = 1.000. Infants whose mothers' CD4 count was ≥500cells/mm3 were more likely to be discharged negative 94.7% (126/133) than those whose mothers' CD4 count was ˂500cells/mm3 91.6% (98/107), but this was not statistically significant, P = 0.331, OR = 1.653 (95%CI 0.595-4.595). Mothers who were on ART before conception were also more likely to have their infants discharged negative 94.1% (191/203) compared to those who started ART during pregnancy or after delivery 93.2% (151/162), but this was not statistically significant, P = 0.731, OR = 1.159 (Table 4). In the multivariate analysis using logistic regression, age at enrolment and treatment facility were the only independent predictors of HIV free survival. The only other factor that remained in the model even though not statistically significant was maternal adherence to ART (Table 5).

Discussion
This study was carried out to evaluate the HIV-free survival rate at 18 months and associated factors among HIV exposed infants enrolled and followed up between 2014 and 2016 in two tertiary health facilities in Gulu district, northern Uganda. Three hundred and sixty five (365) HIV exposed infants registered in the two facilities were sampled for this study using simple random sampling technique ( Figure 1). The mean age of the infants at enrolment was 2.03 (SD 1.91), and more than half of the infants 55.6% (203/365) were males. Majority, 86.8% (317/365) of the infants were enrolled early within the recommended 2 months of age, while 12.1% (44/365) were enrolled between 3-12 months of age and 1.1% (4/365) were enrolled after 12 months of age ( Figure 2). Timeliness of enrolment of infants in care as found in the current study shows a better trend compared to that reported by Hassan et al (2012) in rural Kenya where majority of the infants (68%) were enrolled after 2 months of age (28). This is an important aspect of early infant diagnosis and an indicator which shows the effectiveness of the PMTCT processes. It is likely to relate, albeit as a proxy, to good antenatal attendance and facility delivery under skilled attendance where mothers get opportunities to receive appropriate and effective counselling and interventions for PMTCT.

HIV-free survival
The overall HIV-free survival rate in the current study was 93.7% (342/365), while 6.3% ( to lay credence to the effectiveness of option B+ over the earlier approaches towards elimination of mother-to-child transmission (eMTCT) of HIV among breastfeeding population in developing countries.
The infants' age at enrolment was an important factor significantly associated with HIV free survival.
Infants discharged negative (HIV-free survival) were six times more likely to have a lower mean age at enrolment into care compared to those who were HIV-infected or died, and this was statistically significant, P = 0.014, χ 2 = 6.057. In the multivariate analysis (Table 5)  PCR tests had a 30% excess risk of mother-to-child transmission of HIV compared to those tested early (31). The mortality rate of 3.6% in the current study was relatively low but higher than the 1.

Duration of breastfeeding and HIV-free survival
Without intervention, the absolute risk of a mother transmitting HIV to her child during pregnancy and delivery is estimated at 15-25%, with an additional risk through breastfeeding estimated at 5-20%.
However, with interventions for prevention of mother to child transmission of HIV, the transmission rate reduces to less than 5% in the breastfeeding population. In this study 95.4% ( In the current study, mothers who were on ART prior to pregnancy were more likely to have an HIVfree child compared to those who initiated ART during pregnancy or after delivery, though this was also not statistically significant, P = 0.487, OR = 1.352. This result is similar to that reported in a Cameroonian study in which the 12-month HIV transmission rate where 51% of women were receiving ART prior to pregnancy was low at 1.2% (37), and also compares relatively well with that reported in a study by Hoffman et al (2010) where MTCT rate was lower in women who were on HAART before pregnancy (0.7%) compared to women who initiated HAART during pregnancy (5.7%), P = 0.01 (38).
Similarly, Oluwayemi et al (2015) in Nigeria reported a significantly lower risk of HIV transmission among babies whose mothers commenced HAART before pregnancy (3.4%) compared to those whose mothers initiated HAART during pregnancy (5.4%) and after delivery (28.6%) (39). The above results could be explained by the fact that one of the hypothesized benefits of lifelong ART is protection against HIV transmission in subsequent pregnancies, resulting from greater chances of virologic suppression, as also suggested by Gill et al (2017) in a Rwandan study where the substantial proportion of women on ART prior to pregnancy with suppressed viral load (VL) was thought to have contributed to the high effectiveness of PMTCT (29).
Likewise, available evidences also point to the fact that long duration of ART may be associated with high viral load (VL) or viral rebound postpartum which could be associated with increased risk of mother-to-child transmission (29,40,41), supporting the importance of VL monitoring during pregnancy and breastfeeding, and continued adherence counselling (29). In the current study, data on maternal viral load suppression and CD4 were very limited, making comparison of HIV-free survival based on these parameters unrealistic. This could probably explain the paradoxical finding seen in  (43). According to a report by Kahungu et al (2018) and Fisseha and Bereket (2016), infants who did not receive ART prophylaxis at birth had a five-fold risk of being HIV infected compared to those who received prophylaxis (44,45). Similarly in an Ethiopian study, infants who did not receive ARV prophylaxis were about seven times as likely to test HIV positive compared to ARV-protected infants (31). The above findings are not surprising and are in line with the already known benefits of early infant ARV prophylaxis for PMTCT.

Limitations of the Study
Being a retrospective study with use of secondary data, the information retrieved was incomplete for some infants. Some of the maternal potential factors for vertical transmission could not be exhaustively explored in the study because they were lacking. Our estimates may be an overestimate or underestimate since some infants with undocumented final outcome were not considered.

Conclusion
The overall HIV-free survival rate for breastfed infants in the study setting of 93.7% was below the WHO goal of >95% HIV-free rate in breastfeeding population (virtual elimination of HIV). Overall duration of breastfeeding did not significantly affect HIV-free survival, though exclusive breastfeeding for at least 6 months showed positive relationship with HIV-free survival.To improve on the outcome of HIV free survival, early enrolment of HIV exposed infants into comprehensive care should be emphasized and adhered to. In addition, adherence to the infant feeding guidelines should be encouraged with emphasis on exclusive breastfeeding for at least 6 months, while both mother and infant are on ART and ARV prophylaxis respectively.

List Of Abbreviations
WHO: World Health Organization; HIV: Human Immunodeficiency virus; ART: Antiretroviral Therapy; Declarations Ethics approval and consent to participate The study was approved by Gulu University Research and Ethics Committee (GUREC) and permission to conduct the study and access records was granted by the respective research committees of the respective hospitals. Being a retrospective study the patient consent was deemed unnecessary.
However, the identity of the participants was kept confidential by using participants' codes and not names.

Consent for publication
Participants' consent to publish this manuscript was not applicable since the manuscript does not contain any participants' personal data.

Availability of data and materials
The dataset supporting the conclusions of this article is available on request to the corresponding author. Age-Sex distribution of the study population. Final outcome disaggregation by duration of breastfeeding.