A retrospective study to determine whether there is an association between maternal HIV infection and neonatal hypoxic-ischaemic encephalopathy in a tertiary hospital in Johannesburg, South Africa

Abstract

Hypoxic ischaemic encephalopathy (HIE) due to perinatal asphyxia is prevalent in South Africa. Almost one-third of mothers in South Africa are HIV-positive. This secondary analysis of an existing database sought to determine the association between maternal HIV status and neonatal HIE. There were 5846 neonates weighing more than 1800 g, and 824 (14.1%) had HIE. According to univariate analysis, more neonates without HIE than neonates with HIE were born to HIV-positive mothers (26.2% vs 22.6%, p 0.026). However, according to the binary logistic regression, the variables significantly associated with neonatal HIE were maternal age, vaginal delivery, maternal diabetes, and place of birth. There was no association between maternal HIV and neonatal HIE. This negative finding may be due to the high rate of prevention of mother-to-child transmission in the present study 95.2% of mothers received antiretroviral therapy during pregnancy. This study was a secondary analysis of an existing database from a single centre, so the results may not be generalizable. Factors such as maternal CD4 count, viral load, type and duration of antiretroviral therapy were not available for analysis.

Background

Over the past decade, the number of medicolegal claims in South Africa has increased dramatically [1]. Many of these claims are related to cerebral palsied (CP) children caused hypoxic ischaemic encephalopathy (HIE). The incidence of moderate to severe HIE in South Africa varies between 1.5 and 3.7 per 1000 live births, depending on the definition used [2, 3]. There are multiple causes of neonatal encephalopathy, including HIE [4]. Before attributing HIE as the cause of CP, as per the guidelines of the American College of Obstetricians and Gynaecologists, all other potential causes of neonatal encephalopathy must be excluded [5]. Congenital viral infections, including cytomegalovirus, herpes simplex virus, HIV and Zika virus can all cause neurological manifestations in infected infants [6, 7]. South Africa has the highest number of people living with HIV worldwide, with the highest burden in women of child-bearing age [8].

There is evidence in the literature that maternal HIV infection is associated with a worse perinatal outcome than in mothers without HIV infection [9]. A study from Florida, USA, between 1998 and 2007 reported that seizures and feeding difficulties (which are symptoms of neonatal encephalopathy) were more frequent in mothers with HIV infection [10].

In a 2007 review of the effect of HIV on perinatal outcomes in Mowbray Maternity Hospital, South Africa, Kennedy et al. reported that both perinatal mortality and neonatal encephalopathy were significantly greater in mothers who were HIV positive [11]. A study from Tshwane in 2010 showed a similar association, where maternal HIV infection was associated with a significantly increased risk of perinatal death due to preterm birth, infection and intrapartum asphyxia [12]. South Africa introduced its Prevention of Mother-to-Child Transmission (PMTCT) program in 2002 [5], yet studies exploring the program's impact on neonatal encephalopathy are limited. Moreover, a 2019 study from China found no association between maternal HIV infection and neonatal asphyxia [6]. Given these inconsistencies, the present study seeks to evaluate the relationship between maternal HIV infection and neonatal hypoxic-ischaemic encephalopathy (HIE), contributing to the understanding of the interplay between HIV and neonatal outcomes in South Africa.

Materials and methods

This study was a secondary analysis of an existing database of all neonatal admissions > 1500 g in the neonatal unit at Charlotte Maxeke Johannesburg Academic Hospital (CMJAH) from 1 January 2013 to 31 December 2021. Eligible subjects included all neonates above 1800 g birth weight who were admitted to the CMJAH neonatal unit, either after birth or referral. Neonates with missing information regarding HIE were excluded.

CMJAH is a tertiary referral centre offering high-risk obstetric care and neonatal intensive care, including therapeutic hypothermia, mechanical ventilation, and neonatal surgery. Maternal testing for HIV was part of routine antenatal care. HIV-positive mothers were offered antiretroviral therapy as part of PMTCT. Routine HIV PCR testing of HIV-exposed neonates was introduced in South Africa in 2015 [13].

Demographic, obstetric, and clinical information was collected from all patients by attending medical staff at discharge from the neonatal unit. The data was managed using Research Electronic Data Capture (REDCap) hosted by the University of the Witwatersrand [14]. Data was verified against the original hospital records during collection and entry into the computer database.

Hypoxic ischaemic encephalopathy (HIE) was diagnosed by the attending neonatal medical staff, using the TOBY criteria, which include a history compatible with intrapartum hypoxia, need for resuscitation, biochemical evidence of metabolic acidosis and evidence of encephalopathy. [15] The HIE and was graded 1 to 3 according to modified Sarnat and Sarnat [16]. The mode of delivery was classified as either any vaginal birth (assisted or normal vaginal delivery) or a caesarean section (both emergency and elective). The place of birth was categorized as inborn (at CMJAH) or outborn (including at another hospital, born before arrival and from the midwife obstetric unit). Survival to hospital discharge included neonates transferred to other healthcare facilities as step-downs.

Data analysis

Statistical analyses were conducted using IBM SPSS Statistics for Windows, Version 28.0 (IBM Corp., Armonk, NY, USA). Continuous variables, as their distributions were confirmed to be approximately normal, are presented as means and standard deviations (SDs). Categorical variables are expressed as frequencies and percentages. The study sample was divided into two groups: neonates diagnosed with hypoxic-ischemic encephalopathy (HIE) across all grades, and neonates without HIE. Independent t-tests were employed to compare continuous variables between the two groups, while chi-square tests were used to assess differences in categorical variables. A p value < 0.05 was considered indicative of statistical significance. Only valid cases were analysed for each variable (i.e. missing information for that variable was not included in the analysis).

To explore the relationship between potential risk factors and the presence of HIE, and to adjust for potential confounding variables, multivariable binary logistic regression analyses were performed. The forward likelihood ratio method was used for the selection of variables. All variables that exhibited a possible association with HIE in univariate analyses (those with a p value < 0.1) were included in the multivariable models. The analysis aimed to estimate adjusted odds ratios (ORs) with corresponding 95% confidence intervals (CIs) to assess the independent effect of each variable on the likelihood of HIE. Special attention was given to exploring the association between HIV exposure and the occurrence of HIE while adjusting for other relevant clinical and demographic factors that could potentially confound this relationship.

A secondary multivariable logistic regression analysis was performed to specifically examine the risk factors associated with more severe cases of HIE, defined as grade 2 or 3 HIE, compared to those without HIE (grade 0). The same forward likelihood ratio method was applied in this analysis to identify adjusted ORs for significant variables.

This comprehensive approach allowed for the exclusion of potential confounders and ensured that the impact of HIV exposure on HIE could be independently assessed, alongside other clinical predictors. Only valid cases were included, and missing data were excluded from all regression models to maintain the robustness of the findings.

Results

There were 6 280 neonates in the database, 434 of whom weighed 1800 g or less; thus, 5846 subjects were included in the final sample. There were 824 neonates with HIE (14.1%); 300 (36.4%) had HIE grade 1, 383 (46.5%) had HIE grade 2 and 141 (17.1%) had HIE grade 3. The majority of neonates were male (3460/5896; 58.6%). The mean birth weight was 3.01 kg (SD 0.63), the mean gestational age was 38.2 weeks (SD 2.19), and the mean maternal age was 28.5 years (SD 6.3). The median parity was 1 (range 1–8), and the median Apgar score was 9. There were 1503 HIV positive mothers (25.7%); PMTCT was given in 1432 (95.2%). Neonatal PCR was done in 1208/ 1503 cases (80.3%) of which, 21/1208 (1.7%) were positive. Neonates with HIE weighed significantly more than did those without HIE (3.1 kg (SD 0.5) vs 3.0 kg (SD 0.7) p < 0.001). The gestational age of neonates with HIE was also significantly greater than that of neonates without HIE (38.9 weeks (SD 1.8) vs 38.1 weeks (SD 2.2) p < 0.001). Mothers of neonates with HIE were significantly younger than were those without HIE (26.9 years (SD 6.2) vs 28.8 years (SD 6.3) p < 0.001. Maternal HIV, hypertension, diabetes and chorioamnionitis were more common in neonates without HIE, while neonates with HIE were more likely to be delivered vaginally (Table 1).

Table 1.

Characteristics of neonates weighing more than 1800 g admitted to the CMJAH neonatal unit between January 2013 and December 2021

Variables	Total	HIE	No HIE	p Value
Antenatal care	5083/5502 (92.4)	696/753 (92.4)	4387/4749 (92.4)	0.959
Maternal HIV	1503/5846 (25.7)	186/824 (22.6)	1317/5022 (26.2)	0.026
Maternal Syphilis	21/1287 (1.6)	10/800 (1.3)	11/487 (2.2)	0.087
Maternal hypertension	388/5464 (7.1)	37/752 (4.9)	351/4712 (7.4)	0.012
Maternal diabetes	266/5672 (4.7)	8/795 (1.0)	258/4877 (5.3)	< 0.001
Chorioamnionitis	140/5499 (2.5)	12/770 (1.6)	128/4729 (2.7)	0.012
Vaginal delivery	2627/5803 (45.3)	496/816 (60.8)	2131/4987 (42.7)	< 0.001
Inborn	4799/5792 (82.9)	595/816 (72.9)	4204/4976 (84.5)	< 0.001
Female	2385/5846 (48.5)	329/824 (39.9)	2056/5022 (40.9)	0.79
Survived to discharge	5637/ 5846 (96.4)	705/824 (85.6)	4932/5022 (98.2)	< 0.001

Adjusted odds ratios for these variables significantly associated with HIE and HIE grade 2/3 are shown in Table 2. Maternal HIV was not significantly associated with HIE (OR 1.03 95%CI 0.84–1.28; p = 0.72).

Table 2.

Adjusted odds ratios for those variables significantly associated with HIE and HIE grade 2/3 in neonates admitted to CMJAH between January 2013 and December 2021

Variables	Adjusted odds ratio	95% Confidence intervals	p value
HIE
Gestational age	1.14	1.09–1.18	< 0.001
Vaginal delivery	1.63	1.36–1.96	< 0.001
Inborn	1.35	1.10–1.72	0.018
Maternal age	0.97	0.95–0.98	< 0.001
Maternal diabetes	0.28	0.13–0.6	0.001
HIE grade 2/3
Gestational age	1.15	1.09–1.21	< 0.001
Vaginal delivery	1.75	1.42–2.17	< 0.001
Maternal age	0.95	0.94–0.97	< 0.001
Maternal diabetes	0.36	0.15–0.89	0.027

Discussion

South Africa bears one of the highest burdens of HIV globally and also experiences a significant incidence of perinatal asphyxia and neonatal hypoxic-ischemic encephalopathy (HIE). Despite this, our study found no association between maternal HIV infection and the occurrence or severity of neonatal HIE. This is in contrast to earlier studies from South Africa conducted in 2010 and 2012, which reported a significant increase in HIE among neonates born to HIV-positive mothers [11, 12].

The earlier findings may reflect the state of the Prevention of Mother-to-Child Transmission (PMTCT) program during those periods. Initially introduced in 2002, the PMTCT program in South Africa involved the administration of only a single dose of nevirapine to mothers during labour and a single dose to neonates after birth. However, substantial improvements were made in 2013 when all HIV-positive mothers became eligible for highly active antiretroviral therapy (HAART) regardless of their CD4 count. The program advanced further in 2015 when lifelong antiretroviral treatment became available to all HIV-positive mothers. In our study, most mothers were already receiving HAART as part of the updated PMTCT program, which could explain the lack of association between maternal HIV and neonatal HIE.

Additionally, our study did not observe any differences in the uptake of PMTCT or in the proportion of HIV-infected neonates between those with and without HIE, further suggesting that maternal HIV is not a driving factor for neonatal HIE in the era of effective maternal antiretroviral treatment.

Supporting our findings, a recent study conducted in Zimbabwe between 2018 and 2019 similarly found no association between maternal HIV infection and mortality in neonates with HIE. In that study, 10% of mothers of neonates with HIE were HIV-positive, which was lower than the 16.1% prevalence of HIV among antenatal clinic attendees in 2014 [17]. This aligns with the idea that maternal HIV, especially in well-managed cases with access to HAART, does not increase the risk of neonatal HIE.

Beyond the focus on maternal HIV, our study revealed other significant associations with neonatal HIE. Increased gestational age, vaginal delivery, and maternal age were all associated with a higher likelihood of neonatal HIE, while mothers with diabetes appeared to have a reduced risk. These findings contribute additional insights into the factors associated with neonatal HIE that merit further exploration.

In conclusion, the results of this study suggest that maternal HIV is not a significant risk factor for neonatal HIE in the context of current PMTCT and HAART interventions. This has important implications, particularly in medicolegal cases where the presence of maternal HIV might be considered a potential contributor to adverse neonatal outcomes. Our findings indicate that HIV, under effective treatment, should not be seen as a risk factor for HIE.

Limitations of the study

This study has several limitations. Firstly, it was conducted retrospectively in a single neonatal unit, which limits the generalizability of the findings to other settings. Additionally, while associations between variables were explored, causality could not be established due to the observational nature of the study. As a secondary analysis of an existing database, the scope of the analysis was constrained by the available data, and key information—such as CD4 count, viral load, and the specific antiretroviral drugs used—was unavailable for inclusion. Lastly, a potential source of bias is that the majority of mothers received antenatal care and participated in PMTCT programs, which may not reflect the broader population.

Author contributions

DEB conceptualized the study, collected and analysed the data, wrote the manuscript and drafts, reviewed the final manuscript, and submitted for publication. RMB assisted with data collection and analysis, reviewed the various drafts and final manuscript. TDR assisted with data collection, reviewed the various drafts and final manuscript.

Funding

Not applicable.

Data availability

Data will be made available upon reasonable request to the corresponding author.

Declarations

Ethics approval and consent to participate

Approval to conduct the study was obtained from the Human Research Ethics Committee of the University of the Witwatersrand (Certificate M230137). This study was a secondary analysis of an existing database, so the need for individual participant’s consent was waived.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.