Morbidity and Mortality of HIV-Exposed Uninfected Infants in a Tertiary Referral Facility in Yaoundé, Cameroon

2.1. Study Design

This was a retrospective cohort study covering a period of three years from January 2017 to January 2020. The study took place at the Essos Hospital Center (EHC) in Yaoundé, Cameroon. EHC is a health facility with 315 beds and houses all the technical medical units, including pediatric, maternity, and neonatology services. The facility hosts an approved antiretroviral therapy treatment center. EHC was established as a reference center for the prevention of HIV transmission from mother to child (PMTCT) as previously described.20 The facility is at the center of a health network within the Djougolo Health District, the most populated district of Yaoundé. Data were collected from February 2020 to June 2020. A consecutive sampling was conducted.

2.2. Target Population

The study subjects were confirmed UIH aged more than 6 weeks to 24 months or less, regardless of the feeding mode and who had a follow-up of at least three documented consultations that occurred 24 months post-birth before January 2020.

An HIV-negative result was determined as follows: for subjects between the ages of 6 weeks and 9 months, one PCR result had to be negative. For subjects who were over the age of 9 to 18 months, two PCR and/or one HIV serology test had to be negative. Finally, for subjects over the age of 18 months, one test had to be confirmed as negative via HIV serology.

Any subject who did not have the minimum required consultations for follow-up or had a medical file deemed as ‘unusable’—characterized by incomplete or inconsistent medical records—was excluded from the study.

2.3. Procedure and Data Collection

After the identification of eligible cases of exposed children in the PMTCT register, their medical files were used for analysis. The records of the subjects who completed at least 3 routine follow-up consultations were retained. Their parents were called to visit the hospital with the needed health records for the collection of additional information. Analysis of morbidity was considered for any event other than the routine consultation appearing therein, namely consultation for acute problems or hospitalization. Deaths were reported either verbally by the mother or documented in the subject’s medical file post-hospitalization.

The primary outcomes were the proportion of subjects with clinical-infectious events, morbidity, and the proportion of mortality among the subjects. The recorded pathologies that dictated infection were the following: respiratory infections, malaria, gastroenteritis, cutaneous-mucous infections, conjunctivitis, and sepsis. Meanwhile, the death rate was calculated from the verbal notification of death by the mother or as recorded in the medical files.

2.4. Exclusion and Inclusion

All incomplete medical records were initially excluded due to incomplete data on final HIV status. Call reminders to parents of children who completed at least 3 clinical consultations of follow-up before 24 months were performed for final diagnosis of HIV status. After screening, the child was included in the study if the results were negative.

2.5. Independent Variables

The independent variables that were included in the study were the following: socio-demographic data, mother’s history, data related to pregnancy, childbirth, and antiretroviral treatment, socio-demographic data, infant history, and clinical and biological (HIV status) follow-up.

2.6. Statistical Analysis

The collected data were entered and recorded into Cspro 7.3 after the data cleaning and removal of duplicate data. All analyses were performed using R statistical 3.6.2. The socio-demographic characteristics of the mothers, the clinical history of the mothers and children, and the morbidity rates were determined using univariate distribution tables. The chi-square test of independence was used at the bivariate level to determine the factors or variables associated with infant morbidity. Multivariate analysis was used to determine the factors associated with infant morbidity by estimating the unadjusted and adjusted odds ratio (OR) and the associated 95% confidence interval. Confounders and independent variables selection for adjusted analysis were performed through univariable chi-square test selection. The significance level was 5%.

2.7. Ethical Considerations

Ethical clearance for the study was obtained from the Institutional Review Board (IRB) of the Essos Hospital Center under the reference number 2020/07/CE-CHE; the Hospital Directorate provided an administrative authorization; as per approval from the IRB on the consent procedure, a written proxy-informed consent was obtained from the respective parent/caregiver, including the use of medical records in the research; all data were fully anonymized and processed under strict confidentiality and privacy by using unique identifiers.

3.1. Characteristics of the Study Population

Out of 670 subjects who were recorded during the study period, approximately 240 mother/baby pairs for further analysis were attained (Figure 1). The median age of the mothers was 33 years, with the ages ranging from 19 to 48 years (interquartile range [IQR] 30-38). The profession found for most working mothers was the liberal profession (95/240 or 39.6%). Most of the mothers were cohabiting with a partner (87/240 or 36.2%). Concerning the level of study, a large proportion had a secondary level (155 or 64.6%), followed by a higher-level education (64 or 26.7%). According to the antiretroviral (ART) protocol, the tenofovir (TDF) + lamivudine (3TC) + efavirenz (EFV) regimen was the treatment taken by most mothers during their pregnancy (135/196 or 68.9%), as recommended in guidelines. The duration of ART before delivery was more than 4 weeks in most cases (154/161 or 95.7%). Most of the subjects were born at terms between 37 and 42 weeks (126/148 or 85.1%). Most of the subjects’ parameters had normal anthropometric parameters at birth.

Close

Figure 1

Flow chart of inclusion of UIH in a referral site in Yaoundé, Cameroon

Export to PPT

Exclusive breastfeeding was the predominant method of feeding up to 6 months (68.7%). Approximately 93.2% of infants had received antiretroviral prophylaxis, and 94.7% of subjects were up to date with their vaccines. Conversely, the rate of cotrimoxazole prophylaxis was relatively low at 60.6%. The mean duration of follow-up was 11 months with an overall follow-up duration of 2,662 months for the cohort (Table 1).

3.2. Morbidity

The recorded morbidity rate was 34.2% (82/240). Among the subjects who had clinically infectious events requiring consultations and or hospitalizations, 78% had contracted a single event, 19.5% had contracted two clinical events, and 2.5% had contracted three clinical events. The mean age at disease onset was 11 months. With an overall follow-up duration of 2,662 months, the incidence rate of infectious events was 3 per 100 child months of follow-up. The most frequent pathologies were acute respiratory infections (ARI; 60.79%), followed by malaria (17.65%), which were among the recorded events. (Table 2).

3.3. Mortality

During the study, only three death cases were recorded, which resulted in a mortality rate of 3.7% from the 82 subjects; two deaths were determined to be caused by malaria (66.7%), and one death was determined to be caused by pneumonia (33.3%). With three recorded deaths, the overall mortality rate was 1.25%, and the incidence was 1.1 per 1,000 child-months.

3.4. Factors Associated with Morbidity

Table 3 presents the unadjusted and adjusted analysis of factors that are associated with morbidity. A higher risk of contracting infectious diseases was observed in subjects whose mothers were diagnosed with HIV during pregnancy, exhibiting an odd ratio of 5.78 (2.14 -16.80; p = 0.001). The risk was also increased for those subjects following a maternal protocol based on azidothymidine (AZT), with an odd ratio of 3.83 (1.09 -14.45; p = 0.039). Finally, morbidity was 5 times higher in subjects whose follow-up period was less than 6 months (Table 4). In this study, factors associated with mortality, due to the insignificant number of documented deaths (only 3 cases) were not analyzed, as it could not allow a relevant statistical analysis.

4.1. Strengths and Limitations of the Study

This study has some limitations. Its monocentric character, the small sample size, and the high loss rate of follow-up affect the generalization and the externalization of the results. However, these weaknesses reflect the operational difficulties of observational studies in routine life settings. Despite these limitations, the potential applicability of our results relies on the fact that this reference site covers approximately 2% of HIV-exposed and uninfected children in Cameroon.