Study designand Population:
Retrospective and prospective cohort study, which involved retrospective data collection and analysis of key dates from deliveries to ART initiations and initial CD4 cell and viraemia count outcomes of all HIV-1 proviral DNA positive infants who had their samples collected between 2015 and 2019. Those found initiated on ART were enrolled in cohorts of study group (HIV-1 infected infants delayed in diagnosis and ART initiation, delayed in diagnosis but not in ART initiation, and delayed in ART initiation but not in diagnosis) and control group (HIV-1 infected infants neither delayed in diagnosis nor in ART initiation). less than or equals to (≤) 8 weeks between delivery dates and DBS sample collection dates (not delayed in diagnosis); less than or equals ≤ 4 weeks between sample collection dates and ART initiation dates (not delayed ART initiation); greater than (>) 8 weeks between delivery dates and sample collection dates (delayed in diagnosis) and greater than (>) 4 weeks between sample collection dates to ART initiation dates (delayed in ART initiation); was dichotomized using the adopted recommended timeframe described above. Those on ART adherence from initiated date in the study and control group, a simple direct descriptive analysis of difference between dates ART was initiated and dates whole sample should be collected from each between January and March 2020 was conducted. This was done to identified the eligible (at least 6 months to at most 3 years duration on ART adherence) prospective study participants. Participant’s information sheet that entails the purpose, benefits, and procedures involved in the study together with inform consent form, was used to obtain consent from mothers/caregivers of the identified eligible participants and they were prospectively followed for their infant’s CD4 cell and viraemia count whole blood sample collection. They had the authority to withdraw at any time or stage during the study without any implication.
Study sites:
Fifteen (15) PMTCT-ART health facilities namely: SOS and Sibanor clinics; Serekunda, Banjulinding, Bajakunda and Janjangbureh major health centers; Brikama/Hand-on- Care, Soma, Essau and Basse district hospitals; Bansang, Bwiam, Serrekunda and Farafeni general hospital; and Edward Francis Small Teaching hospital; were accessed for the study .The sites were selected and accessed based on their registration of at least one HIV-1 proviral DNA positively tested infants from the central and 4 study sites laboratories (Brikama/Hands-on-Care, Soma, Bansang and Basse) capacitated for the services.
Sample size:
The sample size included all HIV-1 poviral DNA positive infants who had their samples collected from 2015 to 2019, whether death or alive.
Inclusion criteria and exclusion criteria:
The retrospective study included all HIV-1 poviral DNA positive infants whether death or alive and had their samples collected between 2015 and 2019, while the prospective study included only consented mothers/caregivers of the identified eligible infants. However, HIV-1 infected infants who doesn’t underwent proviral DNA screening detection test or did not have his/her samples collected from 2015 to 2019, whether death or alive were excluded in the entire study. Furthermore, unconsented mothers of eligible prospective study participants were also excluded.
Sampling frame and method:
A standardized and ethically approved data collection form was used to extract the list of all HIV-1 proviral DNA positive DBS PMTCT/year numbers, collected health facilities name (center name) and date tested from the 5 testing laboratory records described in the study sites. This was used to track them at their respective collected site PMTCT ART and Infants ARV registers (PA&IAR) to extract: delivery dates; date samples were collected; date send to testing laboratories; date results received from testing laboratories and survival status. Although, information on every step and date were usually found not routinely collected and available in their records despite, services providers were informed and trained. However, where this instance was encountered, they were documented not available (NA)/missing data thus, a challenge for this study. For those found initiated on ART, treatment folders were reviewed to extract: date ART was initiated; initial CD4 cell and viraemia count results; together with ART status. They were stratified and ranked into study and control groups using the WHO national adopted recommended timeframe described above. The study group included three strata ( 1. those delayed in diagnosis and ART initiation, 2. those not delayed in diagnosis but delayed in ART initiation and 3. those not delayed in ART initiation but delayed in diagnosis), while the control group included one strata ( those neither delayed in diagnosis nor in ART initiation). Among these groups, those identified eligible and were consented as described in the study design and population above, 2 to 3 ml whole blood samples were collected from them by trained laboratory staff for immediate CD4 cell count and processing the whole blood sample to plasma for storage at -20 °C, as equally done in their routine procedures. Becton Dickinson and company (BD) FasCount machine, available at all ART sites were used for CD4 cell counts. This was to ensure the consistent used of machine previously used for their initial CD4 cell counts. The -20 °C stored samples were used for viraemia count by the researcher under the supervision of the laboratory Head. Abbott Real-Time m2000sp/rt PCR machine available only at central laboratory was used, also to ensure consistent used of machine previously used for their initial viraemia counts. Despite sample were stored before viraemia counts was done, due to reasons described above. However, date whole blood samples were collected from each eligible study participant represented his/her CD4 cell and Viraemia count ART duration outcome from initiated date.
Data analysis:
Collected data were entered in excel spread sheet, then imported in STATA (StataCorp LP, College Station, Texas, USA) version 13 software and cleaned for analysis. However, to examine the normality of continuous variables, histogram and quantile-quantile plots, together with Shipiro-Wilk test was used to ascertain the type of test applicable for each variable. This was because the variable time interval (turnaround tine) between key dates was not normally distributed due to, information’s on every step and date was not routinely collected and available in their records as previously described. Therefore, median and interquartile (IQR) were reported for the time interval variables. Wilcoxon rank-sum test was used to compare the time intervals between deliveries and ART initiations, together with between sample collections and ART initiation, while difference in mean initial and mean prospective, CD4 cell and viraemia outcome respectively among the study and control groups were compared using Paired t-test. Turnaround time was dichotomized as: less than or equals to (≤) 8 weeks between delivery dates and DBS sample collected dates (not delayed in diagnosis); ≤ 4 weeks between DBS samples collected dates and ART initiation dates (not delayed in ART initiation); greater than (> 8) weeks between delivery dates and sample collected dates (delayed in diagnosis); and > 4 weeks between DBS samples collected dates and ART initiation dates (delayed in ART initiation). All the categorical variables were coded for the analysis and p-values less than 5% were considered statistically significant.