Study design
This was a retrospective follow-up study conducted between January 2018 and April 2022 among HIV-positive children and adolescents who were attending care and treatment clinics (CTCs) in Tabora, Tanzania.
Study site
The study was done in the Tabora region, which has seven councils, namely Tabora Municipal, Sikonge, Igunga, Kaliua, Uyui Urambo, and Nzega. The 2012 housing census reported 2291, 623 in the region. There is only one HIV VL testing Laboratory in the region, which is located at Kitete Regional Referral Hospital (RRH). We included all seven hospitals in the region, namely Kitete RRH, Igunga DH; Sikonge designated DH, Urambo DH, Nkinga mission hospital, Kaliua mission hospital, and Ndala hospital.
Study population
The study population comprised children (0–9 years) and adolescents (10–19 years) living with HIV who had been registered between January 2018 and April 2022 and were on ART for ≥ 6 months.
Inclusion criteria
All HIV-positive children and adolescents who were on ART for at ≥ 6 months were eligible for the study.
Exclusion criteria
Those who were lost to follow-up, died, transferred out and those whose files were not accessible were excluded from the study analysis.
Sampling method and sample size estimation
Sampling method
A simple random method was used to attain study participants. A list of CTCs in the Tabora region at the hospital level from seven districts was used. In a district with more than one hospital, a balloting method was applied; whereby the hospitals were assigned numbers, and the numbers were written on a different piece of paper and folded and then placed in a box and shaken thoroughly. Then, the hospital name from a selected piece of paper was included in the study. The seven selected CTCs are Kitete RRH, Igunga DH, Sikonge designated DH, Urambo DH, Nkinga mission hospital, Kaliua mission hospital, and Ndala hospital. The selected clinics represent geographical diversity, and significant numbers of HIV-positive patients attended CTCs in the region. In every selected CTC, HIV clients (aged < 20 years) and on ART for ≥ 6 months were arranged and a unique identification number based on the registration was assigned (Fig. 1).
Sample size estimation
We calculated sample size based on data from an African cohort study conducted in Uganda, Kenya, Tanzania, and Nigeria showing the overall proportion of non-suppression to be 9% and an adherence level of 77% (Kiweewa et al. 2019). We used a two-sided 95% confidence interval, and a minimum detectable alternative of ± 5% was used to calculate the power of the study, which was set at 99.9% via Open-Epi Version 3.01. The resulting sample size was 378.
The sample size distribution per CTC facility
Figure 1 shows a scheme showing clients, including children and adolescents who were on ART during the study period. In brief, a total of 101,698 patients living with HIV were ever enrolled in seven CTCs in the Tabora region, of whom 2884 were children and adolescents aged less than 20 years. Of the 2884 children, 725 children and adolescents were enrolled in ART. Of whom, 142 were not on ARV for at least six months, 23 files were not found, 27 were LTFU, 31 died and 123 were transferred out. All the remaining (378) client files were reviewed and included in the study analysis (Fig. 1). Their distribution per facility was as follows: Kitete RRH (n = 67, 17.7%), Igunga DH (n = 61, 16.1%), Sikonge designated DH (n = 40, 10.6%), Urambo DH (n = 49, 12.9%), Nkinga mission hospital (n = 50, 13.2%), Kaliua mission hospital (n = 54, 14.3%), and Ndala hospital (n = 54, 15.1).
Data abstraction
Data were abstracted from the facility CTC database regarding the client’s socio-demographic and clinical, ART drug regimen, and virological and immunological characteristics.
Dependent variable
The dependent variable was VL non-suppression. A binary outcome was categorized as non-suppression status or suppression status.
Independent variables
The independent variables were as follows: socio-demographic features (age, sex, education level, and HIV disclosure status). The regimen at the initiation of ART was categorized as (lopinavir/ritonavir (LPV/r)-based, dolutegravir (DTG)-based, efavirenz (EFV)-based, and nevirapine (NVP)-based (NACP 2019b). Adherence to ART treatment was categorized as 'good' and ‘poor’ according to the percentage of drug dosage calculated from the total monthly dose of ART drugs as follows: good (95% or ≤ 3 doses missed per month), or poor (≥ 95% or ≥ 4 doses missed per month) (NACP, 2019b). WHO clinical stage (categories; I, II, III, or IV)(WHO 2016). CD4 count was categorized as ‘low’ or ‘high’, low CD4 count was categorized as; CD4 count ≤ 350 cells/mm3 for children above 5 years, and CD4% < 25% of total lymphocytes for children less than 5 years, and high CD4 count was categorized as; CD4 count > 350 cells/mm3 for children above 5 years, and CD4% > 25% of total lymphocytes for children less than 5 years (NACP 2019b). Duration on ART of receiving HIV care and treatment in months, this was defined as: the time from initiation of antiretroviral (ARV) regimen (first or second line) to the date of data collection, death, loss to follow-up or transfer out, the categories of ART duration was; ≤ 2 years and > 2 years. Current ARV regimen (first-, second-line or third-line), and TB/HIV co-infection at the initiation of ART; ‘yes’ or ‘no’. Completion of Isoniazid Preventive Therapy (IPT); categorized as ‘yes’ or ‘no’. HIV disclosure status; categorized as ‘Full disclosure’, ‘Complete non-disclosure’ and ‘Partial disclosure’. Full disclosure refers to providing the child or adolescent with the name of the diagnosis and full information and knowledge about HIV. ‘Complete non-disclosure refers to maintaining complete secrecy around diagnoses, and the child or adolescent are not told the truth about their illness Partial disclosure is telling the child the truth, but not the whole truth, usually withholding the name of HIV (EGPAF 2016).
Pre-testing
Tool pre-testing was conducted on ten patient records with information from the CTC analytics database and medical records of children and adolescents HIV-positive and on ART for ≥ 6 months, registered in ART clinics in the study period (January 2018–April 2022) at Kitete RRH.
Data processing and analysis
Data cleaning and analysis were performed using the STATA version 15 package (StataCorp. 2017. Stata Statistical Software: Release 15. College Station, TX: StataCorp LLC). Descriptive statistics were summarized using frequency and proportions for categorical variables, whereas continuous variables were summarized with a measure of central tendency with corresponding measures of dispersion. Incident rates were estimated, and a log-rank test was used to describe the survival experiences of categorical variables for the events of viral non-suppression. Bivariate analysis was done for all covariates. Hazard ratios (HRs) were estimated as a degree of association between viral non-suppression and client features, via a Cox proportional hazards regression. Variables with p-values less than 0.20 in the bivariate analysis were included in the multivariable analysis, whereas variables with a p-value ≤ 0.05 were measured as statistically significant determinants of viral non-suppression, with a 95% CI.