Dataset evaluating the effectiveness of the Konga model to address factors contributing to a low viral load suppression among children with HIV in Tanzania

Data were collected for a cluster-randomized clinical trial of the Konga community-based intervention using a validated questionnaire for children and caregivers. The raw and analyzed data include 82 participants with the following information: sociodemographic characteristics (caregiver's age, sex, and level of education, income, and caregiver's marital status) and clinical characteristics of the children (weight, CD4 cell count, and viral load at baseline and after 6 months of follow-up. The other data included in this dataset were weight, medication adherence, and opportunistic infections. Analysis of covariance (ANCOVA) was performed using the baseline VL. The outcome was viral load at the end of the intervention. Additionally, Omega squared (ω2) was used to calculate the effect size as an estimation of the strength of the intervention. These data will help researchers analyze data from similar studies and evaluate the effectiveness of community-based interventions for viral load suppression.


Value of the Data
• These data are useful because viral load (VL) measurement is essential for monitoring the effectiveness of antiretroviral therapy (ART) and is considered a surrogate marker for disease suppression.VL suppression (VLS) after early ART initiation is the primary treatment goal in children.This cluster-randomized clinical trial of children with a high VL ( > 10 0 0 copies/mL) tested mitigating measures to improve VL suppression.• These data will benefit researchers studying community participation in the care and treatment of children with HIV infection receiving ART.These data were used to test the effectiveness of the Konga model in improving VLS among children living with HIV and receiving ART.Despite their potential, community resources are often not engaged in providing support, and their roles remain undefined.Hence, these data will encourage service providers to engage with the community to promote retention and adherence to ART among children living with HIV.• These data will be valuable to researchers comparing similar studies regarding communitybased services as a contribution to reducing the HIV burden and mitigating its impact.• The data will be valuable for training researchers in the analysis of covariance for analyzing experimental and observational data.Analysis of covariance is underutilized in health research.
• The data provide a framework for future in-depth studies on using community-based organizations (e.g., Konga) to overcome the challenges in HIV care, treatment, and prevention.• The data provide a resource for leaders and policymakers to formulate effective policies for implementing HIV care and services.

Objective
Despite substantial antiretroviral therapy (ART) coverage in people living with HIV in Tanzania, viral load suppression (VLS) (viral load [VL] < 10 0 0 copies/mL) among HIV-positive children receiving ART remains unacceptably low.Of children enrolled in care and treatment centers (CTCs) and receiving ART, 82 % do not achieve VLS [2] .Through the National HIV/AIDS Control Program, the Government of Tanzania has tried to improve retention in care and ART adherence among people with HIV.These effort s have not been effective for VLS among children living with HIV receiving ART [3] .
Therefore, we identified the need for a sustainable intervention promoting retention in care and adherence to ART and addressing the lack of VLS among children living with HIV in Tanzania.Thus, the main goal of this study was to identify a sustainable intervention that promotes retention in care and adherence to ART in children living with HIV receiving care from care and treatment centers (CTCs) in mainland Tanzania.
The objective of collecting these data was to evaluate the effectiveness of a community-based intervention (the Konga model) for addressing the factors contributing to low VLS among children living with HIV in Tanzania.

Data Description
HIV depletes CD4 cells, making individuals with HIV susceptible to illnesses that a healthy immune system would otherwise prevent [4][5][6] .The data presented in this article were collected from children with high VLs > 6 months after ART initiation.The raw data are a set of demographic and clinical characteristics of children and are shared publicly in Figshare [7] .The data descriptor ( Table 1 ) is provided with the shared dataset.The name of the dataset in the Figshare is the Konga Trial Database.It includes sociodemographic characteristics of children and their caregivers, health service and clinical characteristics before and after the intervention.The data include age, VL at the start and end of the study, distance from the child's home to the Date on which the study ended health facility, sex, education level, source of income, and ART adherence at the start and end of the study ( Table 2 ).The adherence of children was measured using the pill count method, a standard method for measuring adherence in the study setting, and was performed by the ART nurse during the clinical visit.Adherence was recorded as "good adherence" if the child  did not miss pills for more than 1 day and as "poor adherence" if the child missed pills for ≥2 days.
The data analysis was descriptive and inferential ( Tables 2-4 ).

Experimental Design, Materials and Methods
The National Council of People Living with HIV (NACOPHA) is a nonprofit, nongovernmental, national grassroots-based organization of people living with HIV (PLHIV) in mainland Tanzania.Since its establishment, NACOPHA has embarked on coordinating the effort s of PLHIV through their district clusters known as Konga to address the needs of PLHIV.Hence, in this study, we used a community-based model called the Konga model to provide services to children.The study was conducted in the Simiyu Region.It was a cluster-randomized clinical trial with concurrent intervention and control groups.The clusters comprised health facilities that provided HIV care and treatment.The randomization unit was the HIV care and treatment facility, and the unit of analysis was the patient.Cluster randomization was chosen for practical reasons and to prevent contamination according to patient or nurse preferences.Health facilities with a care and treatment center (CTC) were classified into three levels: hospitals, health centers, and dispensaries.Health facilities at each level were randomly assigned to the intervention or control group using computer-generated random numbers.The trial was conducted in 45 CTC clinics that delivered HIV care and treatment within the four selected districts (15 CTCs provided the intervention, and 35 CTCs provided standard care).All children living with HIV receiving ART at the 45 study CTCs with an HIV viral load (VL) > 10 0 0 cells/mL were recruited.We used the analysis of covariance (ANCOVA) method (Dekkers, 2016) to calculate the required sample size.We assumed that the correlation between the VL results at the start and end of the study would be 0.6, the mean of the intervention group 0, the mean of the control group 0.5, the power 80 %, and the standard deviation of both groups 1.Using the Stata function (sampsi 0.5, power (0.8) st1(1) std2(1) pre (1) r01(0.6))method (ANCOVA)), the estimated sample size requirement was 82 (41 in the control group and 41 in the intervention group).Healthcare workers (i.e., ART nurses) identified and recruited children aged 2-14 years with a VL > 10 0 0 copies/mL.All children who met these conditions and were receiving ART from one of the study CTCs were eligible to be recruited.If the child was in an intervention cluster, personnel from the Konga visited the child's household to provide adherence counseling, psychosocial support, and screening for opportunistic infection.Standard operating procedures were developed and used to provide the Konga intervention.(The standard operating procedure for the Konga intervention is provided as a supplementary file.) The study data were collected using a structured validated questionnaire for children and caregivers at baseline and after 6 months of follow-up.The inclusion criterion was children aged 2-14 years with a viral load > 10 0 0 cells/mL attending a clinic.Stata software was used to analyze the data.An analysis of covariance (ANCOVA) was determined using the mean change in VL over the 6-month study period.The Omega-squared ( ω 2 ) value was used to calculate the effect size (an estimate of the strength of the treatment effect).F-tests and their corresponding p-values were used as measures of improvement.

Limitations
None.

Table 1
Attributes and description of socio-demographic and clinical data obtained during the trial (N = 82).
logpost_vlr Logarithm viral load at the end of the study post_adhere Adherence status at the end of the study.(Adherencewas measured using the pill count and recorded in the database as "good adherence" if the child did not miss pills for more than 1 day, and as "poor adherence" if the child missed pills for 2 or more days.hvlresultsHIV viral load results( continued on next page )

Table 1 (
continued )Attribute Description categoryThe stability category of the child at the start of the study.Patients were classified as stable or unstable: A stable client is a client who has been on ART for > 6 months, age above 5 years, with a viral load < 50 copies/mL, with good adherence, no history of opportunistic infection within the past 6 months, and not on a 3rd-line ART regimen.

Table 2
Sociodemographic characteristics of children and caregivers before and after the intervention, N = 82.

Table 3
Fitted model of the viral load suppression results at the end of the study.

Table 4
Measure of the effect size of the Konga intervention on viral load suppression.