Integration of EPI and paediatric HIV services for improved ART initiation in Zimbabwe

Publication Details

Prescott, M, Boeke, C, Gotora, T, Mafaune, HW, Motsi, W, Graves, J, Mangwiro, A and McCarthy, E, 2017. Integration of EPI and paediatric HIV services for improved ART initiation in Zimbabwe, 3ie Impact Evaluation Report 63. New Delhi: International Initiative for Impact Evaluation (3ie)

Link to Source
Marta Prescott, Caroline Boeke, Tendai Gotora, Haurovi William Mafaune, Wadzanai Motsi, Justin Graves, Alexio Mangwiro and Elizabeth McCarthy
Institutional affiliations
None specified
Grant-holding institution
None specified
Sub-Saharan Africa (includes East and West Africa)
Health Nutrition and Population
HIV and AIDS, Preventive Health and Health Behavior, Sexual Behavior
Gender analysis
HIV and AIDS, Preventive Health and Health Behavior, Sexual Behavior
Gender analysis
Equity Focus
None specified
Evaluation design
Randomised Control Trials (RCT)
3ie Series Report
3ie Funding Window
Integration of HIV services


The study uses a cluster-randomised control design to measure the impact of integrating Early Infant Diagnosis (EID) of HIV into the Expanded Programme of Immunisation (EPI) on the percentage of infants that initiate ART in Zimbabwe.


Over the last decade, Zimbabwe has developed a strong programme for EID of HIV. Nevertheless, challenges still exist in linking mother and infant pairs (MIPs) to HIV counselling and testing (HTC) as well as antiretroviral therapy (ART) with recent data showing that only 31 per cent of eligible children are currently receiving ART. To remedy this gap, CHAI seeks to integrate EID into the country’s high-performing EPI. Research has found that integrating other health programmes with EPI typically increases the coverage of the additional programme. The study addresses policy questions surrounding effective strategies to increase linkages to HTC and ART for all infants.

Research questions

The objective of the study was to evaluate the impact of integrating infant HIV testing and treatment referral into routine EPI services. Particularly on infant HIV testing, test result turnaround time, ART initiation time for positive infants, and on volume of EPI services.

The primary outcome of interest is to know the proportion of exposed infants that were tested for HIV by the end of 16 weeks.

Secondary outcomes of interest include:

  • proportion of infants tested for HIV by the end of 16 weeks that received results within 20 weeks,
  • average turnaround time between testing and receiving HIV results among those who were tested within 16 weeks,
  • proportion of positive infants initiated on ART by the end of 20 weeks,
  • average turnaround time between results received and initiation among those who were initiated within 20 weeks,
  • average age at initiation, among those who were initiated within 20 weeks,
  • proportion of children receiving Penta1 over the study period,
  • cost per additional HIV test by age 16 weeks,
  • cost per additional ART initiation by age 20 weeks.


Intervention design

Over the seven-month intervention period, MIPs in the intervention group present at the six-week EPI visit were offered HTC services and health education. They received their results during the ten week EPI visit. If the mother and/or infant tested positive, they were referred to ART services. Additionally, researchers provided the intervention facilities with the supplies, personnel training and mentoring, mobile health technology and sample transport needed, to ensure improvements in the integration of services.

Theory of change

The intervention’s theory of change builds upon several previous studies conducted in Sub-Saharan Africa which have shown that integrating other health programs with EPI usually increases the coverage of the additional programme. The theory assumes that by offering HIV tests to infants (and their mothers) who attend EPI clinics, more infants would be tested and diagnosed. Further, those who are HIV positive could then be linked to care and improving treatment initiation rates. The assumptions are that infants are underdiagnosed, while EPI clinics are well-attended, and that testing infants (and their parents) at EPI clinics will identify more HIV positive infants. Thus, this knowledge of status will motivate mothers to enrol their infants into treatment.

Evaluation design

The study used a cluster-randomised control trial design, implemented over a period of seven months after baseline data collection in July and August 2015.  Twenty-nine facilities were randomized according to the minimum-maximum t-statistic methodology to ensure balance on covariates during the pre-intervention period (Essajee et al. 2015); 15 were allocated to the intervention arm and 14 to the control. Data were collected from medical registers and the Logistics Management Information System (LMIS) database. Additional data were collected through interviews with health care workers and patient caregivers.

Main findings


There were no observed significant difference between the intervention arm and the control arm during the study period in rates of HIV testing by 16 weeks, receipt of results within 20 weeks, or treatment initiation within 20 weeks.

Cost efficiency analysis

The total cost of the intervention was over USD 43,000. This figure excludes the USD19,000 spent on personnel costs and the USD 17,500 spent on administrative costs.  Authors found that even if the study underestimated the impact of the intervention, the cost per additional HIV-identified infant and per HIV-positive infant treated was high compared to other interventions.

Implications for policy and practice

Although the intervention did not prove effective, integration of services in a higher prevalence environment may help alleviate some of the challenges associated with EID testing. These include addressing some of the laboratory problems, or providing point-of-care testing for EID, sample transport, reduction in costs and potentially increasing diagnosis rates. This in-turn would increase the effectiveness of an integrated program. Future interventions would benefit from incorporating more intensive training to gain buy-in from facility staff, focusing on facilities with adequate staffing, solving the laboratory challenges that result in delays, and, specifically targeting areas with a high prevalence of HIV-positive infants.

Implications for further research

This study did not conclusively determine that the integration of EPI and EID will not impact early infant testing and initiation in any setting. Additional work could look more specifically at whether EID coverage is really higher (and under-diagnosis lower) than previously estimated.  If modifications of the intervention show improved results, estimations of HIV infections averted could be added to the analysis.

Scroll to Top