Enhanced health systems to improve uptake of early infant diagnosis of HIV in primary care in Malawi

Abstract

Background: The global target for HIV testing in HIV-exposed infants (HEI) at six weeks is 95% by 2025. Half of untreated children with HIV die by the age of two years. Low utilisation of maternal and infant health care services post-birth affects HIV testing for HEI. Research shows that effectiveness of interventions to promote early infant HIV diagnosis (EID) varies by setting, necessitating a context-driven understanding. Many EID strategies adopt evidence-based approaches but lack thorough stakeholder co-design to meet contextual needs. Therefore, I co-designed and evaluated a context-responsive enhanced health system (EEHS) intervention with public, community and private stakeholders in Blantyre, Malawi, a country with one of the highest HIV prevalences in the World, to improve the uptake of EID.

Methods: The Behaviour Change Wheel, Theoretical Domains Framework, and Consolidated Framework for Sustainability Constructs in Healthcare guided the overall study in one rural and one urban primary health facility in Blantyre District. Mixed-methods evaluation was applied in two main phases. In Phase I (November 2019 to March 2020), I assessed HEI enrolment coverage in HIV care and testing at six weeks via a retrospective data review, explored EID service implementation gaps through process mapping of mother-infant pairs, and investigated healthcare workers' (HCW) views on these gaps through group interviews (Study 1). In Phase II, managers of the programme for preventing mother-to-child transmission of HIV and I identified stakeholders comprising HCW, service users and non-governmental organisations using process mapping. In two qualitative workshops (August 2021 and December 2022), stakeholders analysed health system problems and co-designed interventions (Study 2). From October 2022 to June 2023, I evaluated the effectiveness of the co-designed intervention using a controlled before-and-after prospective data review of women with HEI (Study 3). We examined the co-designed intervention's implementation through structured observations with HCW and assessed its acceptability and sustainability via semi-structured interviews with women and HCW (Study 4).

Results: Enrolment of HEI in HIV care at birth and testing at 6 weeks was at 24% (39/163) and 52% (85/163) in 2018. Implementation gaps were observed during the process mapping of the eight mother-infant pairs, such as the failure of HCW to identify and enrol HEI in HIV care. Sixteen HCW attributed implementation gaps to process, capacity and system factors (Study 1). Forty-four stakeholders during workshops clarified that client and HCW factors complicated the identification of HEI, worsened by stigma. Patient services were hindered by resource and logistical issues, necessitating careful planning. The HCW coordination and accountability suffered from poor teamwork, misconduct, and negligence, underscoring the need for effective accountability measures and precise role coordination. Limited training and information sharing led to knowledge gaps. Unique identifiers, booking systems, strengthening leadership, data validation, care pathways, and facility-based training were the context-responsive enhanced health system interventions (Study 2). The co-designed intervention showed the potential to improve EID. The odds of enrolment in HIV care at birth increased 3-fold after the introduction of the intervention compared to before the intervention (odds ratio (OR) 3.04: 95% CI 1.06-10.06, p=0.048). The OR remained marginally significant for testing at six weeks (OR 3.74, 95% CI 1.10-17.23; p=0.052) (Study 3). Of 156 HCW, 151(97%) attended EEH intervention training. Knowledge of EID services and EEH intervention significantly improved (t (20) = 29.7, p= 0.001, (95% CI 26.8, 32.6), comparing before and after training. Implementation fidelity among the observed 15 HCW was 89.8%. The EEHS intervention was acceptable among the interviewed 23 HCW and 15 women with HEI due to positive progress, simplicity, and facilitators (Study 4).

Conclusion: Before the intervention, EID service uptake was low due to the complexities of the health system. A co-designed, context-appropriate intervention potentially improved EID uptake and was accepted by HCW and women with HEI. Findings indicate that client identification, appointment management, coordination, accountability, and facility training are essential for enhancing HIV EID services. Policymakers should update HIV guidelines to help facilities adapt to their contexts and address barriers in client identification, service integration, workforce capacity, and leadership. Evaluation using a randomised design is warranted based on these findings.

Date of Award	9 Dec 2025
Original language	English
Awarding Institution	Liverpool School of Tropical Medicine
Supervisor	Angela Obasi (Supervisor), Nicola Desmond (Supervisor), Augustine Choko (Supervisor) & Peter MacPherson (Supervisor)