BACKGROUND:
Group antenatal care (gANC) combines health assessments with interactive learning and group discussion, with the aim of improving care engagement and effectiveness. A realist review conducted by Mehay et al. identified seven theories of effect: social support, peer learning, active participation in health, health education, satisfaction with care, health professional development and wellbeing, and empowerment.

OBJECTIVES:
To understand mechanisms by which gANC functions to achieve positive care experiences and outcomes across different country and care settings.

METHODS:
Analysis of mixed-methods data in implementation case studies within seven countries. Sources included: interviews and/or focus groups with service providers and participants, observations of gANC training and care, surveys of participants and service data. Using each country as a case, data were first coded inductively and then synthesised, at country level, then across all settings, using Mehay et al’s review findings as a framework.

RESULTS:
The mechanisms theorised in the literature were found to be applicable across widely varied settings but were refined and adjusted. ‘Continuity of carer and participants’, and ‘time for care’ were identified as key underpinning mechanisms, while empowerment was found to be an overarching effect. Understanding of how health professional satisfaction and development and elements such as self-checking and an interactive approach with this model of care may support empowerment was enhanced.

CONCLUSIONS:
The analysis provided a more developed understanding of how gANC may influence care experiences and outcomes, across a diversity of contexts. In cases where specific mechanisms did not function, this was related to implementation challenges influencing fidelity of approach.

KEY MESSAGE:
The mechanisms through which group care supports positive care experiences and outcomes are applicable across a wide range of country and health-system settings but may be undermined if a key aspect of the model such as continuity or interactive approach is not applied. Poster session 2 (Group A)