Abstract
INTRODUCTION: The Integrated Sustainable Childhood Pneumonia and Infectious diseases Reduction in Nigeria-Lagos project introduced pulse oximetry and oxygen as part of a 'stabilisation room' intervention in primary healthcare facilities (PHCs) in Lagos state, Nigeria. However, impact evaluation found minimal effects on treatment and referral practices for children with pneumonia, largely attributed to persistently low pulse oximetry use. We aimed to understand how the stabilisation room intervention worked (or did not work) to improve pulse oximetry practices, for whom, and in what contexts. METHODS: We conducted a preplanned mixed-methods realist evaluation. We used an initial programme theory to guide data collection and analysis, focusing on pulse oximetry adoption. We described oximetry adoption using quantitative data from the impact evaluation, then used training observations, monitoring data, quarterly site reports and qualitative interviews with purposefully selected healthcare workers (HCWs) to explore how pulse oximetry was perceived and adopted. We used inductive content analysis, iteratively triangulating emerging themes with process data to generate and test theories. RESULTS: We identified four inter-related themes explaining low adoption of pulse oximetry for children in this setting: (1) In contexts of frequent staff turnover and competing demands, oximetry training was desired, appreciated and may have increased HCW knowledge and skills, but largely failed to reach frontline HCWs. (2) Pulse oximetry was not perceived as a usual part of PHC practice. HCWs relied on external motivation to adopt a fundamentally new practice, seeing it as extra work and getting easily discouraged without external supervision or strong local leadership. (3) Without institutional norms and expectations regarding pulse oximetry, HCWs needed to be convinced they were capable (self-efficacy) and would typically only use pulse oximetry selectively. (4) Technical challenges and faulty equipment, in the absence of confident and accessible troubleshooters, were demotivating and discouraging to users. CONCLUSION: Implementation of pulse oximetry for children in primary care facilities requires clarity on the intended role of oximetry in primary care, clear normative guidance and close support and supervision during the early adoption phase.