Designing a theory-based digital audit and feedback dashboard for maternal and child health care providers: A qualitative study of preferences and needs in Palestine

Abstract

Audit and feedback to clinical care providers about their performance are important mechanisms for self-reflection and quality improvement in health care provision. Feedback intervention theory (FIT) and the model of actionable feedback (MAF) provide specific recommendations to maximize the impact of audit and feedback, although the evidence in support of these theories has not focused on clinical providers in low-resource settings, such as in health care systems in low- and middle-income countries. The aim off this study is to develop design requirements for the creation of a digital audit and feedback dashboard for maternal health care providers in Palestine, adapted to their needs using a theory-based framework that combines FIT and MAF with a human-centered distributed information design (HCDID) approach. The study relied on FIT and MAF to develop an early version of a digital dashboard, and guidance and questionnaires for in-depth interviews and focus groups discussions to understand the current supervision practice and explore supervision needs. An HCDID approach was used to collect contextual information through in-depth interviews with 18 health care providers; three focus group discussions with 20 healthcare providers and supervisors; observations at three public healthcare clinics; and document and process review. The findings from the interviews, observations and document review were analyzed through the lens of audit and feedback theory to develop the final recommendations for a quality improvement dashboard. This study captures specific observations for the Palestine setting, while also contributing to the overall understanding of audit and feedback, as well as the emerging field of digital health in the global south. Specifically, the recommendations of FIT and MAF were useful in generating a tailored audit and feedback system Palestine, when combined with a contextually-sensitive HCDID process. This approach may have applicability for similar settings.