Family Interviews Improve Health Service Recommendations in Mortality Review Process: A Mixed-Methods Assessment

Abstract

Background: Within the United States, although there is strong motivation for incorporating family interviews into Fetal Infant Mortality Reviews (FIMR) and important potential for expanding it to other types of reviews, there is limited evidence that family interviews make a difference in review team outcomes.

Objective: This study aims to assess the impact of FIMR family interviews identifying health service gaps and generating actionable improvements.

Design: Mixed methods design with quantitative case-control analysis and qualitative semi-structured interviews. Quantitative data collection was secondary analysis of FIMR administrative records, comparing outcomes with and without family interviews using Generalized Estimating Equation (GEE) and descriptive analysis of annual FIMR team evaluation responses. Qualitative data collection included audio-taping, transcribing and consensus coding of semi-structured interviews.

Setting and participants: The quantitative setting was Kalamazoo County, Michigan, FIMR reviews from 2015 to 2023, whereas the qualitative setting was virtual. Quantitative sample was the full population of completed FIMR case reviews (N = 158), and the 15 FIMR team case reviewers completing the 2023 annual evaluation. The qualitative sample was a purposive sample of 28 FIMR administrators across the United States.

Main outcome measures: Quantitative outcomes were review-identified contributing stressors and subsequent recommendations generated in the case review process. Qualitative outcomes were thematic experiences of family interview implementation and impact within mortality review process.

Results: The 53 cases (34%) with family interviews were similar to the 105 (66%) non-interview cases regarding multipleperinatal characteristics and were different regarding death type and manner. Controlling for these differences, GEE analysis found that family interviews were associated with increased identification of stressors, especially medical and socioeconomic, with a 2.6 increase in the number of stressors identified (aOR = 2.6, 95% CI: 1.5-4.7, p < 0.001). Family interviews were associated with a 40% increase in recommendations generated (aOR = 1.4, 95% CI: 1.0-2.0, p = 0.05), especially regarding patient-provider communication. Two-thirds of the FIMR team reported that family interviews were 'Very Impactful' in making meaningful system changes. Qualitatively, three primary themes emerged: Hurdles to Getting Interviews, Completing the Picture and Bringing the Human Connection.

Conclusions: This study drew upon multiple types of data, documenting the challenges obtaining interviews, while emphasizing their value identifying root causes, producing actionable healthcare service recommendations and motivating action.

Patient or public contribution: Locally, members of community advocate organizations representing birthing and parenting individuals and those representing people identifying as Black race contributed by actively refining the implementation of FIMR family interviews, including how interview information was utilized in FIMR case review meetings (study interventions). Mock FIMR case review meetings conducted as part of United Way public 'bus tours' and with undergraduate/graduate university students further refined recommendations coding and dissemination strategies. Leadership from public and private health service organizations shared the facilitators and barriers for applying FIMR recommendations to institutional policy and practice (shaping the development of recommendations coding as well as dissemination and interpretation of study findings). Nationally, a network of mortality review funders and administrators, in a 4-year learning series convened by the National Center for Fatality Prevention and Review, provided input in centreing family interviews within different types of mortality review processes (influencing study objectives, qualitative interview structure and results interpretation). Additionally, US policy leaders in maternal child health and health equity provided historical and current context for FIMR family interview funding and enabling legislation (informing the interpretation of study results and specifically referenced within the manuscript).

Keywords: community and patient engagement; health services improvement; population health.

Figure 1

Study parallel convergent mixed‐methods design.