JMIR Publications

ABSTRACT

Background:

Overdose fatality reviews (OFR) are an important public health tool for developing local overdose prevention strategies by reviewing individual overdose cases. While this approach offers a rich, contextual understanding of drug overdose factors in communities, it examines a small number of cases, providing limited insight into broader population-level risk patterns. To complement OFRs, we developed a real-time dashboard that visualizes trends about five key “touchpoints” (ie, interactions with medical and justice services preceding overdose). We then trained local OFR teams to use this dashboard to identify prevention opportunities.

Objective:

This study examines the integration of population-level data into OFR practices, as well as tensions that emerge between OFRs’ traditional, case-driven review processes and the statistical, population-level analysis typical in public health. We analyze how and when the dashboard was used in meetings, the extent to which the data informed recommendations, and identify barriers that prevented broad uptake of this intervention.

Methods:

We observed 26 OFR meetings across 11 counties in Indiana over 10 months, during which teams conducted case reviews and developed recommendations. We documented instances of dashboard use as well as “missed opportunities”, in which relevant population-level data were available but not incorporated into the discussion. We also conducted semi structured interviews with OFR team members (n=7) to understand their perceptions of the dashboard, including its usefulness, usability, and adoption barriers.

Results:

Despite its intended role, the dashboard was rarely integrated into OFR meetings; it was used only 10 times, compared to 114 missed opportunities in which relevant data could have informed discussion. Interviews revealed that this limited uptake was not solely due to usability barriers but reflected a deeper tension between two distinct analytic approaches. OFR teams prioritized narrative-driven case reviews that are grounded in empathy, local knowledge, and lived experience. This approach seemed at odds with the population-level visualizations shown in the dashboard, which required statistical abstraction and interpretation. Other barriers identified included limited time and resources, staff turnover, and varying levels of data fluency that made it difficult for teams to confidently interpret the dashboard despite training.

Conclusions:

The results highlight a tension between case–based and data-driven approaches to overdose prevention. These approaches are grounded in different workflows, values, and motivations, making it challenging for OFR teams to keep their traditional, empathetic review practices while incorporating population-level trends. Our findings indicate the need for data tools that bridge these approaches, such as visualizations that connect aggregate patterns to individual cases. The results also underscore the need for additional support and training for teams, such as dedicated data specialists who interpret population-level trends and provide insights to augment team discussions and inform prevention strategies.