Importance Death due to preventable medical error is a leading cause of death, with varying estimates of preventable death rates (14%–56% of total deaths based on national extrapolated estimates, 3%–11% based on single-centre estimates). Yet, how best to reduce preventable mortality in hospitals remains unknown.

Objective In this article, we detail lessons learnt from implementing a hospital-wide, automated, real-time, electronic mortality reporting system that relies on the opinions of front-line clinicians to identify opportunities for improvement. We also summarise data obtained regarding possible preventability, systems issues identified and addressed, and challenges with implementation. We outline our process of survey, evaluation, escalation and tracking of opportunities identified through the review process.

Methods We aggregated and analysed 7 years of review data regarding deaths, review responses categorised by ratings of possible preventability and inter-rater reliability of possible preventability. A qualitative analysis of reviews was performed to identify care delivery opportunities and institutional response.

Results Over the course of 7 years, 7856 inpatient deaths occurred, and 91% had at least one review completed. 5.2% were rated by front-line clinicians as potentially being preventable (likely or possibly), and this rate was consistent over time. However, there was only slight inter-rater agreement regarding potential preventability (Cohen’s kappa=0.185). Nevertheless, several major systems-level opportunities were identified that facilitated care delivery improvements, such as communication challenges, need for improved end-of-life care and interhospital transfer safety.

Conclusions Through implementation, we found that a hospital-wide mortality review process that elicits feedback from front-line providers is feasible, and provides valuable insights regarding potential preventable mortality and prioritising actionable opportunities for care delivery improvements.