Analysis of potential clinical management issues in unreported mortality data from an Australian health service.

Background: We have previously reported incomplete data submission to the Victorian Audit of Surgical Mortality (VASM) by a large health service. We have further examined the source health service clinical data to assess whether any clinical management issues (CMI) occurred and should have been reported.

Methods: The previous study identified 46 deaths that should have been reported to VASM. The hospital records of these patients were further analysed. Data recorded included the patient's age, gender, admission type and clinical course. Any potential clinical management issues were recorded and classified using the VASM definitions (area of consideration or concern, adverse event).

Results: Median age of the deceased patients was 72 (range 17-94), with 17 (37%) being female. Patients were under the care of nine different specialties with general surgery being the most common (18/46). Only four (8.7%) of the cases were electively admitted. 17 (37%) patients had at least one CMI with 10 (21.7%) classified as adverse events. Most deaths were not considered preventable.

Conclusion: The proportion of CMI in the unreported deaths was consistent with the previously reported VASM data, however current findings show a high percentage of adverse events. The underreporting may be due to inexperienced medical staff or coders, poor quality notes or confusion about what should be reported. These findings reinforce the importance of data collection and reporting at the health service level, and a number of important lessons and opportunities to improve patient safety have been lost.