The effect of peer review on mortality rates

Objective: Lowering of mortality rates in hospitals with mortality rates higher than accepted reference values for acute myocardial infarction (AMI), congestive heart failure (CHF), pneumonia, stroke, mechanical ventilation (MV) and colorectal surgery by using an external peer review process that identifies areas requiring rectification and implements protocols directed at improving these areas.

Design: Retrospective, observational, quality management study using administrative data to compare in-hospital mortality rates (pre and post an external peer review process that included adoption of improvement protocols) with reference values.

Setting: German general hospitals of a large, private group.

Participants: Hospitals with mortality rates higher than reference values.

Interventions: Peer review of medical records by experienced, outside physicians triggered by in-hospital mortality rates higher than expected. Inadequacies were identified, improvement protocols enforced and mortality rates subsequently re-examined.

Main outcome measures: Mortality rates 1 year before and 1 year after peer review and protocol use.

Results: For AMI, CHF, pneumonia, stroke, MV and colorectal surgery, the mortality rates 1 year post-peer review were significantly decreased as compared to pre-peer review mortality rates. The standardized mortality ratio for all of the above diagnoses was 1.45, 1 year before peer review, and 0.97, 1 year after peer review. The absolute risk reduction of 7.3% translates into 710 deaths in this population which could have been prevented.

Conclusions: Peer review triggered and conducted in the manner described here is associated with a significant lowering of in-hospital mortality rates in hospitals that previously had higher than expected mortality rates.