Australian Health Review Australian Health Review Society
Journal of the Australian Healthcare & Hospitals Association

Just Accepted

This article has been peer reviewed and accepted for publication. It is in production and has not been edited, so may differ from the final published form.

A Framework to Analyse and Act on Elevated Mortality Rates in Hospitals

Crystal Emang-Jok , Chantal Ferguson , Helen Gilchrist , Tanya Gawthorne , Matthew Skinner


Objective To create a tool enabling clinical teams to analyse hospital standardised mortality ratios (HSMR) to identify contributing factors and plan a measurable way to address these. Method Best practice models and the literature were analysed to create a framework to investigate elevated HSMR. The framework was tested in theory using a retrospective clinical audit. Setting A large tertiary hospital in Australia, 2015. Patients Group 1: In-hospital deaths with a diagnosis of infectious and parasitic diseases with a higher than expected HSMR. Group 2: Palliative patients (for validation only). Results An analytic framework was created based on best practice models. This incorporated key steps of data analysis (coding and case-mix review), investigating aggregated cases and creating driver diagrams for the key factors contributing to higher than expected mortality. The framework presented the methodology to investigate HSMR in a logical order. In the trial, the framework identified some potential contributing factors to the elevated HSMR in the high HSMR group and very few in the palliative care (control) group. The tool identified the main issues potentially counteracting care were escalation failure (60%) and limited supervision of junior staff (70%). The framework enables the creation of driver diagrams to identify measurable actions to address the key issues potentially contributing to mortality. Conclusion The framework can guide clinicians step by step to identify aspects of care that could potentially reduce HSMR. The trial confirmed that the framework was successful in identifying modifiable elements of health care potentially contributing towards excess deaths.

AH16235  Accepted 02 April 2017

© CSIRO 2017