MORTALITY RATES IN LEAGUE TABLES – FACT OR FICTION?

Abstract

It is likely that league tables will to some extent determine hospital finance in the future. The major indicator used in league table calculations in orthopaedics is mortality rates following surgery. Therefore, our study audited the accuracy of mortality data.

A previous audit of our department by an external audit company was found to show an apparent excess mortality rate, due to the company’s failure to distinguish between true operations and certain procedures, i.e. urethral catheterisation. We were concerned that these flawed results may find their way into the publicised tables of the Department of Health (DH). We thus audited deaths in 2000/1 and compared the results with DH data.

DH league table figures combine the mortality numbers for all surgical specialities. Our analysis was based on DH criteria [www.doh.gov.uk/performancer-atings/2001), death within 30 days of operation, following non-elective admission and excluded certain procedures. PAS was used to identify deaths and all case notes were reviewed.

From review of the notes, the criteria for post-operative death were fulfilled by 54/131 deaths (41%). By speciality, these included 14/33 deaths in orthopaedics, general surgery (25/73) and neurosurgery (15/25). The DH identified 64 post-operative deaths in this period. DH calculations were applied to compare our postoperative mortality results (54 deaths) with those of the DH (64 deaths). Although there was no significant difference between our observed death rate and the DH’s, using our results the hospital’s ranking improved from twelfth to sixth place in 42 small acute hospitals.

The observed mortality rate in our hospital is very close to that published by the DH and the national average. From the results of our study, we are confident that the flawed data from the external company did not enter the system and distort the DH’s league tables.

Therefore, hospitals should not wast money on audits by external companies.