Monitoring performance of cardiac surgery: the SCTS governance programme

Hickey, Graeme ORCID: 0000-0002-4989-0054, Cosgriff, Rebecca and Bridgewater, Ben (2013) Monitoring performance of cardiac surgery: the SCTS governance programme. 2013, Brighton, UK.

Text Presentation - 457.pdf - Unspecified Download (2MB)

Abstract

OBJECTIVES The SCTS have published mortality rates for cardiac surgery by named hospital since 2001 and by named surgeon since 2005. This clinical governance programme has been associated with improved mortality outcomes despite increasing numbers of high-risk patients undergoing surgery. We describe the process of analysing the 2008-11 in-hospital mortality data, including the management, data processing and statistical framework. METHODS All SCTS data from April-2008 to March-2011 were extracted from the central cardiac audit database. The data were cleaned and summaries, including missing data, returned to units for validation. Afterwards, the final extract was cleaned and procedures classed as emergency, salvage, transplantation, trauma or primary-VAD removed. The primary outcome was in-hospital mortality. A contemporary recalibration of the logistic EuroSCORE model was developed for risk-adjustment. Funnel plots were used to detect outlier units with 95% and 99% two-sided confidence limits. One-sided 95% confidence limits corrected for multiple comparisons and multiplicative adjustment for overdispersion are examined. RESULTS A total of 106,982 records were included from 40 hospitals and 301 consultants. The mean mortality was 2.7% in all cardiac surgery. The recalibrated model was well calibrated (Hosmer-Lemeshow test P=0.56) and had good discrimination (AUC=0.78). Funnel plots were generated for each procedure group comparing 1) hospitals and 2) consultants. CONCLUSIONS The detection of ‘outlier’ healthcare providers is a challenging exercise and requires careful planning and analysis. By combing clinical and statistical expertise with robust methodology, we can reduce the chances of falsely classifying a unit as an ‘outlier’.

Item Type:	Presentation Material
Depositing User:	Symplectic Admin
Date Deposited:	27 May 2015 08:17
Last Modified:	17 Dec 2022 01:26
URI:	https://livrepository.liverpool.ac.uk/id/eprint/2012053