Related Experiment Videos
Clinical governance and the vascular surgeon
C D Irvine1, D Grayson, R J Lusby
1Departments of Vascular Surgery and Centre for Education and Research on Ageing of the University of Sydney, Concord Repatriation and General Hospital, New South Wales, Australia.
The British Journal of Surgery
|June 10, 2000
Summary
Statistical modeling shows that large patient numbers are needed to identify substandard surgical outcomes, even for common vascular procedures. This challenges the feasibility of using event rates alone for clinical governance.
Area of Science:
- Vascular Surgery
- Health Services Research
- Biostatistics
Background:
- Auditing adverse outcomes is crucial for identifying substandard surgical results.
- This study investigates the feasibility of using statistical modeling for this purpose in vascular surgery.
Purpose of the Study:
- To determine the sample sizes required to detect substandard surgical performance using statistical models.
- To assess the feasibility of clinical governance based on morbidity and mortality rates.
Main Methods:
- Binomial statistical models were constructed to analyze adverse event rates for elective abdominal aortic aneurysm repair and carotid endarterectomy.
- Power calculations were performed to estimate the number of cases needed to identify surgeons or units with significantly higher morbidity rates.
Main Results:
- The mean annual procedures were 57 for abdominal aortic aneurysm repair and 70 for carotid endarterectomy, with a 4% adverse event rate for both.
- Detecting a surgeon with twice the base adverse event rate required 130 patients (for 6% base rate) or over 280 patients (for 3% base rate).
- Identifying widely variant surgical practice required fewer patients (21 procedures) when the morbidity rate was four times the base rate.
Conclusions:
- Statistical modeling for clinical governance requires clear assumptions about adverse event rates and confidence intervals.
- Significant patient data are necessary, even for common procedures, questioning the sole reliance on morbidity and mortality rates for governance.
- The feasibility of effective clinical governance solely based on event rates is questionable due to the large data requirements.