Descriptive Study Comparing Routine Hospital Administrative Data with the Vascular Society of Great Britain and Ireland's National Vascular Database

Objective: To compare patient volume and outcomes in vascular surgery between an administrative data set (Hospital Episode Statistics) and a clinical database (National Vascular Database).

Design: Descriptive study.

Methods: Volume of cases determined by age, sex, year and procedure and in-hospital mortality by procedure for both datasets for patients undergoing either repair of abdominal aortic aneurysm, carotid endarterectomy or infrainguinal bypass over a three year period between 1st April 2001 and 31st March 2004.

Results: There were 32,242 admissions with a mention of the three selected vascular procedures within the administrative data set compared to 8462 within the clinical database. For NHS trusts common to both datasets, there were twice as many procedures (16,923) recorded within the administrative dataset compared to the clinical database. Patient characteristics were similar across both databases. Further analysis limiting the administrative data to records attributed to consultants known to contribute to the clinical database showed much closer agreement with only 11% more repairs of abdominal aortic aneurysm recorded within the administrative dataset compared to the National Vascular Database.

Conclusions: There are significant differences in total numbers between HES and the NVD. If the National Vascular Database is to become a credible source of information on activity and outcomes for vascular surgery, there is a clear need to increase the number of contributing surgeons and to increase the completeness of data submitted. Further analysis at individual record level is needed to identify other reasons for discrepancies which could help to enhance data quality, both within Hospital Episode Statistics and within the National Vascular Database.