Validity of the Glasgow Aneurysm Score and the Hardman Index in predicting outcome after ruptured abdominal aortic aneurysm repair (Br J Surg 2005; 92: 570–573)

Sir As with most retrospective analyses, we acknowledge that complete data sets were not available for all patients. However, all case notes were retrieved and the Hardman Index calculated from the recorded data. Where variables were missing, these were scored as zero. This strategy would have underscored the Hardman Index rather than overestimated its severity, and has not compromised our conclusions. We did not criticise the work of Boyle and colleagues, but merely recognised that their series reported on data collated from two geographically disparate centres with statistically significant differences in patient comorbidity1. It is surprising that Boyle and Raptis challenge the relationship between hospital volume and outcome (Br J Surg 2005; 92: 1179). The evidence base for this is widely established for elective aortic aneurysm repair and is now suggestive for ruptured aneurysms too2. Furthermore, their own operative mortality of 32 per cent stems from tertiary centres1. Though our study used the more concise definition of myocardial ischaemia (as did Boyle et al.) rather than the original Hardman definition, the original criteria, regarding ST segment and T wave changes, were in fact applied in the present study. Boyle et al. challenge the number of patients with three or more Hardman criteria, yet in their own series 12 (12 per cent) of 100 patients had a score of 3 or more1. It is also noteworthy that Gatt and colleagues have recently also shown that the Hardman Index lacks validity3. These data support the conclusion that the Hardman Index and Glasgow Aneurysm Score lack precision as predictive tools in patients with a ruptured aneurysm.