Use of Hospital Episode Statistics to investigate abdominal aortic aneurysm surgery

Background A coding framework was evaluated to study patients undergoing open surgical replacement of an abdominal aortic aneurysm (AAA) in the English Hospital Episode Statistics (HES) database. The objective was to create groups of patients who are homogeneous with respect to diagnosis, prognosis and treatment. Methods The frequency and consistency of potentially relevant diagnosis (International Classification of Diseases, 10th revision) and procedure (Office of Population Censuses and Surveys Classification, 4th revision) codes were assessed in patients admitted to English National Health Service hospitals between April 2003 and March 2008. Administrative codes were compared with diagnosis and procedure codes to check that patients who had undergone emergency surgery for a ruptured AAA were admitted as an emergency. Results Of 20 290 patients undergoing AAA replacement, 19 250 (94·9 per cent) had a consistent diagnosis (unruptured or ruptured AAA); 79·3 per cent of patients with an emergency replacement were coded as having a ruptured AAA and 95·7 per cent of those with a non-emergency replacement as having an unruptured AAA. Of patients who had undergone emergency replacement of a ruptured AAA, 93·3 per cent were coded as having been admitted as an emergency. Conclusion Coding consistency was high. The proposed framework could define homogeneous groups by combining diagnosis, procedure and administrative codes. It also allows an assessment of potential miscoding at national and hospital level.