Coding errors: a comparative analysis of hospital and prospectively collected departmental data

Objectives To assess the accuracy of a hospital coding database at a busy tertiary referral urological unit. Methods Prospectively collected departmental coding data for all urological patients attending the Churchill Hospital, Oxford between 1 May 1999 and 30 April 2000 were compared with the coding data entered by hospital coding clerks on the Oxford Radcliffe NHS Trust database. Results There were significant discrepancies between the number of patients on the hospital and the departmental database (639 vs 1109). There were gross procedural coding errors in 74 cases. Conclusion Hospital‐coded data in this study were incomplete and inaccurate. This has important implications when considering the validity of hospital‐trust databases when used as a source for medical research, clinical audit and as a representation of a consultants' clinical workload.