Objectives: To investigate a method to identify and understand patterns of adverse events by utilising secondary data analysis; to identify the types of complications associated with elective surgery; to identify any specific "adverse event-prone" elective procedures; and to consider the implications of these patterns for hospital patient safety programs.
Setting: Public hospitals in Victoria.
Design: Secondary analysis of data on acute hospital admissions for elective surgery in the period 1 July 2000 to 30 June 2001, for non-obstetric patients older than 15 years (n = 177 533).
Main outcome measures: Estimated rates of adverse events for the most commonly performed elective surgery procedures; frequency of the most commonly recorded adverse event types.
Results: Of all admissions, 15.5% had at least one complication of care. The most frequent first-recorded single complication code, in 9.6% of cases with a complication, was "Haemorrhage and haematoma complicating a procedure". The most common adverse event categories were cardiac and circulatory complications (23%), symptomatic complications (18%), and surgical and drug-related complications (17%). The procedure blocks most frequently associated with an adverse event were coronary artery bypass surgery (67%), colectomy (52%), hip and knee arthroplasty (42% and 36%, respectively), and hysterectomy (20%). The types of complications associated with the four most adverse event-prone procedures were cardiac arrhythmias, surgical adverse events (haemorrhage or laceration), intestinal obstruction, anaemia, and symptomatic complications.
Conclusion: Routinely collected data are valuable in obtaining information on complication types associated with elective surgery. International Classification of Diseases codes and surgical procedure "blocks" allow very sophisticated investigation of types of complications and differences in complication rates for different surgical approaches. The usefulness of such data relies on good documentation in the medical record, thorough coding and periodic data audit. The limitations of the method described here include the lack of follow-up after discharge, variable coding standards between institutions and over time (potentially distorting information on rates), lack of information on the causative factors for some adverse events, and a limited capacity to support investigation of particular cases. Hospitals should consider monitoring complication rates for individual elective procedures or blocks of similar procedures, and comparing adverse event rates over time and with peer hospitals as an integral part of their patient safety programs.