Comparison of data extraction from standardized versus traditional narrative operative reports for database-related research and quality control

Abstract

The purpose of this study was to compare the completeness and reproducibility of data extracted from a standardized operative report (SOR) with the non-standardized operative report (NSOR). Between July and December 2003, operative data were collected from all laparoscopic cholecystectomy procedures performed at the Peter Lougheed Centre Hospital. A standardized format for dictating laparoscopic cholecystectomy operative reports was introduced on October 1, 2003. Non-standardized operative reports dictated in the first 3 months of the study period were compared with SORs dictated in the final 3 months. Two physicians independently extracted data from each operative report into a surgical database. During the study period, 221 cholecystectomy reports were analyzed (119 SOR and 102 NSOR). Completeness of data extraction for identifying variables (eg, patient name, age, and date of procedure) was similar in the 2 types of reports. However, most other operative and perioperative details were more completely reported in the SOR (95% to 100%) when compared to the NSOR (14% to 100% complete). Furthermore, interobserver agreement between 2 independent data extractors was better for the SOR than the NSOR (0.9972 vs 0.9809, P < .0001). Standardized operative reports result in more complete and reliably interpretable operative data compared with NSORs.