Accuracy of administrative and trauma registry databases.

OBJECTIVE: Accurate data are needed to evaluate clinical outcomes, therapeutic modalities, and quality of care in trauma. Administrative data, usually used for billing, and trauma registries, have been used to perform these functions. This study compares data for trauma patients from administrative and trauma registry databases at a Level I trauma center.
METHODS: Data from patients injured in 1998 were obtained from both the trauma registry and administrative database. These International Classification of Diseases, Ninth Revision, Clinical Modification codes signify an admitting diagnosis of trauma. Patients from each database were "matched" by admission date, medical record number, age, and name. The two matched data sets were compared for accuracy in recording data. Chi-square analysis was used to compare groups.
RESULTS: There were 2,702 patients found in both databases. One hundred eighteen patients with significant trauma were recorded in the trauma registry, but not in the administrative database. Comparison of recorded data for "matched" patients is as follows. The underreporting of mechanism of injury, diagnoses, diagnostic interventions, surgical procedures, and complications was rampant throughout the administrative database. Statistical significance was seen in the comparison between the trauma registry and the administrative database with motor vehicle collisions (458 vs. 391), abdominal injuries (346 vs. 293), orthopedic injuries (1,243 vs. 1,101), and thoracic injuries (486 vs. 397). Diagnostic interventions such as diagnostic peritoneal lavage, head computed tomographic scans, and abdominal computed tomographic scans were all grossly underrecorded, with only 40%, 12%, and 9% captured by the administrative database, respectively. Analysis of surgical procedures revealed these same trends, with statistical significance seen in abdominal and orthopedic procedures. Complications such as acute respiratory distress syndrome and deep venous thrombosis showed statistically significant differences. Mortality was underreported in the administrative database, with 14 deaths omitted.
CONCLUSION: This study shows that administrative data have copious omissions of specific injuries, diagnostic and therapeutic interventions, as well as complications. The trauma registry recorded more of the diagnoses, diagnostics, procedures, and outcomes in the care of trauma patients. Trauma registries may be more useful than administrative databases in assessing quality of care and diagnostic and therapeutic interventions.