Timeliness of Operation as Audit Filters in Trauma Care

ABSTRACT

While Joint Commission on Accreditation of Healthcare Organizations(JCAHO) and American College of Surgeon(ACS) have identified certain audit filters in trauma care, there are few studies to substantiate the value of these audit filters. Some researchers found that audit filters qualifiers were significantly associated with adverse outcomes, however, others were unable to reproduce such association. It is also necessary to test their validity and applicability in Korea. The purpose of this present study was to validate two trauma audit filters proposed by the JCAHO and the ACS, through the analysis of the relationship between timeliness of operation and risk-adjusted mortality. Among trauma audit filters, timeliness of operation in epidural or subdural hematoma(EDH/SDH) and intraabdominal injury were selected. By stratified random cluster sampling, 19 emergency medical centers (EMCs) were selected from 30 EMCs and all patients who received craniotomy or laparotomy in 1996 were evaluated in each hospital. Six medical records administrators reviewed medical records of 463 patients with EDH/SDH and of 508 patients with intraabdominal injury retrospectively. In other to adjust risk of mortality, timeliness of operation, age, Revised Trauma Score(RTS), ICD-9CM based ICISS, and experiences of transfer were included in logistic regression model. In the logistic regression models of all EDH/SDH or intraabdominal injury patients, timeliness of operation was not significant predictor of mortality. However, if patients who have been operated later than 12 hours were excluded from the statistical model, timeliness of operation showed significant or marginally significant relationship with mortality in the following situations; craniotomy > 4 hours in EDH(OR=30.46, p=0.032), craniotomy > 8 hours in SDH(OR=6.50, p=0.020), laparotomy > 2 hours in shock patients(OR=9.26, p=0.055). In addition to timeliness of operation, RTS and ICISS were significant variables in every logistic regression model, and experience of transfer and types of EMC were significant or marginally significant only in EDH. Timeliness of operation as audit filters for trauma care could not be applied to all cases. Early operations seem to improve clinical outcome only in the patients for whom emergent craniotomy or laparotomy were indicated. It could be interpreted as a phenomenon of 'confounding by indication'. Additional studies to establish more objective eligibility criteria for these audit filiters are needed.