• Belinda J Gabbe, Pam M Simpson, Ann M Sutherland, Owen D Williamson, Rodney Judson, Thomas Kossmann, and Peter A Cameron.
• Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Victoria, Australia. belinda.gabbe@med.monash.edu.au
• Ann. Surg. 2008 May 1;247(5):854-9.

ObjectiveTrauma registries are integral to trauma systems, but reliance on mortality as the primary outcome measure remains a limitation. Some registries have included measures of discharge function, usually the modified Functional Independence Measure (FIM) or the Glasgow Outcome Scale (GOS), with the potential benefit being the ability to identify patients at risk for poor outcome. This study investigates the ability of these measures to predict longer term outcomes.MethodsTwo hundred forty-three blunt major trauma patients participated. Data were captured from the trauma registry and discharge function was assessed using the modified FIM, FIM, and GOS. At 6 months postinjury, the GOS, FIM, modified FIM, return to work/study, and other outcome measures were collected by telephone interview. Multivariate analyses were used to assess the performance of discharge functional measures as predictors of 6-month outcomes.ResultsTwo hundred thirty-six (97.1%) participants were followed at 6 months postinjury. Disability was prevalent at 6 months; 42% had not returned to work/study, and only 32% were categorized as a "good recovery" by the GOS. Neither the GOS nor modified FIM at discharge were independent predictors of 6-month outcomes, whereas the FIM score and the FIM motor score were independent predictors of functional recovery (adjusted odds ratios 0.97; 95% confidence intervals: 0.96-0.99) and return to work/study (adjusted odds ratios 1.03, 95% confidence intervals: 1.01-1.04), respectively.ConclusionsFor trauma registries to compare outcomes between regions and improvements over time, it is important that survivors with poor long-term outcomes are identified. Present measurement of discharge outcomes for trauma patients is inadequate for this purpose.

21 keywords...

hide…