RESUMO
OBJECTIVE: To assess the ability of the SOFA score (Sequential Organ Failure Assessment) to describe the evolution of organ dysfunction/failure in trauma patients over time in intensive care units (ICU). DESIGN: Retrospective analysis of a prospectively collected database. SETTING: 40 ICUs in 16 countries. PATIENTS: All trauma patients admitted to the ICU in May 1995. MAIN OUTCOME MEASURES AND RESULTS: Incidence of dysfunction/failure of different organs during the first 10 days of stay and the relation between the dysfunction, outcome, and length of stay. Included in the SOFA study were 181 trauma patients (140 males and 41 females). The non-survivors were significantly older than the survivors (51 years+/-20 vs 38+/-16 years, p < 0.05) and had a higher global SOFA score on admission (8+/-4 vs 4+/-3, p < 0.05) and throughout the 10-day stay. On admission, the non-survivors had higher scores for respiratory ( > 3 in 47% of non-survivors vs 17% of survivors), cardiovascular ( > 3 in 24% of non-survivors vs 5.7% of survivors), and neurological systems ( > 4 in 41% of non-survivors vs 16% of survivors); although the trend was maintained over the whole study period, the differences were greater during the first 4-5 days. After the first 4 days, only respiratory dysfunction was significantly related to outcome. A higher SOFA score, admission to the ICU from the same hospital, and the presence of infection on admission were the three major variables associated with a longer length of stay in the ICU (additive regression coefficients: 0.85 days for each SOFA point, 4.4 for admission from the same hospital, 7.26 for infection on admission). CONCLUSIONS: The SOFA score can reliably describe organ dysfunction/ failure in trauma patients. Regular and repeated scoring may be helpful for identifying categories of patients at major risk of prolonged ICU stay or death.