Abstract
Objective: To describe the pattern of injuries of our intensive care unit (ICU) trauma patients, to identify the features related to a poor outcome, and to audit severe trauma care in our hospital using the trauma injury severity score (TRISS) method. Design: Prospective, cohort study of consecutive admissions to ICU. Setting: A six-bed general ICU in a 500-bed general hospital. Subjects: One hundred and ninety severe trauma patients admitted from January 1992 to December 1993 were considered eligible. Patients without the data necessary to calculate the TRISS probability of survival, or where the outcome was unknown, were excluded. One hundred and sixty-two patients were included in the study. Interventions: None. Measurements and main results: One hundred and ten (67.9%) patients suffered severe trauma (injury severity score [ISS] >15). Sixty-five (40.1%) had severe head injury (abbreviated injury scale [AIS] ≥3), and 54 (33.3%) had a Glasgow coma score (GCS) <9 during the first 24 hours. Sixty-one (37.6%) patients had severe chest trauma, and 28 (17.2%) severe abdominal injury; 52 (32.1%) patients had severe injuries to the pelvis and/or extremities. Non-survivors were older than survivors and had significantly higher APACHE II scores, ISS, and AIS of the head. Moreover, non-survivors had lower revised trauma score (RTS) and GCS in the first 24 hours after admission, and a shorter ICU stay. A GCS ≤⃒8 was a strong predictor of in-hospital death (OR 28.7; 95% CI 6.3-129.7), as was age >65 years (OR 4.6; 95% CI 1.78-12.2), and haemorrhagic shock (OR 4.22; 95% CI 1.6-11.0). There were 142 (87.6%) ICU survivors; 141 (87.0%) were discharged alive from the hospital. The TRISS method predicted a mean survival rate of 82.1±28.9% (95% CI 77.8-86.5%); the W value was 4.85, and Flora's Z statistic was 2.369 (p=0.017). The standardised Ws was 5.96 (95% CI 3.31-8.61), and Zs was 4.415 (p<0.0001). Conclusions: Our data show that hospital mortality is strongly associated with severe head injury, advanced age and haemorrhagic shock. Flora's Z statistic shows that the mortality in the trauma population considered was significantly lower than the death rate predicted by major trauma outcome study (MTOS) norms. This result is also confirmed by the standardised Zs statistic. Due to the relatively small number of patients with operable intracranial injuries included, some caution in extending these results to such an important group of trauma patients is advisable.