Ineffectiveness of the trauma score and the CRAMS scale for accurately triaging patients to trauma centers.

Both the Trauma Score (TS) and the CRAMS scale have been advocated for field triage of trauma victims to trauma centers. To determine which scale best serves this purpose, both scores were calculated by computer for 5,130 trauma cases in our statewide computer file of ambulance rescue runs from 1981 through 1983. A total of 3,231 patients (63%) were treated in the ED and released; 1,857 (36.2%) were admitted to the hospital (202 of whom were transferred directly from the ED to the operating room [OR] for immediate operation); and 42 (0.8%) were DOA or died in the ED. Major trauma was defined as a TS of 12 or less or a CRAMS scale of 8 or less. Both scales correctly identified as major trauma all patients who were DOA or died in the ED. The TS identified as major trauma more patients admitted to the hospital than did the CRAMS scale (33% vs 21%; P less than .0001). The TS identified as major trauma more patients transferred from the ED to the OR than did the CRAMS scale (37% vs 21%; P less than .002). Both scales failed to identify as major trauma almost two out of three patients brought directly from the ED to the OR. The paramedics', or emergency medical technicians' qualitative judgements about injury severity, reflected in their coding injuries as life- or limb-threatening, was almost as good (more than 90% sensitive and specific) as either score (100% sensitive and specific) in identifying patients who died, and was better in identifying patients sent from the ED to the OR.(ABSTRACT TRUNCATED AT 250 WORDS)