An observational cohort study of triage for critical care provision during pandemic influenza: 'clipboard physicians' or 'evidenced based medicine'?

We assessed the impact of a United Kingdom government-recommended triage process, designed to guide the decision to admit patients to intensive care during an influenza pandemic, on patients in a teaching hospital intensive care unit. We found that applying the triage criteria to a current case-mix would result in 116 of the 255 patients (46%) admitted during the study period being denied intensive care treatment they would have otherwise received, of which 45 (39%) survived to hospital discharge. In turn, 69% of those categorised as too ill to warrant admission according to the criteria survived. The sensitivity and specificity of the triage category at ICU admission predicting mortality was 0.29 and 0.84, respectively. If the need for intensive care beds is estimated to be 275 patients per week, the triage criteria would not exclude enough patients to prevent the need for further rationing. We conclude that the proposed triage tool failed adequately to prioritise patients who would benefit from intensive care.