Triaging patients to the ICU: a pilot study of factors influencing admission decisions and patient outcomes

To assess the appropriateness of ICU triage decisions.

Prospective descriptive single-center study.

Ten-bed, medical-surgical ICU in an acute-care 460-bed, tertiary care hospital.

All patients triaged for admission were entered prospectively.

None.

Age, underlying diseases, admission diagnoses, Mortality Probability Model (MPM0) score, information available to ICU physicians, and mortality were recorded. Of the 334 patients (96% medical), 145 (46.4%) were refused. Reasons for refusal were being too-sick-to-benefit (48, 14%) and too-well-to-benefit (93, 28%). Factors independently associated with refusal were patient location, ICU physician seniority, bed availability, patient age, underlying diseases, and disability. Hospital mortality was 23% and 27% for patients admitted to our ICU and other ICUs, respectively, and 7.5% and 60% for patients too well and too sick to benefit, respectively. In the multivariate Cox model, McCabe = 1 [hazard ratio (HR), 0.44 (95% CI, 0.24–0.77), P=0.001], living at home without help (HR, 0.440, 95% CI, 0.28–0.68, P=0.0003), and immunosuppression (HR, 1.91, 95% CI, 1.09–3.33, P=0.02) were independent predictors of hospital death. Neither later ICU admission nor refusal was associated with cohort survival. MPM0 was not associated with hospital mortality.

Refusal of ICU admission was related to the ability of the triaging physician to examine the patient, ICU physician seniority, patient age, underlying diseases, self-sufficiency, and number of beds available. Specific training of junior physicians in triaging might bring further improvements. Scores that are more accurate than the MPM0 are needed.