The H1N1 pandemic has highlighted the importance of reliable and valid triage instruments. A Sequential Organ Failure Assessment score of >11 has been proposed to exclude patients from critical care resources quoting an associated mortality of >90%. We sought to assess the mortality associated with this Sequential Organ Failure Assessment threshold and the resource implications of such a triage protocol.
Three multisystem intensive care units.
Consecutive patients admitted from January 2003 to December 2008. Subsequently, a comparison H1N1 cohort was assembled consisting of all patients admitted in 2009 with confirmed H1N1.
Sequential Organ Failure Assessment was collected daily by use of an electronic bedside clinical information system (n = 10,204 patients, 69,913 patient days). Mean admission Acute Physiology and Chronic Health Evaluation was 19.1. 13.4% of the cohort (9% of total patient days) had an initial Sequential Organ Failure Assessment of >11. Mortality in patients with an initial Sequential Organ Failure Assessment score of >11 was 59% (95% confidence interval: 56%, 62%). The mortality associated with an initial Sequential Organ Failure Assessment >11 across diagnostic categories varied from 29% for poisoning to 67% for neurologic patients. Hospital mortality exceeded 90% only when initial Sequential Organ Failure Assessment was >20 (0.2% of patients). H1N1 patients were younger, had a longer intensive care unit length of stay, and more commonly had a respiratory admission diagnosis than the nonH1N1 cohort. Hospital mortality in H1N1 patients with an initial Sequential Organ Failure Assessment score of >11 was 31% (95% confidence interval: 5%, 56%).
A Sequential Organ Failure Assessment score of >11 was not associated with a hospital mortality of >90% at any time during intensive care unit stay. Only a small proportion of patients have the extreme initial Sequential Organ Failure Assessment values associated with a hospital mortality of >90% limiting the usefulness of Sequential Organ Failure Assessment as a triage instrument for pandemic planning. Application of a Sequential Organ Failure Assessment threshold of >11 to the recent H1N1 pandemic would have excluded patients with a markedly lower mortality than seen in a large regional cohort of intensive care unit patients.
From the Departments of Critical Care Medicine (HTS, CJD, DAZ), Community Health Sciences (HTS, CJD, PJB, DAZ), and Clinical Neurosciences (DAZ), University of Calgary, Calgary; and the Department of Critical Care Medicine (RS, HTS, CJD, PJB, DAZ), Alberta Health Services, Calgary, Alberta, Canada.
Supported, in part, by funding from a New Investigator Award from the Canadian Institutes of Health Research. Dr. Shahpori is an employee of Alberta Health Services. The remaining authors have not disclosed any potential conflicts of interest.
For information regarding this article, E-mail: firstname.lastname@example.org