Tool uses out-of-hospital factors to predict critical illness during hospitalization
TUESDAY, Aug. 17 (HealthDay News) -- A prediction score based on out-of-hospital factors may be useful for stratifying non-trauma patients and predicting who will develop critical illness during hospitalization, according to research published in the Aug. 18 issue of the Journal of the American Medical Association.
Christopher W. Seymour, M.D., of the University of Washington in Seattle, and colleagues studied data on 144,913 non-trauma, non-cardiac arrest patients transported by emergency medical services to hospitals. They linked records with complete data to hospital discharge data and randomized the patients into development and validation cohorts, 87,266 and 57,647, respectively, to determine out-of-hospital predictors of clinical illness and to assess the performance of a score for prediction of critical illness development during hospitalization.
The researchers found that critical illnesses occurred in 5 percent of both cohorts during hospitalization; predictors included greater age, lower systolic blood pressure, abnormal respiratory rate, lower Glasgow Coma Scale score, lower pulse oximetry, and residence at a nursing home during out-of-hospital care. The prediction-score model had a sensitivity of 0.22, a specificity of 0.98, a positive likelihood ratio of 9.8, and a negative likelihood ratio of 0.80, with a score threshold of four or higher (range, zero to eight).
"In a population-based cohort, the score on a prediction rule using out-of-hospital factors was significantly associated with the development of critical illness during hospitalization. This score requires external validation in an independent population," the authors write.
Abstract
Full Text (subscription or payment may be required)
Editorial (subscription or payment may be required)