ABSTRACT
BACKGROUND: Simple measures of acute physiologic compromise, functional status and comorbidity may help clinicians to make decisions relating to clinical care and resource utilisation. AIMS: To explore the usefulness of common assessment tools in predicting outcomes of (i) death or intensive care unit (ICU) admission and (ii) length of hospital stay at a busy tertiary hospital in Singapore. METHODS: Three hundred and ninety-eight consecutive admissions to two general medicine teams were prospectively assessed during 2 months in 2011. Patients were followed until discharge or transfer to ICU/high dependency unit (HDU). Data collected included routine demographic data, final diagnosis, comorbid conditions including a weighted prognostic comorbidity index (the updated Charlson index) and the modified Early Warning Score (MEWS) at presentation to the emergency department. The admission modified Barthel Index was recorded for patients aged 65 and over. Death and total length of hospital stay were recorded in all cases. RESULTS: Of 398 patients, 16 (4 %) died or were transferred to ICU and 99 (25%) stayed for more than 7 days. Medical early warning (MEW) scores of ≥5 were significantly associated with death or ICU admission (hazard ratio 5.50, 95% confidence interval 1.77-17.07, P = 0.003). There was no independent association between this outcome and the Charlson score or admission Barthel Index. Excess length of stay was associated with a modified Barthel Index ≤17 and altered mental status at presentation. CONCLUSION: Among unselected general medical patients, MEW scores of ≥5 were significantly associated with death or ICU admissions and only functional status and altered mental status were independent predictors of excess length of stay.
