October 1, 2016

Predicting Adverse Health Outcomes in Older Emergency Department Patients: The APOP Study

de Gelder, J., et al, Netherlands J Med 74(8):342, October 2016

METHODS: These Dutch authors performed a prospective study to develop and validate a prediction tool for adverse health outcomes in elderly patients seen in the ED. Included were medically stable patients (52% female) aged 70 or older, excluding those sent directly to cardiac catheterization or thrombolytic therapy. A questionnaire gathered baseline information on patient demographics, disease characteristics and severity, and preexisting geriatric comorbidities (functional and cognitive impairment). The tool was derived in one cohort (n=751; median age 78) in 2014 and validated in another cohort (n=881; median age 80) in 2015. The primary study outcome was the composite of functional decline or death at 90 days after the ED visit.

RESULTS: In the derivation cohort, 230 patients (30.6%) met the primary composite outcome and 71 (9.5%) died. The final prediction model included age, ED arrival by ambulance, number of medications, help needed for bathing or showering, hospital admission within six months, and history of dementia. The area under the curve (AUC) for the composite outcome was 0.73 (95% CI, 0.69-0.77), and the AUC for the separate outcome of 90-day mortality alone was 0.79 (95% CI, 0.73-0.85). In the validation cohort, 247 patients (28.0%) met the primary outcome and 84 (9.5%) died; AUC values were 0.71 (95% CI, 0.67-0.75) for the composite outcome and 0.67 for mortality (95% CI, 0.60-0.73). For patients deemed to be at highest risk, up to 69% experienced the composite outcome and 36% died.

CONCLUSIONS: This clinical tool appears to have good accuracy for predicting 90-day decline in function and mortality among elderly patients in the ED, but confirmation of these results is required. 26 references (J.de.Gelder@lumc.nl – no reprints)

See the Abstract in PubMed

Copyright 2017 by Emergency Medical Abstracts – All Rights Reserved 3/17 - #18

[ Feedback → ]