Article Text
Abstract
Background We sought to evaluate the ability of the Identification of Seniors At Risk (ISAR) tool to differentiate between older adult patients having a poor outcome within 30 days of emergency department (ED) care and those who do not. We compare prognostic accuracy of subjective versus objective risk factors.
Methods 202 community-dwelling patients age 65 years and older presenting to two EDs were prospectively enrolled. Participants completed the six-question ISAR and objective testing (cognition, ambulation, vision). We reviewed electronic medical records for current medications, hospitalisations in the past six months, ED disposition, length of hospital stay, subsequent ED visits or inpatient admissions or death at 30 days. Participants were given a point for each risk factor present; subjective and objective risk factors were scored separately. We tested ability of individual risk factors and scores to predict a composite outcome of subsequent ED visit, postdischarge hospitalisation or death by day 30 after the index ED visit. We computed receiver operating curve area under the curves (AUC) to determine tool discrimination.
Results 23% of participants had a poor 30-day outcome. The optimum subjective ISAR cut-off score for screening was ≥2, which was present in 84% of participants, had a sensitivity of 91% and specificity of 19%. Using the subjective ISAR tool, the AUC was 0.66. The optimum objective ISAR-related risk cut-off score for screening was ≥3, which was present in 82% of participants, had a sensitivity of 87% and specificity of 40%. Using the objective ISAR-related tool, the AUC was 0.69.
Conclusions The self-reported ISAR tool did not discriminate well between older adults with or without 30-day hospital revisit or death. An optimum score of ≥2 would identify many older adults at no apparent increased risk of poor outcomes at 30 days. Using objective ISAR-related risk factors did not improve overall discrimination.
- patient support
- hospitalisations
- geriatrics