Geriatric Screening Tools to Select Older Adults Susceptible for Direct Transfer From the Emergency Department to Subacute Intermediate-Care Hospitalization

Abstract

Objectives Early transfer to intermediate-care hospitals, low-tech but with geriatric expertise, represents an alternative to conventional acute hospitalization for selected older adults visiting emergency departments (EDs). We evaluated if simple screening tools predict discharge destination in patients included in this pathway.

Design, Setting, and Participants Cohort study, including patients transferred from ED to the intermediate-care hospital Parc Sanitari Pere Virgili, Barcelona, during 14 months (2012–2013) for exacerbated chronic diseases.

Measurements At admission, we collected demographics, comprehensive geriatric assessment, and 3 screening tools (Identification of Seniors at Risk [ISAR], SilverCode, and Walter indicator).

Outcome Discharge destination different from usual living situation (combined death and transfer to acute hospitals or long-term nursing care) versus return to previous situation (home or nursing home).

Results Of 265 patients (mean age ± SD = 85.3 ± 7.5, 69% women, 58% with acute respiratory infections, 38% with dementia), 80.8% returned to previous living situation after 14.1 ± 6.5 days (mean ± SD). In multivariable Cox proportional hazard models, ISAR >3 points (hazard ratio [HR] 2.06, 95% confidence interval [95% CI] 1.16–3.66) and >1 pressure ulcers (HR 2.09, 95% CI 1.11–3.93), but also continuous ISAR, and, in subanalyses, Walter indicator, increased the risk of negative outcomes. Using ROC curves, ISAR showed the best prediction among other variables, although predictive value was poor (AUC = 0.62 (0.53–0.71) for ISAR >3 and AUC = 0.65 (0.57–0.74) for continuous ISAR). ISAR and SilverCode showed fair prediction of acute hospital readmissions.

Conclusions Among geriatric screening tools, ISAR was independently associated with discharge destination in older adults transferred from ED to intermediate care. Predictive validity was poor. Further research on selection of candidates for alternatives to conventional hospitalization is needed.