The Geriatric Emergency Department Intervention model of care: a pragmatic trial.

Background: To evaluate a Geriatric Emergency Department Intervention (GEDI) model of service delivery for adults aged 70 years and older.

Methods: A pragmatic trial of the GEDI model using a pre-post design. GEDI is a nurse-led, physician-championed, Emergency Department (ED) intervention; developed to improve the care of frail older adults in the ED. The nurses had gerontology experience and education and provided targeted geriatric assessment and streamlining of care. The final format included 2.4 full time equivalent nurses working 7 days from 0700 h to 1730 h (1530 h at weekends). There were three implementations periods: pre-implementation (2012); a developmental phase from January 2013 to August 2015; and full implementation from September 2015 to August 2016. The outcomes measured were disposition (discharged home, admitted or died); ED length of stay; hospital length of stay; all cause in-hospital mortality within 28 days; time to ED re-presentation up to 28 days post-discharge; in-hospital costs. The setting was a tertiary hospital ED, with 385 beds, in Queensland, Australia. Approximately 53,000 patients presented to the ED annually with 20% aged 70 years and older. All patients over the age 70 who presented to the ED between January 2012 and August 2016 (n = 44,983) were included in the trial.

Results: Older persons who presented to the ED when the GEDI team were working had increased likelihoods of discharge (Hazard ratio (HR) = 1.19; 95% CI: 1.13-1.24) and reduced ED length of stay (HR = 1.42; 95% CI: 1.33-1.52) compared with those who presented when GEDI were not working. There was no increase in the risk of mortality (HR = 1.01; 95% CI = 0.23-4.43) or risk of same cause re-presentation to 28 days (HR = 1.21; 95% CI: 0.99-1.49). The GEDI service resulted in average cost savings per ED presentation of $35 [95% CI, $21, $49] and savings of $1469 [95% CI, $1105, $1834] per hospital admission.

Conclusions: Implementation of a nurse-led physician-championed model of ED care, focused on frail older adults, reduced ED length of stay, hospital admission and if admitted, hospital length of stay and cost, without increasing mortality or same cause re-presentation. These increases were sustained over time and after the initial implementation team had changed roles.

Trial Registration: Australian Clinical Trials Registration Number ACTRN12615001157561 - retrospectively registered on 29/10/2015. Data were retrieved via retrospective access to clinical information systems. First data access was on 1/7/2015.