Diversion of hospital admissions from the emergency department using an interprofessional team: a propensity score analysis.

Purpose: To examine if an ED interprofessional team ("ED1Team") could safely decrease hospital admissions among older persons.

Methods: This single-center, retrospective, propensity score matched study was performed at a single ED during a control (December 2/2018-March 31/2019) and intervention (December 2/2019-March 31/2020) period. The intervention was assessed by the ED1Team, which could include an occupational therapist, physiotherapist, and social worker. We compared admission rates between period in persons age ≥ 70 years. Next, we compared visits attended by the ED1Team to (a) control period visits, and (b) intervention period visits without ED1Team attendance.

Secondary outcomes: ED length-of-stay, 7-day subsequent hospital admission and mortality in discharged patients.

Results: There were 5496 and 4876 eligible ED visits during the control and intervention periods, respectively. In the latter group, 556 (11.4%) received ED1Team assessment. After matching, there was an absolute 2.3% (p = 0.07) reduction in the admission rate between control and intervention periods. After matching the 556 ED1Team attended visits to control period visits, and to intervention period visits without the intervention, admission rates decreased by 10.0% (p = 0.006) and 13.5% (p < 0.001), respectively. For discharged patients, median ED length-of-stay decreased by 1.0 h (p < 0.001) between control and intervention periods and increased by 2.3 h (p < 0.001) compared to intervention period without the intervention. For patients discharged by the ED1Team, subsequent readmissions after 7 days were slightly higher, but mortality was not significantly different.

Conclusion: ED1Team consultation was associated with a decreased hospital admission rate in older ED patients. It was associated with a slightly longer ED length-of-stay and subsequent early hospitalizations. Given that even a small increase in freed hospital beds would release some of the pressure on an overextended healthcare system, these results suggest that upscaling of the intervention might procure systems-wide benefits.