Exploring Predictive Factors for Potentially Avoidable Emergency Department Transfers: Findings From the FINE Study

To assess the prevalence of potentially avoidable transfers (PAT) and identify factors associated with these transfers to emergency departments (EDs) among nursing home (NH) residents.

This is a secondary outcome analysis of the FINE study, a multicenter observational study collecting data on NH residents, NH settings, and contextual factors of ED transfers.

NHs in the former Midi-Pyrénées region of the southwest of France (n = 312); a total of 1037 NH residents who experienced ED transfers (n = 1017) between January 2016 and December 2016.

The analysis included resident baseline characteristics and NH and transfer decision-making characteristics. An expert group categorized the transfer status as either PAT or unavoidable. Multivariable analysis using a mixed logistic model, accounting for intra-NH correlation, was conducted to assess factors independently associated with PAT.

Among 1017 included transfers, 87.02% (n = 885) were identified as PAT and 12.98% (n = 132) unavoidable transfers. Multivariable analysis revealed that the following patient-related factors were associated with a likely high rate of PAT: usual behavior disturbances before transfer, including productive trouble (OR 2.04, 95% CI 1.25-3.33; P = .0044) and unusual symptom of falling during the week preceding the transfer (OR 4.55, 95% CI 1.76-11.82; P = .0019). On the other hand, distance between ED and NH (OR 0.98, 95% CI 0.97-0.998; P = .0231), NH staff trained in palliative care in the last 3 years (OR 0.52, 95% CI 0.29-0.95; P = .0324), the impossibility of direct hospitalization to an appropriate unit (OR 0.54, 95% CI 0.34-0.87; P = .0117), and the resident Charlson Comorbidity Index (OR 0.90, 95% CI 0.82-0.99; P = .0369) were associated with a lower probability of PAT.

Transfers from NHs to hospital EDs were frequently potentially avoidable, meaning that there are still significant opportunities to reduce PAT. Our findings may help to specifically identify interventions that should be targeted at both NH and resident levels.