Examining the variation across acute trusts in patient delayed discharge

Delayed transfers of care, or delayed discharges, adversely affect patient care and increase costs to England's National Health Service. The main objective of this paper is to explain variation in the probability of delayed discharge from an acute trust and patient perspective. A novel approach is employed in using the Adult Inpatient Survey over the period 2007-2014. We use a two stage regression model to assess the impact of various patient, acute hospital trust, and regional characteristics on the probability of delayed discharge. In the first stage we model the patient-level probability of delayed discharge and estimate hospital trust-specific fixed-effects. Stage two includes multiple linear regressions to explain acute trust fixed effects from stage one by using acute trust characteristics and regional observable characteristics as explanatory variables. Results indicate the probability of delayed discharge varies among acute trusts and patients. Patient-mix complexity, staff skill-mix, size and scope of acute trust are among those factors affecting the trust-specific discharge efficiency.