Mortality and readmission rates of patients discharged in-hours and out-of-hours from a British ICU over a 3-year period

Excess in-hospital mortality following out-of-hours ICU discharge has been reported worldwide. From preliminary data, we observed that magnitude of difference may be reduced when patients discharged for end-of-life care or organ donation are excluded. We speculated that these patients may be disproportionately discharged out-of-hours, biasing results. We now compare in-hospital mortality and ICU readmission rates following discharge in-hours and out-of-hours over 3 years, excluding discharges for organ donation or end-of-life care. This single-centre retrospective study includes patients discharged alive following ICU admission between 01/07/2015–31/07/2018, excluding readmissions and discharges for end-of-life care/organ donation. A multiple logistic regression model was fitted to estimate adjusted odds ratio of death following out-of-hours versus in-hours discharge. Characteristics and outcomes for both groups were compared. 4678 patients were included. Patients discharged out-of-hours were older (62 vs 59 years, p < 0.001), with greater APACHE II scores (15.7 vs 14.4, p < 0.001), length of ICU stay (3.25 vs 3.00 days, p = 0.01) and delays to ICU discharge (736 vs 489 min, p < 0.001). No difference was observed in mortality (4.6% vs 3.7%, p = 0.25) or readmission rate (4.1% vs 4.2%, p = 0.85). In the multiple logistic regression model out-of-hours discharge was not associated with in-hospital mortality (OR = 1.017, 95% CI 0.682–1.518, p = 0.93). Our findings present a possible explanation for reported excess mortality following out-of-hours ICU discharge, related to inclusion of organ donation and end-of-life care patients in data sets rather than standards of care delivered out-of-hours. We are not aware of any other studies investigating the influence of this group on reported post-ICU mortality rates.

Abstract licence: CC BY-NC-ND

18 more references, whose wording the paper’s licence keeps private.

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.