Background While transfer of aged care facility (ACF) residents to an acute hospital is sometimes necessary, for those at end of life this can cause fragmented care and disruption. Aim To explore the characteristics of ACF residents transferred to hospital in the last 24 h of life and factors that might influence this decision, including access to medical review, advance care planning (ACP) and pre‐emptive symptom management prescribing, an area not previously researched. Methods A retrospective observational audit of ACF residents transferred to a metropolitan hospital between 2012 and 2017 who died within 24 h of transfer. Results A total of 149 patients met the criteria. The median age was 87 years, and 63 (42%) were male. Eighty‐three (56%) were transferred ‘out‐of‐hours’, the majority (71%) having no medical review in the 24 h prior, and 43 (29%) died within 4 h of arrival. The most common reasons for transfer were dyspnoea (46%) and altered conscious state (32%), and the most common cause of death was pneumonia (37%). Some form of ACP documentation was available in 48%. Of the 86 (58%) patients who required injectable opioid for symptom management in hospital, only 7 (8%) had this pre‐emptively prescribed on their ACF medication chart. Conclusions Appropriate decision‐making around hospital transfers and end‐of‐life care for ACF residents may be influenced by access to professionals able to diagnose dying and access to appropriate symptom management medications. ACP is important, but often requires the aforementioned to be enacted. Further research is needed to better inform how we can identify and meet the end‐of‐life care needs of this cohort.

Mental health nurses are exposed to high levels of aggressive and challenging patient behaviours. This can cause stress and burnout which is associated with poor staff, patient, and organization outcomes, including unplanned nursing staff leave (UNSL). This study explores the correlation between a patient behaviour tool (RAGE), variations of which are frequently used in mental health and psychogeriatric nursing, and the staff outcome of UNSL. The study is reported according to the STROBE Statement for reporting of observational studies. RAGE scores and ward characteristics were recorded weekly for 26 weeks on an Australian metropolitan psychogeriatric ward and correlated with UNSL for the same week and the following week (allowing for any ‘lag effect’ behaviours may have on leave). There was a moderate negative correlation between mean RAGE score and UNSL for the same week (r = −0.34) and no correlation the following week (r = 0.08). Similarly, there was low to no correlation between ward characteristics and UNSL. The trends seen in this exploratory study should be further interrogated in adequately powered future studies. UNSL is likely influenced by complex factors including staff experience, coping mechanism, and specific patient behaviours such as physical or verbal aggression and whether the behaviours resulted in injury, which should be included in future studies. The novel strategy of utilizing patient‐centred tools to predict staff outcomes is feasible and warrants further exploration.

Reassuringly, patients had a reasonable understanding of diagnosis and purpose of GEM admission. Lack of understanding of medications, especially among NESB patients, should be improved.