Introduction Visits to the Emergency Department by cancer patients are increasing but are associated with poor outcomes and capacity concerns. Current strategies aim to decrease hospital admissions and manage needs in community e.g. home/hospice. However it is not known whether needs of patients admitted from hospital (HA) to hospice could have alternatively been met by hospice IPU admission at outset.

Methods Cross-sectional retrospective analysis of consecutive HA to Hospice IPU (14 bed unit) August 16–March 17. Data collected: Demographics and outcome data from hospital/hospice admission. Two clinicians assessed whether management could have been met in hospice IPU at outset. Data anonymised and treated confidentially. Results analysed and descriptive statistics utilised.

Results Data from 50 consecutive transfers collected. Mean age 71.3 years (SD 13.7) and 60% male. 82% had a cancer diagnosis (Lung (22%) most common). More than half (52%) of patients were known to SPC Services prior to hospital admission. 38% (19/50) of patients self-referred to A and E. Half of admissions were in working hours (52% admitted 9 am-5 pm Mon-Fri). Most common phase of illness on hospital 1 st assessment was unstable (n=22, 44%). Mean AKPS 38.9 (SD 22). Mean acute hospital inpatient length of stay was 14.7 days (SD 11.2; range 1–49 days). In 38 cases it was assessed that hospice IPU could not have met immediate needs at outset. Reasons: needing procedure/operation/hospital specialty review 9/38, needs acute investigations 9/38; acute sepsis treatment 9/38; chemotherapy related problems 3/38, patient previously not known to have palliative diagnosis/unknown to hospice 8/38. The majority of HA (90%) died during that hospice admission, 4 (10%) discharged. Total length IPU stay 13.2 days (SD 16.4) but range broad (1–73 days).

Conclusions Most HA patients needed initial acute hospital admission at outset (rather than direct IPU admission). Findings will guide 24/7-service development and future planning of hospice beds.