Introduction Though medical input to hospice inpatients is well-established, the evidence detailing the nature and level of medical staffing is lacking. Moreover, there are calls for hospices to develop less medical, more public health models.

To facilitate nursing skill-mix changes, and broaden our ‘reach’ (to include ‘low-complexity’ patients), our inpatient unit was split equally into generalist nursing (GN) and specialist nursing (SN) beds. It was suggested GN-suitable patients would correspondingly have few medical needs. To facilitate medical workforce planning, we wanted to evaluate the medical needs across these potentially contrasting populations.

Methods We developed a tool to detail the nature and intensity of medical interventions. We completed a 1 month prospective pilot, in a 28-bedded UK hospice, scoring perceived patient need each day.

Results A tool reflecting overall medical need was generated; with 3-ratings (low, moderate, high), across 9 items (e.g. urgency, clinical complexity, trajectory, discord).

284 patient assessments were completed (100%); the range of medical need for patients in GN beds was; low=78, moderate=41, high=14 and for SN beds; low=41, moderate=63, high=46. Concordance of medical and nursing complexity for GN patients=58% and SN patients=31%

Discussion A spectrum of need for medical input to hospice inpatients was confirmed; the level fluctuated during a patient’s stay and high needs were not restricted to SN patients.

The limited concordance between a patient’s perceived need for medical input and their suitability for SN or GN, questioned the wider applicability of this differentiation. There was a trend for lower medical input for ‘GN’ compared to ‘SN patients’. However, 41% of ‘GN patients’ had moderate or high medical needs and only 31% of ‘SN patients’ had high medical needs.

The pilot tool appeared suitable for benchmarking the need for medical input; informing our workforce planning and warranting further evaluation, to include other care settings.