Background Good referral forms should cover enough key areas to allow effective triage (Donaldson, Carter & Green, 2000) and elicit complete and clear information from referrers (Depasquale & Crockford, 2005). Feedback from current referrers to two neighbouring hospices covering the same NHS Trust suggested existing referral forms were complicated, time consuming, often necessitating further information gathering before they could be triaged.

Aims To assess the completeness of a sample of referrals to specialist palliative care and the proportion of these that could be effectively triaged given the information provided. To utilise the findings in redesigning the referral form.

Methods Retrospective analysis of all referral forms for new patients received by both hospices during August 2017. Internal/existing patient referrals were excluded. Analysis of completion rates of different sections was undertaken and subjective opinion given by one clinician per site as to whether the form could be effectively triaged. The findings influenced referral form redesign.

Results 205 completed referral forms were evaluated. Of 100 referrals to hospice A, the mean percentage of sections completed was 75% and 82% of referrals could be triaged based on form data alone. Of 105 referrals to hospice B, a mean of 73% of sections were completed and 89% could be triaged. Particular sections of the form were repeatedly left blank. However, even at the lowest completion rates (23%) forms could be triaged, usually due to information provided in ‘any other comments’ section. Given these results, the forms were made more concise and included an ‘open comments’ section. Following stakeholder engagement, ‘phase of illness’ and ‘Australian Karnofsky Performance Status’ were added (National End of Life Care Intelligence Network, 2016; Abernethy, Shelby-James, Fazekas, Woods et al., 2005).

Discussion The pilot of the new referral form is currently underway with initial positive feedback from referrers. Preliminary analysis from evaluation of the first pilot forms received, suggests a high proportion can be adequately triaged. Pending further results adoption throughout clinical community is planned.