Consultations' demand for a hospital palliative care unit: how to increase appropriateness? Implementing and evaluating a multicomponent educational intervention aimed at increase palliative care complexity perception skill.

Background: Planned, multidisciplinary teams' discussions of cases are common in cancer care, but their impact on patients' outcome is not always clear. Palliative care (PC) needs might emerge long before the last weeks of life. Many palliative care patients could be managed from the usual care staff, if appropriately trained; specialist palliative care should be provided to patients with more complex needs. Staff needs adequate training, so that only patients presenting a higher complexity are properly referred to the second level ("specialized") PC services. In the considered hospital setting, "tumour boards" (multidisciplinary discussions) refer often to a low number of patients. Overall complexity of patients' needs is hardly considered.

Methods: A mixed method pilot study with data triangulation of professionals' interviews and an independently structured evaluation of complexity of referred patients, before and after the intervention, using the PALCOM instrument. We trained four teams of professionals to deliver first-level palliation and to refer patients with complex needs detected in multidisciplinary discussions. A multicomponent, first level PC educational intervention, including information technology's adaptation, a training course, and bedside training was offered from the specialized PC Services, to all the HPs involved in multidisciplinary pancreas, lung, ovarian, and liver tumour boards.

Results: While the level of complexity of referred patients did not increase, trainees seemed to develop a better understanding of palliative care and a higher sensitivity to palliative needs. The number of referred patients increased, but patients' complexity did not. Qualitative data showed that professionals seemed to be more aware of the complexity of PC needs. A "meaning shift" was perceived, specifically on the referral process (e.g., "when" and "for what" referring to specialist PC) and on the teams' increased focus on patients' needs. The training, positively received, was adapted to trainees' needs and observations that led also to organizational modifications.

Conclusions: Our multicomponent intervention positively impacted the number of referrals but not the patients' complexity (measured with the PALCOM instrument). Hospital staff does not easily recognize that patients may have PC needs significantly earlier than at the end of life.