Estimating the short-term prognosis to adjust the transfer of patients with terminal cancer to medium-stay palliative care units

Abstract Background and objectives Clinical management for terminal patients should consider various aspects, particularly the patient's functional assessment, which correlates well with the short-term prognosis. The prognosis could improve if the presence of symptoms strongly associated with a poorer progression were included. The study's main objective was to assess whether the prognosis according to the Palliative Performance Scale (PPS) improved with the presence/absence of pain-dyspnea-delirium symptoms. The secondary objective was to determine caregiver satisfaction with the transfer to medium-stay palliative care units (MSPCUs), which are prepared for medium stays of approximately one month. Patients and method We conducted a prospective, observational, multicentre (regional) study that analyzed survival in MSPCUs according to the PPS dichotomized to >20% and ≤20%. We estimated the mean survival functions using the Kaplan–Meier method and compared them according to the Cox proportional hazards ratios (HR). Caregiver satisfaction was studied using an anonymous self-administered Likert questionnaire. Results The study included 130 patients. The PPS ≤ 20% and PPS > 20% subgroups had a median survival of 6 (3–13) days and 21 (11–42) days, respectively, with an unadjusted mortality HR 3.1-fold greater in the PPS ≤ 20% subgroup. The HR did not change when adjusted for the symptoms. Eighty-three percent of the caregivers found the transfer beneficial, and 40% observed better patient care. Conclusions For patients transferred from general hospitals to MSPCUs, PPS scores ≤ 20% were associated with survival shorter than one week, with a 3-fold higher mortality HR than patients with PPS scores > 20%, without the analysis adjusted for the presence of pain-dyspnea-delirium providing greater prognostic accuracy. The caregivers found benefits mainly in the convenience of the facilities and distance.