Palliative care provides an extra layer of support to patients and families facing a serious illness. To date, several studies support the use of early, integrated palliative care for patients with cancer, based upon documented improvements in quality of life, symptoms, mood, satisfaction, utilization, and even overall survival. Despite this, patients with cancer continue to have unmet palliative care needs, and palliative care services are often engaged late in their care, if at all. Amid this under-utilization, questions remain about the optimal timing and nature of palliative care integration. To answer this question, we briefly review the evidence based for palliative care in oncology, and discuss three approaches to optimizing the timing of palliative care integration: (1) prognosis-based, (2) needs-based, and (3) trigger-based models. Prognosis-based models most closely mirror the approach of randomized trials to date, but are overly dependent on prognostication, and may miss patients with unmet needs who do not meet standard definitions of poor-prognosis disease. Needs-based models may better capture patients in a personalized manner, based on actual needs, but require sophisticated screening systems to be integrated into routine care processes, along with clinician buy-in. This may lead to excessive referrals, which strain the already limited palliative care workforce. As such, a blended, trigger-based approach may be best, allowing one to utilize certain disease-based and prognosis-based triggers for referral, plus screening of unmet needs, to identify those patients most likely to benefit from integrated palliative care when they need it most.
Keywords: Integration; Oncology; Palliative care; Prognosis; Referral; Unmet needs.