Symptom Burden Poorly Responsive to Palliative Care Intervention and Karnofsky Predict Survival in an Acute Palliative Care Unit.

Abstract

Background/Objective: Survival prediction in the advanced cancer care setting plays a vital role in treatment planning and patients' arrangements. The aim of this study was to examine the association of the global Edmonton Symptom Assessment System (GESAS) and Karnofsky scale (KPS) with overall survival (OS) in patients with advanced cancers admitted to an acute palliative care unit (APCU). The second aim was to assess if GESAS changes after comprehensive palliative treatment could influence OS. Methods: This is a prospective planned sub-analysis of advanced cancer patients. A consecutive sample of 521 patients admitted to an APCU. Patients with available survival in follow-up phone calls, having complete ESAS, and discharged alive were selected. KPS and GESAS were measured at admission and after seven days of individual comprehensive palliative care. Results: Two hundred forty-three of 521 screened patients were assessed according to inclusion criteria. The mean age was 67.1 years (SD 11.5), and 121 patients were male. The mean KPS was 43.5 (SD 9.3). The mean OS was 74.6 (SD 136.2) days. Significant changes in GESAS were observed after one week. Univariate linear regression analysis showed that KPS and GESAS at T0 and at T7 were correlated with OS (p < 0.0005; p = 0.020; p < 0.0005, respectively). At multivariate analysis, OS was correlated with KPS and GESAS at discharge (B = 3.349, 95% CI = 1.560-5.137; B = -2.430, 95% CI = -3.831--1.029). Discussion: KPS and poor response to intensive treatment, maintaining high GESAS scores, can be considered predictive factors of shorter OS. Further studies should confirm whether a specialized intervention in other settings can improve OS.