A Comparison of Hospitalized Patients With Heart Failure and Cancer Referred to Palliative Care

Importance Growing evidence shows that palliative care (PC) improves treatment outcomes in patients with heart failure (HF), but few large-scale studies have prospectively evaluated the processes and outcomes associated with PC consultation for such patients in the real world. Objective To characterize processes and outcomes of PC consultations for hospitalized patients with HF compared with patients with cancer. Design, Setting, and Participants This cohort study of inpatient encounters at community and academic hospitals in the Palliative Care Quality Network enrolled participants between 2013 and 2017. Of a total of 135 197 patients, 57 272 adults with a primary diagnosis of HF or cancer receiving PC consultation were enrolled. Data analysis was performed from April 2018 to December 2019. Exposures Primary diagnosis of HF or cancer. Main Outcomes and Measures Symptom improvement and changes in care planning documentation after PC consultation. Results At the time of consultation, patients with HF were older (mean age, 75.3 years [95% CI, 75.0-75.5 years] vs 65.2 years [95% CI, 65.0-65.3 years]; P < .001), had lower Palliative Performance Scale scores (mean, 35.6% [95% CI, 35.3%-35.9%] vs 42.4% [95% CI, 42.2%-42.6%]; P < .001), and were more likely to be in a critical care unit (5808 of 16 741 patients [35.3%] vs 4985 of 40 531 patients [12.5%]; P < .001) or a telemetry or step-down unit (5802 of 16 741 patients [35.2%] vs 7651 of 40 531 patients [19.2%]; P < .001) compared with patients with cancer. Patients with HF were less likely than patients with cancer to be referred to PC within 24 hours of admission (6773 of 16 741 patients [41.2%] vs 19 348 of 40 531 patients [49.0%]; P < .001) and had longer hospitalizations before receiving PC consultation requests (mean, 4.6 days [95% CI, 4.4-4.8 days] vs 3.9 days [95% CI, 3.8-4.0 days]; P < .001). Patients with HF were referred less frequently for symptoms other than pain (1686 of 16 488 patients [10.2%] vs 8587 of 39 609 patients [21.7%]; P < .001), but were equally likely to report improvements in anxiety (odds ratio, 0.85; 95% CI, 0.71-1.02; P = .08) and more likely to report improvements in dyspnea (odds ratio, 2.17; 95% CI, 1.83-2.57; P < .001) compared with patients with cancer. Patients with HF were less likely than those with cancer to be discharged alive (odds ratio, 0.78; 95% CI, 0.64-0.96; P = .02) or to be referred to hospice (odds ratio, 0.50; 95% CI, 0.47-0.53; P < .001). Conclusions and Relevance These findings suggest that PC referral comes late for patients with HF and is used primarily to discuss care planning. Practitioners caring for patients with HF should consider involving PC experts earlier for symptom management.