Characteristics of Advance Care Planning in Patients With Cancer Referred to Palliative Care

Abstract

QUESTION ASKED: Within a high-need population of patients with cancer-patients referred to palliative care-is advance care planning (ACP) being addressed? SUMMARY ANSWER: Nearly half (47%) of the patients with cancer in our study lacked advance directives, with deficits in ACP noted particularly in younger patients and non-White patients. When we account for differences in demographics and performance status, we find no difference in ACP completion in patients with cancer versus without cancer referred to palliative care. WHAT WE DID: We performed a retrospective cross-sectional analysis of initial outpatient palliative care visits using data from the Quality Data Collection Tool for Palliative Care (QDACT-PC) database from 2015 to 2019. QDACT-PC is a web-based point-of-care registry used by palliative care providers across the United States to track quality metrics associated with patient consultations. We compared completion of ACP and resuscitation status between patients with cancer and patients with a noncancer diagnosis. WHAT WE FOUND: In our unadjusted comparison, patients with cancer were less likely to have a do-not-resuscitate or do-not-intubate (DNR/DNI) order (37% v 53%; P , .0001) and less likely to have an advance directive (53% v 73%; , .0001) than patients without cancer; rates of healthcare proxy identification were similar (92.8% v 94.5%; P 5 .10). These differences in ACP did not persist when we accounted for age, race, sex, and performance status, with age being the primary explanatory factor. BIAS, CONFOUNDING FACTOR(S): Although we applied inverse probability weighted models to mitigate the impact of selection bias on our results, we had significant data missingness in our sample. QDACT registry participation is voluntary, which may introduce a reporting bias and limit external validity. REAL-LIFE IMPLICATIONS: Rates of ACP in patients with cancer with serious illness remain low, with age being a key driver. Although ACP is a shared responsibility among multiple providers, younger patients may have fewer touchpoints within the health system for addressing this need beyond their oncology visits. abstract PURPOSE Advance care planning (ACP) is a process in which patients share their values, goals, and preferences regarding future medical care. ACP can improve care quality, yet may be challenging to address for patients with cancer. We sought to characterize key components of ACP in patients with cancer as compared with patients with noncancer serious illness referred to palliative care (PC). METHODS We performed a retrospective cross-sectional analysis of initial outpatient PC visits from the Quality Data Collection Tool for PC database from 2015 to 2019. Quality Data Collection Tool is a web-based point-of-care specialty PC registry to track quality metrics. RESULTS We analyzed 1,604 patients with cancer and 1,094 patients without cancer: 44% of patients were female, 87% were White, and 98% were non-Hispanic. The average age was 72.2 years (standard deviation [SD] 15.4). Patients with cancer were on average younger than patients without cancer (66.5 [SD: 13.9] v 80.5 [SD: 13.8]) and had a higher Palliative Performance Scale (PPS) (59.5 [SD: 22.4] v 33.4 [SD: 25.1]). In our unadjusted comparison, patients with cancer were less likely to be DNR/DNI (37% v 53%; P , .0001) and less likely to have an advance directive (53% v 73%; , .0001); rates of healthcare proxy identification were similar (92.8% v 94.5%; P 5 .10). These differences did not persist when we accounted for age, race, sex, and PPS, with age being the primary explanatory factor. CONCLUSION Despite having serious illness meriting PC referral, many patients with cancer in our study lacked advance directives. This highlights both the important role of oncologists in facilitating ACP and the utility of PC playing a complementary role.