Association of Advance Care Planning Visits With Intensity of Health Care for Medicare Beneficiaries With Serious Illness at the End of Life.

Importance: Advance care planning (ACP) is intended to maximize the concordance of preferences with end-of-life (EOL) care and is assumed to lead to less intensive use of health care services. The Centers for Medicare & Medicaid Services began reimbursing clinicians for ACP discussions with patients in 2016. Objective: To determine whether billed ACP visits are associated with intensive use of health care services at EOL. Design Setting and Participants: This prospective patient-level cohort analysis of seriously ill patients included Medicare fee-for-service beneficiaries who met criteria for serious illness from January 1 to December 31, 2016, and died from January 1, 2017, to December 31, 2018. Analyses were completed from November 1, 2020, to March 31, 2021. Main Outcomes and Measures: Five measures of EOL health care services used (inpatient admission, emergency department visit, and/or intensive care unit stay within 30 days of death; in-hospital death; and timing of first hospice bill) and a measure of EOL expenditures. Analyses were adjusted for age, race and ethnicity, sex, Charlson Comorbidity Index, Medicare-Medicaid dual eligibility, and expenditure by hospital referral region (high, medium, or low). The primary exposure was receipt of a billed ACP service classified as none, timely (>1 month before death), or late (first ACP visit ≤1 month before death). Results: Of the 955 777 Medicare beneficiaries who met criteria for serious illness in 2016 and died in 2017 or 2018, 522 737 (54.7%) were women, 764 666 (80.0%) were 75 years or older, and 822 684 (86.1%) were non-Hispanic White individuals. Among the study population, 81 131 (8.5%) had a timely ACP visit, and an additional 22 804 (2.4%) had a late ACP visit. After multivariable adjustment, compared with patients without any billed ACP visit, patients with a timely ACP visit experienced significantly less intensive EOL care on 4 of 5 measures, including in-hospital death (adjusted odds ratio [aOR], 0.85; 95% CI, 0.84-0.87), hospital admission (aOR, 0.84; 95% CI, 0.83-0.85), intensive care unit admission (aOR, 0.87; 95% CI, 0.85-0.88), and emergency department visit (OR, 0.83; 95% CI, 0.82-0.84). Only small or insignificant differences in late hospice use or mean total EOL expenditures were noted. Compared with patients without ACP, patients with late ACP experienced more intensive EOL care, including in-hospital death (aOR, 1.22; 95% CI, 1.19-1.26), hospital admission (aOR, 5.28; 95% CI, 5.07-5.50), intensive care unit admission (aOR, 1.57; 95% CI, 1.53-1.62), and emergency department visit (aOR, 3.87; 95% CI, 3.72-4.02). Conclusions and Relevance: In this cohort study of US Medicare beneficiaries, billed ACP services during the EOL course of patients with serious illness were relatively uncommon, but if they occurred before the last month of life, they were associated with less intensive use of EOL services. Further research on the variables affecting hospice use and expenditures in the EOL period and the differential effect of late ACP is recommended to understand the relative role of ACP in achieving goal-concordant care.