A Quality Improvement Initiative for Inpatient Advance Care Planning

The Centers for Medicare & Medicaid Services (CMS) implemented advance care planning (ACP) billing codes in 2016 to encourage practitioners to conduct and document ACP conversations, and included ACP as a quality metric in the CMS Bundled Payments for Care Improvement Initiative in 2018. Use of this billing code in the inpatient setting has not been studied.

To determine whether a quality improvement intervention to increase inpatient ACP is effective in increasing ACP billing rates or changing hospital treatment plans or patient outcomes.

This nationwide cohort study and difference-in-differences analyses compared changes in ACP billing, treatment, and outcomes in Medicare fee-for-service beneficiaries aged 65 years and older who were hospitalized and cared for by 3 different groups: practitioners employed by a national acute care staffing organization who underwent an ACP quality improvement intervention, nonintervention practitioners at the same hospital, and control group practitioners from other hospitals. Using data from the Master Beneficiary Summary File, acute care hospital stays for nonsurgical conditions were linked to Medicare enrollment, claims, and vital status data from 1-year preadmission to 1-year postadmission from 2015 to 2019. The ACP billing rates for each group were assessed for associations with 6 inpatient treatments and 4 outcomes. Data analyses were performed from January 2022 to December 2024.

Billed ACP conversations, receipt of intensive care and life support (intensive care unit admission, gastrostomy tube placement, mechanical ventilation, tracheostomy), treatment limitations (newly initiated do-not-resuscitate orders) and outcomes (discharge to hospice, inpatient death, 30-day postadmission death, and 1-year postadmission death).

The total study sample included 109 intervention hospitals, 1691 control hospitals, nearly 12 million Medicare fee-for-service beneficiaries aged 65 years and older, and 738 309 practitioners associated with admissions from 2016 to 2018. ACP billing rates increased more for the intervention (1.3% in preintervention to 14.0% in postintervention; P < .001) than for the nonintervention (same hospitals) and control groups (odds ratio [OR], 2.6; 95% CI, 1.7-4.1 intervention vs control). Increased ACP billing rates were significantly associated with decreased rates of inpatient death in the intervention group (OR, 0.95; 95% CI, 0.90-1.00) compared to the nonintervention (OR, 1.10; 95% CI, 1.04-1.17) and control groups (reference). All other associations were nonsignificant.

This nationwide cohort study suggests that while the ACP quality initiative increased ACP billing, changes in clinical outcomes were inconsistent with the hypotheses. Further study is needed to address questions regarding confounding by unobserved measures of care quality.