Trainee-Led Patient Education to Increase Advance Care Planning in a Geriatric Primary Care Clinic

BACKGROUND: Advance care planning (ACP) is critical for communicating patients' preferences for future medical care, especially in geriatric populations. Despite the acknowledged importance, completion rates remain low. OBJECTIVE: We aimed to increase ACP in a geriatric primary care clinic (geri-PCC) using a trainee-led educational intervention. DESIGN: Retrospective analysis of a quality improvement intervention. PARTICIPANTS: In total, 498 patients, > 60 years old with no advance directives (AD) or Portable Medical Orders (POLST) were seen in the geri-PCC during the intervention period. Sixty-five received the intervention while 433 did not. INTERVENTION: We incorporated an ACP education session led by medical trainees into the geri-PCC workflow 3/2023-4/2023. MAIN MEASURES: New AD (includes durable power of attorney for healthcare (dPOA) and Living Will (LW)) and POLST captured in the EMR, ICD-10 codes billed for ACP/goals of care (GOC) were compared between the intervention and standard care groups during 6 months following the intervention period. Secondary logistic regression analyses adjusted for age, gender, mood disorder, and cognitive impairment when evaluating the intervention's association with primary outcomes. KEY RESULTS: The intervention group had more new dPOA forms (13.8% vs 5.3%, p = 0.025) and ACP/GOC codes (32% vs 13%, p < 0.001). These findings remained significant after adjusting for age, gender, mood disorder, and cognitive impairment (dPOA OR = 2.96, 95% CI 1.24-6.59, p = 0.010; ACP/GOC OR 3.07, 95% CI 1.67-5.52, p < 0.001). Cognitive impairment was associated with more completion of POLST (OR = 3.29, 95% CI 1.15-11.0, p = 0.035) but less completion of LWs (OR = 0.34, 95% CI 0.13-0.79, p = 0.018). CONCLUSIONS: Incorporating an ACP education session led by medical trainees was associated with increased AD documentation in the EMR and increased ACP/GOC codes billed in our geri-PCC. The emergence of cognitive impairment as a predictor of POLST completion needs further exploration.