English and Spanish-speaking vulnerable older adults report many barriers to advance care planning.

Background/objectives: Advance care planning (ACP) rates are low in diverse, vulnerable older adults, yet little is known about the unique barriers they face and how these barriers impact ACP documentation rates.

Design: Validated questionnaires listing patient, family/friend, and clinician/system-level ACP barriers and an open-ended question on ACP barriers.

Setting: Two San Francisco public/Department of Veterans Affairs hospitals.

Participants: One thousand two hundred and forty-one English and Spanish-speaking patients, aged 55 and older, with two or more chronic conditions.

Measurements: The open-ended question on ACP barriers was analyzed using content analysis. We conducted chart review for prior ACP documentation. We used chi-square/Wilcoxon rank-sum tests and logistic regression to assess associations between ACP barriers and demographic characteristics/ACP documentation.

Results: Participant mean age was 65 ± 7.4 years; they were 74% from racial/ethnic minority groups, 36% Spanish-speaking, and 36% with limited health literacy. A total of 26 barriers were identified (15 patient, 4 family/friend, 7 clinician/system-level), and 91% reported at least one ACP barrier (mean: 5.6 ± 4.0). The most common barriers were: (patient-level) discomfort thinking about ACP (60%), wanting to leave health decisions to "God" (44%); (family/friend-level) not wanting to burden friends/family (33%), assuming friends/family already knew their preferences (31%); (clinician/system-level) assuming doctors already knew their preferences (41%), and mistrust (37%). Compared with those with no barriers, participants with at least one reported barrier were more likely to be from a racial/ethnic minority group (76% vs 53%), Spanish-speaking (39% vs 6%), with fair-to-poor health (48% vs 34%), and limited health literacy (39% vs 9%) (p < 0.001 for all). Participants who reported barriers were less likely to have ACP documentation (adjusted odds ratio = 0.64, 95% confidence interval [0.42, 0.98]).

Conclusion: English- and Spanish-speaking older adults reported 26 unique barriers to ACP, with higher barriers among vulnerable populations, and barriers were associated with lower ACP documentation. Barriers must be considered when developing customized ACP interventions for diverse older adults.