Examining the Influence of Social Determinants on Advance Care Planning in End-Stage Renal Disease

Background and Purpose: Advance care planning (ACP) allows individuals to express their preferences for future medical care in case they become unable to make their own decisions. Given the reduced survival rate, potential for cognitive decline, and diminished physical and social functioning, individuals on dialysis would benefit from engaging in ACP early. Early ACP conversations offer patients and their families time to become comfortable with medical decision-making and to reach consensus on end-of-life care. They also ease the burden on family members through improved information and preparedness to make decisions aligned with patient’s wishes. Despite these benefits, ACP engagement remains low. This study aimed to examine the impact of social determinants on ACP engagement among patients with end-stage renal disease (ESRD).

Methods: Data were from the United States Renal Data System (USRDS). ACP engagement was identified by Current Procedural Terminology (CPT) codes 99497 or 99498, used for billing ACP discussions with patients, families, or surrogates. Social determinants assessed included gender, age, race/ethnicity, insurance type, and relevant clinical factors. The analysis included 1,531,695 ESRD patients between 2016 and 2021.

Results: Only 2,563 individuals (0.2%) received an ACP consultation. However, engagement varied significantly across patient demographics and insurance types. After controlling for dialysis duration, dialysis modality and comorbidities, logistic regression analysis revealed that gender, age, race/ethnicity, and insurance types were key predictors of ACP discussions. Males were less likely than females to engage in ACP (OR=0.891, p=0.004). Younger adults were significantly less likely to engage compared to those aged 65 and older, with odds ratios ranging from 0.150 for the 18-30 age group to 0.537 for those aged 41-64 (all p<0.001). Regarding race/ethnicity, although Black individuals were not significant, Hispanics (OR=0.702, p < 0.001), non-Hispanic Asians (OR=0.622, p<0.001), and non-Hispanic individuals of other races (OR=0.472, p<0.001) were significantly less likely to engage in ACP discussions compared with non-Hispanic Whites. Insurance types also played a significant role. Patients covered by traditional Medicare Fee-for-Service (MPAB + MPO) were more than twice as likely to engage in ACP (OR=2.284, p<0.001). In contrast, those with Medicare Advantage HMOs (OR=0.910, p=0.040) or Medicare as a secondary payer (MSP, OR=0.717, p<0.001) were less likely to participate in ACP discussions. Additionally, Medicare and Medicaid dual-eligible patients had higher odds of engaging in ACP compared to non-dual eligible individuals (OR=1.404, p<0.001). We acknowledge the substantial imbalance in the outcome variable; however, model diagnostics indicated satisfactory convergence. We also applied robust standard errors to enhance the reliability of our inference.

Conclusion and Implications: The study highlights the persistently low engagement in ACP discussions among patients with ESRD. The findings indicated that ACP engagement is not uniform across the ESRD population, but is influenced by key social determinants. These findings underscore the need for targeted strategies to promote ACP, particularly among younger patients, racial and ethnic minorities, and those with more restrictive insurance plans and provider networks. Future research should also explore providers’ perspectives on engagement in ACP with ESRD patients.