Latent Profiles of Advance Care Planning Engagement and Death Attitudes in Dialysis Patients with End-Stage Renal Disease.

Background: Early identification of individuals with low advance care planning (ACP) engagement remains a critical component of clinical care. However, we know little about the heterogeneity of ACP engagement at the individual level. This study identified latent subgroups of ACP engagement using latent profile analysis (LPA), and explored their associations with death attitudes. Methods: This study recruited 302 end-stage renal disease (ESRD) patients undergoing dialysis. Data included sociodemographic characteristics, the Advance Care Planning Engagement Survey (ACPES; Chinese version), and the Death Attitude Profile-Revised (DAP-R). Based on multidimensional indicators, LPA was employed to identify distinct ACP engagement profiles. Model fit and classification quality in LPA were evaluated based on class sizes and entropy values. All analyses were completed in SPSS 26.0 and Mplus 8.3, with R3STEP and BCH methods employed to uncover underlying patterns and relationships. Results: Among dialysis-dependent ESRD patients, ACP engagement was categorized into two latent profiles: a "low-ACP Engagement" profile (n = 162, 53.6%) and a "high-ACP Engagement" profile (n = 140, 46.4%), with good classification quality (entropy = 0.909). The profile membership was significantly associated with dialysis vintage, and educational level (both p < 0.05). BCH analyses revealed that patients with low ACP engagement reported significantly higher fear of death and death avoidance, whereas those with high ACP engagement showed greater neutral acceptance of death (all p < 0.001). Conclusion: This study identifies two distinct ACP engagement profiles among dialysis-dependent ESRD patients. Findings emphasize the need for tailored interventions, particularly for patients with shorter dialysis vintage and lower education level, and highlight the role of death attitudes in shaping ACP engagement. These findings should be interpreted with caution due to the cross-sectional design and single-center setting.