Answers are grounded in this paper and cite their sources.
Background Postoperative follow-up is essential for maintaining arteriovenous fistula (AVF) function in maintenance hemodialysis (MHD) patients, yet completion rates remain suboptimal. Determinants of follow-up adherence are multifactorial and may involve interactions that traditional regression models fail to capture. This study aimed to identify factors influencing post-AVF follow-up completion and to compare the predictive performance of logistic regression and decision tree modeling. Methods We conducted a retrospective single-center cohort study of 818 MHD patients who underwent AVF creation between January 2023 and December 2024. Follow-up completion rate was calculated as attended visits divided by scheduled visits, and patients were dichotomized at the median rate (60%). Candidate predictors were entered into a binary logistic regression (Enter method) and a classification and regression tree (CRT) model. Model discrimination was compared using ROC-AUC and the DeLong test. Results The mean follow-up completion rate was 62.9% (median 60.0%). Logistic regression identified seven independent factors favoring higher completion: age ≥ 60 years (OR = 1.596), senior high school education (OR = 1.751), urban residence (OR = 1.716), monthly household income of 4001–6000 CNY (OR = 1.849), receiving hemodialysis at the surgical center (OR = 1.563), number of post-AVF reinterventions (OR = 3.015), and higher MBI score (OR = 1.027). The CRT model achieved a higher AUC than logistic regression (0.762 vs. 0.693, DeLong P < 0.001). The Modified Barthel Index was the primary splitting variable, followed by reinterventions, residence, education, income, and age, revealing hierarchical interaction pathways. Conclusions Follow-up adherence after AVF creation is multifactorial. CRT modeling provides superior risk stratification and reveals interaction structures invisible to logistic regression. Reinterventions and urban residence drive adherence, whereas functional impairment and rural residence identify vulnerable subgroups. These findings support targeted, risk-stratified interventions to improve long-term AVF outcomes. Trial registration Not applicable.
No takes yet. Share an insight, caveat, or question.
Yu et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: