Latent Profiles of Decision Regret and Associated Factors Among Patients Undergoing Medium- and Long-Term Vascular Access Device Placement in Medical Oncology: A Cross-Sectional Study

Background: Patients receiving medical oncology care must weigh multiple factors when selecting a vascular access device, such as a peripherally inserted central catheter, implantable port, or midline catheter. Decision regret is an important patient-reported outcome that may adversely affect treatment adherence and quality of life. From the perspectives of shared decision-making and precision nursing, identifying subgroups of patients with distinct patterns of decision regret may help provide individualized decision support. Objective: To identify latent subgroups of decision regret among patients with cancer who underwent vascular access device placement, based on the three-dimensional structure of decision regret—process regret, option regret, and outcome regret—and to examine the associations of demographic characteristics, clinical factors, decisional conflict, and shared decision-making with subgroup membership. Methods: A cross-sectional survey was conducted using convenience sampling. A total of 309 patients who had undergone vascular access device placement and participated in treatment-related decision-making were recruited from the oncology department of a tertiary hospital in China between February and June 2026. Data were collected using a researcher-designed demographic and clinical characteristics questionnaire, the Three-Dimensional Decision Regret Scale, the Decisional Conflict Scale, and the Shared Decision-Making Questionnaire. Latent profile analysis was performed to identify subgroups of decision regret, and multinomial logistic regression was used to examine factors associated with subgroup membership. Results: A total of 309 patients were included in the final analysis. Based on model fit indices, LMR/BLRT results, entropy, class proportions, and clinical interpretability, the three-profile solution was selected and identified three latent profiles: low decision regret (77.35%), moderate decision regret (13.59%), and high decision regret (9.06%). Univariable analyses showed that place of residence, educational attainment, employment status, monthly household income, type of vascular access device, family involvement in decision-making, receipt of decision-related education, decisional conflict score, and shared decision-making score differed significantly across profiles (all p < 0.05). In the multinomial logistic regression model, place of residence, educational attainment, employment status, family involvement in decision-making, receipt of decision-related education, decisional conflict score, and shared decision-making score were associated with profile membership. After correcting for classification uncertainty using R3STEP, DCS and SDM-Q-9 remained significantly associated with profile membership, whereas the association with employment status was no longer significant. However, the high-regret group was small (n = 28), and the confidence intervals for several predictors were wide (e.g., unemployment: relative risk ratio (RRR) = 30.5, 95% CI 3.24–288). Monthly household income and type of vascular access device did not remain statistically significant. Conclusions: Substantial heterogeneity in decision regret was observed among patients with cancer who underwent vascular access device placement. Particular attention should be given to patients who experience high levels of decisional conflict or receive insufficient preprocedural education. Profile-based nursing interventions and enhanced shared decision-making before device placement are promising strategies, but their effects need to be evaluated in prospective studies.