Health Belief Clusters Among Stroke Patients: Insights for Tailored Interventions Using Latent Profile Analysis.

Aim: To determine the latent profiles of health beliefs for stroke patients and analyze the differences among different clusters of stroke patients. Methods: 1358 stroke patients were recruited by a stratified cluster random sampling approach in Henan Province, China, from March to April 2023. The instruments used were the general situation questionnaire, Stroke Knowledge Questionnaire, National Institute of Health Stroke Scale, modified Rankin Scale, Health Behavior Scale for Stroke Patients, and Short Form Health Belief Model Scale. Questionnaires were distributed through the Questionnaire Star Platform. The health belief clusters were determined by Mplus8.3. Pearson's chi-square test and one-way ANOVA in SPSS 26.0 were applied to examine the profiles' differences. Results: A total of 1358 questionnaires were returned, resulting in an effective response rate of 87.33%. Model fitting indicators supported the three-class model of health beliefs. Latent profile analysis identified three distinct clusters, labeled Class 1 (health belief absence cluster totally), Class 2 (self-deficiency deprivation cluster), and Class 3 (higher perception behavior disorder cluster). Total Short Form Health Belief Model Scale scores among Class 1-3 differed significantly. Single-factor analysis showed statistical significance in the modified Rankin Scale, health behavior, health knowledge, education, residence, profession, medical insurance mode, and hypertension. Conclusion: Individual heterogeneity is evident in health beliefs among stroke patients. The belief can be divided into three potential clusters. The study provides reference for the developing regular health belief and provide methodological guidance for other clinical studies. Clinicians are able to offer tailored guidance according to the different health beliefs of patients.