On path analysis, the model in which pain intensity affected psychological factors had the most explanation. Ninety-three patients diagnosed with frozen shoulder were included in this study. The model was modified repeatedly based on the posterior prediction p value, deviance information criterion (DIC), and Bayesian information criterion (BIC) the model with the highest explanatory power was adopted as the final model. Pain intensity (Numerical Rating Scale (NRS)), pain catastrophizing (Pain Catastrophizing Scale (PCS)), and self-efficacy (Pain Self-Efficacy Questionnaire (PSEQ)) were measured at the first examination, and the relationship was examined using the Bayesian estimation method.
Patients aged 18 years or older, who had been symptomatic for < 1 year and reported localized pain in one shoulder, experienced night pain, and had restricted active and passive shoulder motions were included. Participants who were diagnosed with frozen shoulder between January 2016 and March 2017 were recruited from a single orthopedic clinic. Thus, this study aimed to examine the structural relationship between pain catastrophizing, self-efficacy, and pain intensity in patients with frozen shoulder. However, it remains unclear whether the effects are direct or indirect in patients with frozen shoulder the impact on each variable has also not been clarified. Pain catastrophizing and self-efficacy are useful for predicting pain these are associated with pain intensity and facilitate evaluation of psychological factors.