Ethical considerations
This cross-sectional study was conducted with approval of the local ethics review board (approval number: 16-48). The content and purposes of the study were explained in advance to the patients, and consent to participate in this study was obtained from the prospective patients. This investigation was conducted according to the principles expressed in the Declaration of Helsinki.
Patients
We enrolled 93 patients who were diagnosed with frozen shoulder and started rehabilitation on any day from January 2016 until March 2017. Patients who had calcium deposits or rotator cuff injuries, fractures, required surgery, work-related injuries or injuries due to accidents, bilateral disorders, apparent dementia were excluded. The content and purpose of the study was explained in advance to these patients, and consent to participate in this study was obtained from prospective subjects.
Outcome measures
Pain was evaluated in all patients at the first examination. In order to control the measurement bias, same physical therapists performed all evaluations for each subject. In addition, the answer of the questionnaire, only the patient was performed.
Pain intensity was determined using the Numerical Rating Scale (NRS) [10]. The NRS is a common scale for quantifying pain and is assessed using an 11-point Likert-type scale (0=no pain; 10=worst pain). The scale was used to record the average pain intensity.
Pain catastrophizing was determined using the Japanese version of the Pain Catastrophizing Scale (PCS) [11]. The PCS is a 13-item self-report questionnaire that measures maladaptive thoughts regarding pain. Each item is rated on a 5-point Likert-type scale (0=not at all; 4=all the time). The PCS contains three dimensions of pain catastrophizing. Higher scores reflect a greater degree of pain catastrophizing. Rumination is paying too much attention to pain-related thoughts. Helplessness is to feel helpless in dealing with a painful situation. Magnification is an overview of the threat of pain.
Self-efficacy was assessed using the Japanese version of the Pain Self-Efficacy Questionnaire (PSEQ) [12]. The PSEQ is a 10-item self-report questionnaire that measures the degree of self-confidence in performing an activity in a painful state. Each item is rated on a 7-point Likert-type scale (0=not at all confident; 6=completely confident). Higher scores reflect stronger self-efficacy beliefs.
Statistical analysis
Covariance structure analysis was performed using the hypothesized model. Previous studies [13-15] have reported that there is no correlation between self-efficacy and pain intensity, and there is a correlation between self-efficacy and pain catastrophizing, pain catastrophizing, and pain intensity. Based on these reports, we developed a hypothesis model in which self-efficacy affects pain catastrophizing and pain catastrophizing affects pain intensity (Figure 1). Bayesian estimation implemented via Markov Chain Monte Carlo simulation was used for the estimation method. The Bayesian estimation method is one of the estimation methods in structural equation modeling, and is a method that can estimate the true value or a value close to it even with a complicated model, data that is not normally distributed, or an under sample size. The reason for adopting the Bayesian estimation method in this study is that there is a concern that the distribution of the data may not be stable because the sample size is small and the data is an ordinal scale. The estimation values are presented as median; the chain number was 5, and the number of simulations ranged from 10,000 to 50,000. The convergence of the model determined that the potential scale reduction would be less than 1.1. The posterior prediction p-value was referenced for model fit.
Mplus, version 8.0 (Muthen & Muthen, http://www.statmodel.com/) was used to perform the statistical analyses. The significance level was set at P<0.05 Individuals with missing values were excluded from the specific outcome measure analysis.