The purpose of this study was to establish clinically informed sub-group cut-offs for the PSEQ by examining correlations, interactions, and main effects of the PSEQ with other clinical measures and demographic data. Past studies established the PSEQ is associated with other clinical measures, but none have established cut-offs that clinicians could use when assessing and treating a patient with chronic pain. Results of this study can help inform clinicians of the use of the PSEQ as a predictive and benchmarking measure.
PSEQ admission scores were moderately positively correlated with PSEQ discharge scores, meaning those starting with a greater PSEQ score at admission tended to have a greater PSEQ score at discharge and vice versa. This positive correlation was anticipated, as the rehabilitation program aims to increase patients’ self-efficacy through cognitive-behaviourally informed treatment. Furthermore, PSEQ admission and discharge scores were weakly to moderately negatively associated with the PCS, PDI, and TSK admission and discharge scores. These results align with those of previous studies.
12,20–22 Again, these associations were expected, as self-efficacy fundamentally opposes catastrophizing, kinesiophobia, and belief that one is disabled. Understandably, catastrophizing, kinesiophobia, and belief that one is disabled co-occur in patients with chronic pain, as discussed by Sullivan et al.
18 and Woby et al.
19 There was also a significant between-subjects effect of the PSEQ sub-group levels, indicating patients of each sub-group were truly different from one another. Also, all four measures changed both statistically and clinically significantly from admission to discharge at the interdisciplinary rehabilitation program.
Interactions and main effects
Importantly, the PSEQ sub-groups interacted with the two clinical measures: PSEQ and TSK. The interaction between the sub-groups and the PSEQ revealed change in an individual’s self-efficacy score from admission to discharge was dependent on their level of self-efficacy at admission. As previously mentioned, the mean differences of PSEQ scores between admission and discharge of patients with low PSEQ were 18.01 (
p < 0.001), medium PSEQ was 11.11 (
p < 0.001), and high PSEQ was 5.44 (
p = 0.006), with the sub-groups determined by the patients’ PSEQ scores at admission. These results indicated those in the low PSEQ group made the greatest increase from admission to discharge, those in the medium PSEQ group made a moderate increase from admission to discharge, and those in the high PSEQ group made the smallest increase from admission to discharge. Results also showed patients with low PSEQ had a mean difference twice that required for MCID, those with medium PSEQ had a mean difference well above the MCID threshold, and those with high PSEQ scores had a mean difference below the MCID threshold.
36 This understanding of the different possibilities of improvement between PSEQ levels can be very informative to clinicians treating individuals with chronic pain. This differential change could show the program is effective in helping those most adversely affected or simply that those with higher self-efficacy have less possible room for improvement.
The interaction between the sub-groups and the TSK revealed the reduction of an individual’s kinesiophobia score from admission to discharge depended on their self-efficacy at admission. Results showed those in the low and medium PSEQ sub-groups had similar TSK scores at admission, while those in the high PSEQ group significantly differed from both low and medium groups. Interestingly, at discharge, the medium and high PSEQ groups had similar TSK scores, while those in the low PSEQ group made significantly less improvement than the medium group.
The PSEQ sub-groups did not interact with the PCS or the PDI scores. This may simply mean that catastrophizing and belief that one is disabled are not wholly dependent on self-efficacy, but that does not mean they are not related. The relationship between PSEQ sub-groups and the PCS or PDI can be an avenue for future research, and a test of interactions with the sub-groups shown here should be replicated.
In addition to the clinical measures’ interactions with the three PSEQ levels, both the PCS and the TSK significantly interacted with the referral source. The interaction between referral source and PCS scores revealed that both Veterans and civilians had similar PCS admission scores; however, Veterans had lower PCS scores than civilians at discharge. Additionally, the interaction between referral source and TSK scores revealed that Veterans had larger TSK admission scores than civilians; however, Veterans had lower TSK discharge scores than civilians. In these two interactions, Veterans experienced significantly greater improvements than civilians following treatment. These results supported those of Jomy and Hapidou
38 in that Veterans experienced significantly greater differences between admission and discharge in pain catastrophizing, kinesiophobia, and a number of other measures compared with civilians.
The final interaction involves gender and TSK scores at admission and discharge. Males had significantly larger TSK admission scores than females; however, at discharge, both males and females had similar scores, meaning that while both males and females had significant improvements, males had greater improvements on the TSK compared to females. The role of gender on psychological factors continues to be a source of debate in the field, as beliefs tend to elicit different coping behaviours in males and females. These results support those of Bränström and Fahlström,
39 who found men had higher levels of kinesiophobia than women. Studies also found men tend to report greater fear than women, despite women reporting greater pain and anxiety. These findings suggest that fear influences the types of coping behaviours in which one engages, which differ by gender.
40Limitations and future studies
A limitation of this study was the use of a convenience sample of patients with complete data sets to help establish a preliminary estimate of clinical cut-offs for the PSEQ in the literature. Future research should refine these estimates using a broad sampling of diverse patient populations across various settings to also understand any between and within group differences in pain self-efficacy and any interaction between these factors and patient outcomes (e.g., program completers vs. non-completers).
In addition, a caveat to these findings is that participants were from heterogeneous groups based on the context that resulted in the onset of their chronic pain problem, and with associated third-party involvement (e.g., MVA vs. Canadian Armed Forces Veteran). Veterans also demonstrated a greater level of improvement relative to civilians, raising questions about between-group differences in these populations. While there are shared factors, such as perceived injustice,
41 underlying the chronic pain-related outcomes in Veterans and civilians, future research should investigate between-group differences in pain self-efficacy and other pain-related factors associated with outcome. As this was simply an analysis of demographics, the authors recommend further research on the differentiation between Veterans’ and civilians’ psychological improvement.
Another important change that should be implemented in future studies is the analysis of PSEQ cut-offs in relation to occupational measures. Due to the large number of patients with chronic pain not returning to work, it would be important to understand whether a specific sub-group PSEQ level is linked to return to work.
23 If so, clinicians can identify patients that need more, or less, focus on specific treatments to increase self-efficacy and encourage returning to work. Thus, it is important to validate the results found with PSEQ cut-offs within this study by replicating this study with larger sample sets, data sets, and clinical measures with the ability to run multivariate ANOVA models.