Skip to main content

Table 4 Predictive utility of psychological factors on the outcome after surgical treatment for shoulder complaints. Bold font indicates high quality studies based on the SIGN review

From: Do psychological factors affect outcomes in musculoskeletal shoulder disorders? A systematic review

Authors

Quality

Outcome

Beliefs

Cognitive Style

Affect: Distress

Effect

1.Self efficacy / Coping

2. Expectation of recovery

3. Catastrophizing

4. Avoidance Coping Style

5. Depression

6. Anxiety / Worry/ Fear

Cho et al. 2015 [52]

(+)

Pain (VAS)

Pain And function (UCLA)

Pain and function (ASES)

    

+

+

+

+

+

+

Twelve months follow-up association in the multivariate linear regression analysis HADS-D with VAS − 0.073 (CI − 0.298 - 0.152), with UCLA score − 0.027 (− 0.565–0.511), ASES score − 0.235 (− 1.49–1.96). HADS-A with VAS 0.12 (− 0.05–0.28), UCLA − 0.09 (− 0.49–0.31), ASES − 0.62 (− 1.91–0.67).

Dambreville et al. 2007 [53]

(+)

Pain (VAS)

    

+

–

Preoperative depression (HADS) associated with pain at one month in a multivariate analysis (p = 0.03), not significant in postoperative pain; Anxiety (HADS) N.S.

Dekker et al. 2016 [54]

(++)

Pain (VAS)

Shoulder Pain and function (OSS)

    

+

+

 

Preoperative depression score revealed a strong negative correlation between preoperative HADS score and 6-week OSS (r = −  0.490, p < .01), HADS and 6-month OSS (r = −  0.626, p < .01) and HADS and 6-month satisfaction (r = − 0.259, p < .05). There as strong positive correlation (r = − 0.508, p = 0.01) between HADS score and 6-month pain scores.

George et al 2016 [56]

(++)

Pain (BPI)

Shoulder disability (QuickDASH)

  

+

+

+

-

+

+

Additional analysis using the risk groups for George et al. 2015 and Simon et al. 2020 (included in this review). Strong statistical evidence was found for ADRB2 and depressive symptoms for postoperative course (pain and disability), and GCH1 and anxiety symptoms for 12-month pain-intensity outcome. Interactions involving inflammatory genes with strong statistical evidence for the 12-month postoperative course outcome were: two different IL6 single-nucleotide polymorphisms and pain catastrophizing, and IL6 and depressive symptoms; KCNS1 and kinesiophobia for preoperative pain intensity but not for postoperative pain.

George et al. 2008 [55]

(+)

Pain (BPI)

  

+

–

  

Postoperative pain measured by BPI > 4 points.

Baseline PCS was associated with baseline pain, PCS baseline high score and low-COMT-phenotype the relative risk of high postoperative shoulder pain 6.8 (CI 2.8–16.7); Fear of pain or kinesiophobia were not associated with baseline pain or postoperative outcome (FPQ-III and TSK-11), however postoperative outcome was not systematically analysed.

George et al. 2015 [57]

(++)

Recovery (aADI)

  

+

+

  

Additional analysis using the risk groups for George et al. 2016 and Simon et al. 2020 (included in this review). Additional analysis using the risk groups identified in Simon 2020 (included in this review). Pain recovery was defined by: current pain intensity at VAS 0/10 and worst pain intensity 2/10. PCS, the catastrophizing high risk subgroup (combination of COMT and PCS score) were less likely to recover at 12 months (HR 0.51, P = 0.002); FABQ-score high risk subgroup (combination of COMT and FABQ-score) was less likely to recover at 12 months (HR 0.69, p = 0.043).

Henn et al. 2007 [59]

(+)

Pain (VAS)

Shoulder function (SST)

Shoulder disability (DASH)

Physical and mental health

(SF-36)

 

+

+

+

+

    

Preoperative expectation regarding the treatment (MODEM questionnaire): 6 questions, mean score: expectations were a significant independent predictor of better postoperative outcome scores (VAS (Beta 9.91, p = 0.005), DASH (Beta 11.93, p = <  0.001), SF-36, SST (Beta 15.34, p < 0.001)) at 12 months; Workers compensation in the multivariate model significant for VAS (Beta − 12.88, p = 0.009), DASH (Beta −9.12, p = 0.011), SST (Beta − 1.33, p = 0.038), SF-36.

Jain et al. 2018 [60]

(++)

Pain and disability (SPADI)

   

+

–

 

Linear mixed prediction models incorporating a covariance structure using all available follow-up time points (3, 6, 12, and 18 months) for a given patient. Higher FABQ physical activity score predicted higher SPADI scores (worse shoulder pain and function), p for interaction =0.001. Mental Health Inventory (MHI-5, distress) N.S.

Koorevaar et al. 2018 [61]

(+)

Shoulder disability (DASH)

    

–

 

Additional analysis using the risk groups for Korevaar et al. 2016 (included in this review). Comparison of group 1 (≥1 psychological disorder before and 12 months after surgery n = 32) and group 2 (no psychological disorders, n = 110).

DASH scores before (Group 1 55.5 [SD 19.8], Group 2 35.3 [SD 21.2], p < 0.001) and 12 months after shoulder surgery (Group 1 34.8 [SD 20.5], (Group 2 12.1 [SD 12.1], p < 0.001) were significantly higher in patients with symptoms of psychological disorders. Change of DASH score (p = 0.559) and MCID (% complete recovery, p = 0.284) were not different between the two groups. No adjustment for differences in baseline variables.

Koorevaar et al. 2016 [62]

(++)

Shoulder disability (DASH)

    

+

+

Additional analysis using the risk groups for Korevaar et al. 2018 (included in this review). Preoperative 4DSQ (distress, depression, anxiety, and somatization) was adjusted for age, gender and preoperative DASH score, associated with less of an improvement in DASH score.

Lau et al. 2019 [63]

(+)

Pain and function (ASES)

    

+

+

Same patient samples as Lau et al.2019 (included in this review). A higher score of depression or anxiety related to the

shoulder had a negative correlation with the postoperative (r = −0.31, p = 0.0001; and r = − 0.31, p = 0.0003, respectively) ASES scores, but a positive correlation

(r = 0.50, p < 0.0001; and r = 0.43, p < 0.0001, respectively) with the change in ASES score (pre to post operatively). No multivariate analysis and no adjustment for other factors.

Lau et al. 2020 [64]

(+)

Pain and function (ASES)

    

+

+

Same patient samples as Lau et al.2019 (included in this review). Subjects were classified as those with diagnosed clinical depression/anxiety and those with symptoms but no diagnosis. Regardless classification, there was a strong association for depression between improvement in ASES scores and changes in shoulder-related depression (r = 0.68 [with clinical diagnosis], r = 0.75 [without clinical diagnosis]). Regarding anxiety, there was a moderate association between improvement in ASES scores and changes in shoulder-related depression (r = 0.56 [with clinical diagnosis], r = 0.74 [without clinical diagnosis]. No multivariate analysis and no adjustment for other factors.

Oh et al.72012 [65]

(++)

Shoulder function (SST)

Shoulder function (Improvement Constant-Murley score)

Physical and mental health

(SF-36)

 

+

+

+

    

Patients were classified into low (33%), middle (33%), and high (33%) expectation or concern groups (based on mean expectation (MODEMS score) or concern score).

High-expectation group more improvement on SST (p = 0.24), Constant Murley scores (P < .001), and the SF-36 Physical Function (P = 0.006) compared to low expectation group.

High-concern group no significant improvement compared with low-concern group on SST (p = 0.9), Constant Murley scores (p = 0.7), and SF-36 physical function (p = 0.4).

Potter et al. 2015 [66]

(+)

Pain (VAS)

Shoulder function (SST)

Pain and function (ASES)

    

-

-

-

 

Score stratified based on Distress Risk Assessment Method. No significant differences between group with preop distress and those non-distressed.

VAS MCID in non-distressed group 59% and in distressed group 81% (OR, 2.91; 95% CI, 0.92–9.14; p = 0.06).

SST MCID in non-distressed 89% and distressed 81% (OR, 0.54; 95% CI, 0.14–2.07; p = 0.36).

ASES MCID in non-distressed 86% and distressed 88% (OR, 1.21; 95% CI, 0.28–5.32; p = 0.80).

Ravindra et al. 2018 [67]

(+)

Pain (VAS)

Pain and function (ASES)

    

-

-

 

Correlation coefficients were calculated for VAS and ASES at 1 year for the following independent variables: preoperative demographic factors, MRI tear characteristics. Correlation coefficients were calculated for preoperative VAS scores and ASES and WORC, SST, and SF-36 scores

Significant correlation found for higher 1-year VAS scores and higher preoperative VAS pain scores, narcotic use, and low WORC scores (both composite and emotion). Correlation with higher ASES scores at 1-year was found for higher preoperative VAS scores and increased supraspinatus atrophy.

Simon et al. 2020 [58]

(++)

Active shoulder range of motion (flexion and abduction)

Movement evoked pain (NPRS)

    

-

+

-

-

Additional analysis using the risk groups identified in George et al. 2015 and 2016 (included in this review). There were no significant findings for psychological factors and active range of motion.

Depressive symptoms were found to mediate the causal pathway in the high-risk subgroup for increased movement-evoked pain intensity at 12 months (p = 0.038). The mediation effect accounts for 53% of the total effect of the high-risk group on 12-month movement-evoked pain.

Thorpe et al. 2018 [68]

(+)

Pain and function (ASES)

  

+

 

+

+

After adjustment for gender, workers compensation status, alcohol use and confidence in surgical outcome, cluster with poor psychological health was independently associated with worse ASES score at all time points (regression coefficient for ASES: 3 months after surgery −15 [95% CI, −23 to −8], p < 0.001); and 12 months after surgery −9 [95% CI, − 17 to − 1], p = 0.023).

ASES scores improved in both clusters from before surgery to 12 months after surgery equally (regression coefficient for ASES: cluster 2 31 [95% CI, 26–36], p < 0.001); cluster 1 31 [95% CI, 23–39], p < 0.001).

Valencia et al. 2014 [69]

(+)

Pain (BPI)

Shoulder disability (DASH)

  

-

+

 

-

+

 

PCS no significant correlation with 6 months pain, significant correlation with DASH at 6 months (r = 0.225); PHQ 9 not significant for pain but significant for disability (DASH, r = 0.287).

Woollard et al. 2017 [70]

(+)

Disability (aADI)

   

+

  

Criteria for functional disability postoperative: (1) Global rating of change ≥ + 5, (2) ≥17-point improvement on the WORC from baseline to 6-months postoperative.

Logistic regression model including (1) surgery on dominant shoulder, (2) work compensation status, (3) modified job duty, (4) baseline FABQ-work, internal rotation strength. FABQ-Work was associated with a lower success rate (OR 0.92, 95% CI 0.85–1.00). FABQ work subscale of ≤25 and surgery on the dominant shoulder were both strongly predictive of being a responder to surgery (FABQ work ≤25 points Beta 2.73, OR 15.29 (95% CI 2.30–101.9), p = 0.005)

Yeoman et al. 2012 [71]

(+)

Pain (VAS)

Shoulder pain and function (OSS)

    

-

-

-

-

HADS (> 7 points) no significant difference in the postoperative function and VAS in the depression versus the no depression group (6 weeks follow-up); HADS (> 7 points) no significant difference in the postoperative function and VAS in the anxiety versus the no anxiety group (6 weeks follow-up).

  1. Overall study quality, high (++), moderate (+), low (0)
  2. + Statistically significant relationship was found; − no statistically significant relationship was found
  3. a ADI Author defined instrument, 4DSQ Four-Dimensional Symptom Questionnaire, ASES the American Shoulder and Elbow Surgeons’ Scale, BPI Brief Pain Inventory, CBT cognitive behavioural therapy approach, DASH (and quickDASH) (Quick) Disability of the Arm, Shoulder and Hand Questionnaire, EQ-5D EuroQol- 5 Dimension, FABQ Fear Avoidance Beliefs Questionnaire (FABQ-P: physical activity subscale; FABQ-W, work subscale), FLEX-SF Flexilevel Scale of Shoulder Function, GE graded exercise, HADS Hospital Anxiety and Depression Scale, HSCL-25 Hopkins Symptoms Checklist, MCID A minimal clinically important difference, MODEMS Musculoskeletal Outcomes Data Evaluation and Management System, NPRS Numeric Pain Rating Scale, NRS Numeric Rating Scale, OSS Oxford Shoulder Score, PCCL Pain Coping and Cognition List, PCS Pain Catastrophizing Scale, PSFS Patient Specific Functional Scale, PT physical therapy, RCT randomized controlled trial, SDQ Shoulder Disability Questionnaire, SF-36 Short Form Survey, SPADI Shoulder Pain and Disability Index, SST Simple Shoulder Test, TSK Tampa Scale of Kinesiophobia, UC usual care, UCLA Scale The University of California at Los Angeles Shoulder Score, VAS Visual Analog Scale, WORC Western Ontario Rotator Cuff Index