Outcome measures and treatment of shoulder capsulitis (frozen shoulder) by corticosteroid injections
Not peer reviewed
MetadataShow full item record
- To examine intertester reliability of measuring passive range of motion (PROM) bilaterally using plurimeter in patients with adhesive shoulder capsulitis over an 8-weeks period and examine whether the measurement error remained the same.
- Investigate the effect, if any, of multiple corticosteroid injections with distension as compared to corticosteroid injections alone and to treatment-as-usual.
- Determine whether treatment outcome can be predicted by subjective health complaints and neuroticism in patients with frozen shoulder as measured by Shoulder Pain and Disability Index (SPADI) and change in SPADI.
Design/method: The first study is a prospective intertester reliability study for measurement of PROM in the shoulder of 50 patients with frozen shoulder. Two testers measured PROM with a plurimeter several times during an 8-week period. The second study is a randomised controlled single blinded three-armed trial comparing effect of two different interventions with treatment-as-usual in 106 patients. Treatment-as-usual in this scenario means any other conservative treatment like non-steroid antiinflammatory drugs, painkillers, physiotherapy and acupuncture but no steroids to be used orally or as injection. The intervention consisted of four intraarticular steroid injections with or without distension. SPADI was the primary outcome measure, Numerical Pain Rating Scale (NPRS) secondary, and PROM, the tertiary outcome measure. The third study is observational where we investigated whether comorbid factors measured with the questionnaires Subjective Health Complaints (SHC) and Neuroticism can predict outcome of the given treatment in 105 patients. We collected data from patients answering the questionnaires at baseline and at the end of 8-week clinical follow up.
Results: Study I: Intertester agreements varied from very good to excellent for PROM for all three time-points for the affected arm. Very reliable to excellent values were achieved for intraclass correlation coefficient (ICC 2.1). The measurement error remained the same throughout. Study II: At short-term (4 and 8 weeks) statistically significant differences (p<0.01) in change scores for SPADI, NPRS and PROM were observed when comparing those receiving corticosteroid injections, with or without distension, to treatment-as-usual. At long-term (12 months) there was no difference between the three groups for SPADI (p>0.05). A large effect size (ES) was observed between both injection groups and treatment-as-usual (ES 1.2) at short term. The effect size between the injection groups and treatment-as-usual was reduced to low (ES 0.3 and 0.4) at 12 months. The difference between the two injection groups at short-term (4-and 8 weeks) or long-term (12 months) was not statistically significant. All participants recovered. Study III: Little comorbidity was observed in the 105 patients included in the study, as measured with the questionnaire Subjective Health Complaints (SHC). Significant predictive power (p<0.001) was exhibited by the Pseudoneurology subscale in SHC for outcome at 8 weeks. All included patients scored within normal range on Neuroticism.
Conclusion: Study I: Intertester reliability between the two testers over a time-period of 8 weeks measuring PROM in patients with adhesive shoulder capsulitis with a plurimeter was very good. This method can reliably determine passive range of motion in this patient population and be a reliable outcome measure. Study II: This randomised controlled trial indicated that four serial injections with corticosteroid with or without distension during 8 weeks were better than treatmentas- usual in treatment of patients with adhesive shoulder capsulitis. However, no difference was found between any of the groups at 12 months, indicating that natural healing takes place independent of treatment. Study III: Comorbidity as measured by the Pseudoneurology subscale in the SHC questionnaire did predict the treatment outcome in frozen shoulder as measured by SPADI at 8 weeks, whereas when measured by change in SPADI from baseline to 8 weeks, it did not. Comorbidity may affect symptoms but do not predict the rate of recovery.