Versus Corticosteroid for Treatment of Shoulder Pain: A Systematic Review and Meta-Analysis
Published Online: May 16, 2014
Abstract
Objective
To compare the treatment efficacy between corticosteroid injection and nonsteroidal anti-inflammatory drugs (NSAIDs) for patients with shoulder pain.
Data Sources
PubMed and EMBASE databases were searched from inception to January 2014. Reference lists of the retrieved studies were additionally scrutinized.
Study Selection
Randomized controlled trials (RCTs) comparing corticosteroid injection with NSAIDs for treatment of shoulder pain were included. The primary outcome was remission, and the secondary outcomes were pain relief and improvement of range of active abduction. Study selection was conducted by 2 researchers independently. Any disagreements were solved by discussion and confirmed by the third reviewer.
Data Extraction
Two reviewers independently conducted data extraction and the quality assessment. Data regarding patients, intervention, control, and outcomes were extracted from the included trials.
Data Synthesis
Six high-quality RCTs of 267 patients meeting the inclusion criteria were included. For an outcome of remission, NSAIDs were less effective than corticosteroid in 4 or 6 weeks (relative risk, .64; 95% confidence interval, .45–.92). NSAIDs did not significantly differ with corticosteroid in pain relief and improvement of range of active abduction.
Conclusions
Current meta-analysis suggests that NSAIDs are less effective than corticosteroid in achieving remission in patients with shoulder pain at 4 or 6 weeks after treatment. Considering the limited number of studies and small size of each trial, the results should be interpreted with caution, and more high-quality RCTs are encouraged.
List of abbreviations:
CI (confidence interval), NSAID (nonsteroidal anti-inflammatory drug), RCT (randomized controlled trial), SMD(standardized mean difference)
Shoulder pain is a common musculoskeletal disorder with a prevalence varying from 7.5% to 21%.1, 2, 3, 4 Various clinical disorders, such as rotator cuff tendinitis, adhesive capsulitis, shoulder impingement syndrome, and acromioclavicular joint disease, are diagnosed in patients with a painful shoulder.3 Patients with shoulder pain usually have active and passive motion restriction and function limitation.5 Since shoulder mobility affects daily life and pain influences sleeping, some patients also have anxiety and distress. Although shoulder pain is often described as having a self-limited course with recovery occurring gradually within 1 to 3 years, intervention can help in rehabilitation and improve outcomes.6
Many treatments attempt to relieve shoulder pain, including surgical and nonsurgical options (rest, ice, nonsteroidal anti-inflammatory drugs [NSAIDs], corticosteroid injection, physical treatment, electromagnetic radiation, laser).7, 8, 9 NSAIDs and corticosteroid are the most commonly used treatments, but the most effective treatment option remains to be established. A meta-analysis involving only 3 trials was conducted by Arroll et al10 and published in 2005, and no significant difference between corticosteroid and NSAIDs for treatment of shoulder pain was verified by the results. Moreover, this meta-analysis assessed only the clinical outcome of remission, whereas other outcomes, such as pain and range of motion, were not studied. With accumulating evidence of 3 additionally relevant high-quality randomized controlled trials (RCTs) published,11, 12, 13 we therefore performed a systematic review and meta-analysis of RCTs to investigate the treatment efficacy of corticosteroid and NSAIDs for patients with shoulder pain.
Methods
Literature search and inclusion criteria
We conducted this meta-analysis in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement.14 PubMed and EMBASE databases were searched by 2 authors independently (X.-Q.Z., K.L.) to identify the relevant RCTs (from inception to January 2014). The search strategy was set by combining the subject term and free term with the following words: ((“shoulder pain” OR “shoulder impingement syndrome” OR “rotator cuff” OR “bursitis” OR “adhesive capsulitis” OR “frozen shoulder”) OR (“shoulder pain”[MeSH] OR “shoulder impingement syndrome”[MeSH])) AND (steroid$ OR corticosteroid$ OR “adrenal cortex hormone”[MeSH]) AND (“anti-inflammatory agents, nonsteroidal” [MeSH] OR “nonsteroid antiinflammatory agent” [MeSH]) OR acetylsalicyl∗ OR carbasalaatcalcium OR diflunisal OR aceclofenac OR alclofenac OR diclofenac OR indometacin OR sulindac OR meloxicam OR piroxicam OR dexibuprofen OR dexketoprofen OR fenoprofen OR flurbiprofen OR ibuprofen OR ketoprofen OR naproxen OR tiapro* OR metamizol OR phenylbutazone OR phenazone OR propyphenazone OR celecoxib OR etoricoxib OR nabumeton OR parecoxib). The study design was restricted to RCTs, and the detailed limitation for them is presented in appendix 1. Additionally, the reference lists of retrieved studies were also rechecked manually to identify other eligible trials. This process was repeated until no further relevant studies were detected. RCTs included met the following criteria: (1) patients—people with shoulder pain; (2) intervention/control—NSAIDs versus corticosteroid; (3) outcomes—remission, pain relief, and motion improvement; and (4) study design—RCT.
Data extraction and outcome measures
Two reviewers (X.-Q.Z., K.L.) independently extracted the following information from the included studies: first author, publication year, number of patients (NSAIDs/corticosteroid), patient characteristics, intervention of NSAIDs, intervention of corticosteroid, Jadad score, remission, pain relief, active abduction, definition of remission, and definition of pain. The extracted data were input into an Excela file and checked by the third author (Y.-D.W.). Disagreements were solved by discussion. The primary outcome was remission, and secondary outcomes were pain relief and improvement of range of active abduction.
Quality assessment
The Jadad Scale was used to evaluate the methodological quality of each RCT included in this meta-analysis.15 This scale consists of 3 evaluation elements: randomization (0–2 points), blinding (0–2 points), and dropouts and withdrawals (0–1 points). One point would be allocated to each element if it has been mentioned in the article, and an additional point would be given if the methods of randomization or blinding, or both, have been detailed and appropriately described. If methods of randomization or blinding, or both, were inappropriate, or dropouts and withdrawals had not been recorded, then 1 point was deducted. The score of the Jadad Scale varies from 0 to 5 points. An article with a Jadad score ≤2 is considered to be of low quality, whereas a study with a Jadad score ≥3 is thought to be of high quality.16
Statistical analysis
The difference for a dichotomous variable was expressed as relative risk and 95% confidence interval (CI). Because pain relief and range of active abduction differed across studies, pain relief and range of active abduction were expressed as standardized mean difference (SMD) with 95% CI. In some studies, the SD was transformed from standard error or median, and reduction of pain and improvement of range of active abduction were calculated through subtraction if they were not available according to the methods introduced by Hozo et al17 and Cochrane Handbook for Systematic Reviews of Interventions.18 Heterogeneity across studies was assessed by the Q statistic and the I2 statistic, the latter indicating inconsistency by a quantitative number. Studies with an I2 statistic of 25% to 50% were considered to have low heterogeneity; studies with an I2 statistic of 50% to 75%, moderate heterogeneity; and studies with an I2 statistic >75%, high heterogeneity. The models were determined by the Q statistic and the I2 statistic: fixed-effects models were used if no or low heterogeneity was confirmed by the I2 statistic and heterogeneity (P>.1) verified by the Q statistic; otherwise random-effects models were used. In the analysis, the patients who received corticosteroid injections were regarded as the control group, and those who took NSAIDs were considered the experimental group. Because of the limited number of trials included (<10), publication bias was not evaluated.19 A 2-sided P value <.05 was considered statistically significant. All statistical analysis was performed with STATA 12.0.b
Results
Study identification and selection
A total of 99 articles were identified from the initial search. Fifteen articles were excluded because of duplication. Based on inclusion criteria, 74 of 84 studies were excluded after reading titles and abstracts. The remaining 10 full articles were reviewed for a more detailed evaluation. Two studies20, 21 were excluded because physiotherapy was also used in the NSAIDs group, which could confuse a comparison between NSAIDs and corticosteroid. One study22 was excluded because it was a meta-analysis of 3 trials. Another study23 was excluded because it was an article from a meeting without available data, and no data were acquired from the author. Six RCTs were finally included in this meta-analysis.11, 12, 13,24, 25, 26 The selection process is presented in figure 1.
Fig 1
Flow diagram of included and excluded studies.
Study characteristics and quality assessment
The baseline characteristics of the 6 included RCTs in this meta-analysis are shown in table 1. These studies were published between 1986 and 2013. The sample size varied from 30 to 58 patients (267 total in all 6 studies). Among the 6 studies, 2 studies focused on patients with rotator cuff tendonitis24, 26; 2 studies focused on patients with shoulder impingement syndrome11, 12; 1 study focused on diabetic patients with frozen shoulder13; and 1 study focused on patients with shoulder pain.25 Patients in these trials were randomly allocated into 2 groups: 1 group received corticosteroid injection, and the other received NSAIDs. Jadad scores of the 6 included studies varied from 1 to 5, with 5 of the studies considered to be of high quality according to quality assessment.
| Study | No. of Patients (Corticosteroid/NSAID) | Population Characteristics | Intervention | Study Design/Jadad Score | |
|---|---|---|---|---|---|
| Corticosteroid Group | NSAID Group | ||||
| White et al,261986 | 15/15 | Rotator cuff tendonitis (<3mo) | Subacromial injection of 40mg triamcinolone acetonide + placebo indomethacin tablets 4 times daily + home exercise program | 25mg indomethacin 4 times daily + placebo (1mL saline) injection + home exercise program | RCT/1 |
| Petri et al,251987 | 25/25 | Painful shoulder (>3mo) | 3mL of 1% lidocaine + 1mL of 40mg/mL triamcinolone + placebo pill twice a day + home exercise for 30d | 4mL of 1% lidocaine + naproxen (500mg) twice a day + home exercise for 30d | Double-blind RCT/5 |
| Adebajo et al,241990 | 20/20 | Rotator cuff tendonitis (<3mo) | Diclofenac placebo tablets + subacromial injection of 2mL 0.5% lignocaine + 1mL of 80mg/mL triamcinolone hexacetonide | 50mg diclofenac 3 times a day for 28d + subacromial injection of 3mL of 0.5% lignocaine | Double-blind RCT/5 |
| Karthikeyan et al,122010 | 27/31 | Subacromial impingement (>3mo) | 40mg methylprednisolone + 5mL 1% lignocaine | 20mg tenoxican + 5mL 1% lignocaine | Double-blind RCT/5 |
| Min et al,112013 | 15/17 | Shoulder impingement syndrome | 6mL 1% lidocaine with epinephrine + 40mg triamcinolone | 6mL of 1% lidocaine with epinephrine + 60mg ketorolac | Double-blind RCT/5 |
| Dehghan et al,132013 | 29/28 | Frozen shoulder of diabetic patients | 500mg naproxen twice a day + single injection of 40mg triamcinolone + home exercise | 500mg naproxen twice a day + 2 naproxen tablets daily + home exercise | RCT/3 |
Primary outcome: remission
Four studies reported the primary outcome of remission. Details of remission and its definition are presented in table 2. The aggregated data produced by the fixed-effect model showed that NSAIDs were significantly less effective than corticosteroid (relative risk, .64; 95% CI, .45–.92) (fig 2), with low heterogeneity (I2=45.1%, P=.141).
| Study | Time (wk) | Pain Relief | Active Abduction (deg) | Improvement | Measurement of Pain | Definition of Improvement | |||
|---|---|---|---|---|---|---|---|---|---|
| NSAID | Steroid | NSAID | Steroid | NSAID | Steroid | ||||
| White et al,261986 | 6 | 5.5±8.3 | 4.3±5.2 | 16±45 | 30±37 | 10/15 | 9/15 | 9cm VAS | Global assessment score: Sum of patient's and physician's estimate of severity of pain and severity of motion deficit (0–9 points) |
| Petri et al,251987 | 4 | 1.71±1.55 | 2.04±1.55 | 50.04±55.8 | 56.16±43.2 | 5/25 | 7/25 | Linear scale from 0 to 5 level | Remission is defined as a perfect score in active abduction, pain, and limitation of function. |
| Adebajo et al,241990 | 4 | 3.6±2.99 | 4.95±3.31 | 46.8±25.22 | 50.4±36 | 6/20 | 14/20 | 10cm VAS | NA |
| Karthikeyan et al,122010 | 6 | NA | NA | NA | NA | 7/30 | 12/26 | Constant-Murley Shoulder Score | Patients' feeling |
| Min et al,112013 | 4 | 1.83±2.25 | 0.90±1.86 | 22.06±37.96 | 0±31.68 | NA | NA | VAS | NA |
| Dehghan et al,132013 | 6 | 2.57±2.32 | 3.31±2.19 | 46.3±34.9 | 65.3±32.9 | NA | NA | Pain assessment ruler | NA |
NOTE. Values are mean ± SD, n, or as otherwise indicated.
Abbreviations: NA, not available; VAS, visual analog scale.
Fig 2
Pooled estimate effect of NSAIDs versus corticosteroid for remission. Abbreviations: ID, identification; RR, relative risk.
Secondary outcomes
Five studies reported the pain score before and after treatment (4 or 6wk) or the change in the pain score. Four studies measured pain using a visual analog scale, and the other study used a linear scale from 0 to 5 levels (see table 2). Meta-analysis using the fixed-effect model showed that the effectiveness of NSAIDs and corticosteroid for relieving pain did not differ significantly (SMD, −.16; 95% CI, −.43 to .12), with little evidence of heterogeneity (I2=44.2%, P=.127) (fig 3).
Fig 3
Pooled estimate effect of NSAIDs versus corticosteroid for pain relief. Abbreviation: ID, identification.
Five studies reported the range of active abduction before and after treatment (4 or 6wk) or the range changes of active abduction, which was the most commonly reported index of motion. Range of motion was measured through a goniometer. The pooled analysis using the fixed-effects model showed that compared with corticosteroid, NSAIDs did not significantly improve active abduction (SMD=−.13; 95% CI, −.40 to .14), with low heterogeneity (I2=16.4%, P=.31) (fig 4).
Fig 4
Pooled estimate effect of NSAIDs versus corticosteroid for improvement of range of active abduction. Abbreviation: ID, identification.
Discussion
This meta-analysis of 6 RCTs comparing treatment effects of NSAIDs and corticosteroid on shoulder pain found that NSAIDs and corticosteroid treatment had similar effects on pain relief and active abduction in 4 or 6 weeks, but corticosteroid injection significantly achieved more remissions than NSAIDs in 4 or 6 weeks after treatment. The results show that the use of corticosteroid injections has short-term treatment advantages.
A previous meta-analysis10 did not find any superiority of corticosteroid injection or NSAIDs in the treatment of shoulder pain. A systematic review27 from The Cochrane Library concluded that corticosteroid might have an early benefit on adhesive capsulitis; however, data from only 2 trials were available, and consequently it was insufficient for pooling. Only 3 trials24, 25, 26 were included in both of these articles,10, 27 and no specific suggestions were made about the treatment of shoulder pain by corticosteroid injection or NSAIDs. Additionally, only the outcome of remission was analyzed in the former meta-analysis.10 With more RCTs included and outcomes analyzed, the present meta-analysis showed that corticosteroid was more advantageous than NSAIDs in the short-term for the treatment of shoulder pain.
Pharmacologic mechanisms determine different treatment effects of corticosteroid and NSAIDs for the treatment of shoulder pain. With multiple pharmacologic actions, corticosteroid not only relieves pain but also exerts a positive influence on the patients' mood by both localized steroid injection and systemic steroid treatment. NSAIDs, however, are only able to alleviate pain.28 Additionally, researchers have tried to determine the optimum dose, frequency, volume, and type of corticosteroid, as well as anatomic site of needle placement, to achieve the best treatment effects with corticosteroid. Shah and Lewis29 found that up to 3 injections of corticosteroid reduced pain and improved range of motion for adhesive capsulitis in 6 to 16 weeks, and limited evidence existed for up to six injections. Yoon et al30 showed no significant difference between 20mg and 40mg of corticosteroid for adhesive capsulitis, suggesting that corticosteroid may not produce a dose-dependent effect. Hence, although corticosteroid was proved to be effective for shoulder pain, further investigation is needed to determine the dosage of corticosteroids that will provide the optimum effect with the fewest complications.
The endpoint was at 4 or 6 weeks in our analysis, which is a relatively short observation period compared with the course of shoulder pain. A limited number of high-quality studies evaluating the long-term effects of corticosteroid versus NSAIDs are available for meta-analysis. Dehghan et al13 reported no significant difference between patients treated with corticosteroid and those treated with NSAIDs for pain relief, external rotation, internal rotation, flexion, and abduction in the sixth month. Additionally, in an 8-month follow-up prospective study,31 steroid injection did not show any advantages over mobilization and ice therapy in the long term, while it could relieve pain and increase the range of motion in the early stage of disease. Compared with aceclofenac, an NSAID, corticosteroid injection was shown to be superior in the early-stage treatment of shoulder pain and to have similar efficacy to aceclofenac 24 weeks after treatment.32 Moreover, 2 systematic reviews27, 33 also suggested that corticosteroid had advantages in the short-term treatment, but not in the long-term treatment, of shoulder disorders. Another review34 also found that the advantages of corticosteroid injections for adhesive capsulitis decreased with time, from 6 to 52 weeks. To better evaluate different treatment options, more high-quality trials focusing on long-term effects are needed.
Although corticosteroid is widely used in clinical practice, a standard dose has not been recommended, and evidence that directly supports the use of corticosteroid is limited. The only advantage found for corticosteroid treatment in this meta-analysis was in the achievement of remissions; however, a trend of superiority of corticosteroid could be seen in the alleviation of pain and improvement of range of motion. Therefore, the findings of this meta-analysis suggest that corticosteroid benefits patients with shoulder pain more than NSAIDs in the short-term. In addition, treatment with corticosteroid usually requires a few injections, while treatment with NSAIDs requires continuous intake of tablets for a relatively long time. However, the potential complications of corticosteroid should also be taken into account, especially for the elderly. Therefore, our study implies that corticosteroid injection in the early period and NSAID treatment in the later period might be the best approach to the treatment of shoulder pain.
Study limitations
Limitations exist in this meta-analysis. Only 6 RCTs with a total of 267 patients were included in the analysis. Additionally, only 3 clinical outcomes of efficacy were analyzed and safety of corticosteroid and NSAIDs for treatment of shoulder pain in the 6 RCTs included were not analyzed. Moreover, the types of corticosteroid and NSAIDs used varied across the studies. Although no heterogeneity or low heterogeneity existed across the included RCTs, more relevant studies of high quality would make the results more robust. Hence, more high-quality RCTs evaluating the short-term and long-term effects of corticosteroid versus NSAIDs are needed.
Conclusions
Our results suggest that compared with NSAIDs, corticosteroid is associated with a significant increase of remissions in 4 or 6 weeks after treatment; however, corticosteroid was not found to be significantly superior to NSAIDs for pain relief and improvement of range of active abduction. Hence, our findings partially support the short-term benefit of corticosteroid usage for shoulder pain. More high-quality, well-designed trials are required.
Suppliers
- a.Microsoft Corp, One Microsoft Way, Redmond, WA 98052-6399.
- b.StataCorp LP, 4905 Lakeway Dr, College Station, TX 77845-4512.
Appendix 1. Detailed Limitation for RCTs
(randomized controlled trial[pt]OR randomized controlled trials[mh]OR random allocation[mh] OR random allocat* [tw] OR randomly allocat* [tw] OR double-blind method[mh] OR single-blind method[mh] OR double blind* [tw] OR single blind[tw] OR triple blind*[tw]OR clinical trial [pt] OR clinical trials[mh]).
References
- Urwin, M., Symmons, D., Allison, T. et al. Estimating the burden of musculoskeletal disorders in the community: the comparative prevalence of symptoms at different anatomical sites, and the relation to social deprivation.Ann Rheum Dis. 1998; 57: 649–655
- Palmer, K.T., Cooper, C., Walker-Bone, K., Syddall, H., and Coggon, D. Use of keyboards and symptoms in the neck and arm: evidence from a national survey. Occup Med (Lond). 2001; 51: 392–395
- Ostor, A.J.K. Diagnosis and relation to general health of shoulder disorders presenting to primary care.Rheumatology. 2005; 44: 800–805
- Chard, M.D., Hazleman, R., Hazleman, B.L., King, R.H., and Reiss, B.B. Shoulder disorders in the elderly: a community survey. Arthritis Rheum. 1991; 34: 766–769
- Badcock, L.J., Lewis, M., Hay, E.M., McCarney, R., and Croft, P.R. Chronic shoulder pain in the community: a syndrome of disability or distress?. Ann Rheum Dis. 2002; 61: 128–131
- Silman, A. and Hochberg, M. Epidemiology of the rheumatic diseases. 2nd ed. Oxford Univ Pr, Oxford; 2001
- Kelle, B. and Kozanoglu, E. Low-level laser and local corticosteroid injection in the treatment of subacromial impingement syndrome: a controlled clinical trial. Clin Rehabil. 2014 Feb 11; ([Epub ahead of print])
- Hsieh, L.F., Hsu, W.C., Lin, Y.J., Wu, S.H., Chang, K.C., and Chang, H.L. Is ultrasound-guided injection more effective in chronic subacromial bursitis?. Med Sci Sports Exerc. 2013; 45: 2205–2213
- Bateman, M., McClymont, S., and Hinchliffe, S.R. The effectiveness and cost of corticosteroid injection and physiotherapy in the treatment of frozen shoulder—a single-centre service evaluation. Clin Rheumatol. 2014 Feb 2; ([Epub ahead of print])
- Arroll, B. and Goodyear-Smith, F. Corticosteroid injections for painful shoulder: a meta-analysis. Br J Gen Pract. 2005; 55: 224–228
- Min, K.S., St. Pierre, P., Ryan, P.M., Marchant, B.G., Wilson, C.J., and Arrington, E.D. A double-blind randomized controlled trial comparing the effects of subacromial injection with corticosteroid versus NSAID in patients with shoulder impingement syndrome. J Shoulder Elbow Surg. 2013; 22: 595–601
- Karthikeyan, S., Kwong, H.T., Upadhyay, P.K., Parsons, N., Drew, S.J., and Griffin, D. A double-blind randomised controlled study comparing subacromial injection of tenoxicam or methylprednisolone in patients with subacromial impingement. J Bone Joint Surg Br. 2010; 92: 77–82
- Dehghan, A., Pishgooei, N., Salami, M.A. et al. Comparison between NSAIDs and intra-articular corticosteroid injection in frozen shoulder of diabetic patients; a randomized clinical trial. Exp Clin Endocrinol Diabetes. 2013;121: 75–79
- Moher, D., Liberati, A., Tetzlaff, J., Altman, D.G., and PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med. 2009; 151: 264–269 (W64)
- Jadad, A.R., Moore, R.A., Carroll, D. et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary?. Control Clin Trials. 1996; 17: 1–12
- Kjaergard, L.L., Villumsen, J., and Gluud, C. Reported methodologic quality and discrepancies between large and small randomized trials in meta-analyses. Ann Intern Med. 2001; 135: 982–989
- Hozo, S.P., Djulbegovic, B., and Hozo, I. Estimating the mean and variance from the median, range, and the size of a sample. BMC Med Res Methodol. 2005; 5: 13
- Higgins, J.P. and Green, S. Cochrane handbook for systematic reviews of interventions. Wiley Online Library.2008; (Available at: http://www.cochrane.org/handbook. Accessed December 2013.)
- Song, F., Eastwood, A.J., Gilbody, S., Duley, L., and Sutton, A.J. Publication and related biases. Health Technol Assess. 2000; 4: 1–115
- Arslan, S. and Celiker, R. Comparison of the efficacy of local corticosteroid injection and physical therapy for the treatment of adhesive capsulitis. Rheumatol Int. 2001; 21: 20–23
- Saeidian, S.R., Hemmati, A.A., and Haghighi, M.H. Pain relieving effect of short-course, pulse prednisolone in managing frozen shoulder. J Pain Palliat Care Pharmacother. 2007; 21: 27–30
- Chavero Carrasco, V., Salguero Molpeceres, O., Lopez Ruano, P., and Alvarez Montero, S. [Treatment of shoulder pain; non steroidal anti-inflammatory drug (NSAID) or intraarticular injection?] [Spanish]. Rev Med Fam Comun. 2002; 12: 341–343
- Shirokov, V., Kudryavtseva, M., Krivtsova, I., and Parchomenko, E. Efficiency of using periarticular injection of betamethasone acetate/betamethasone sodium phosphate and peroral etoricoxib in adhesive capsulitis therapy. Eur J Pain Suppl. 2011; 5: 225–226
- Adebajo, A.O., Nash, P., and Hazleman, B.L. A prospective double blind dummy placebo controlled study comparing triamcinolone hexacetonide injection with oral diclofenac 50 mg TDS in patients with rotator cuff tendinitis. J Rheumatol. 1990; 17: 1207–1210
- Petri, M., Dobrow, R., Neiman, R., Whiting-O'Keefe, Q., and Seaman, W.E. Randomized, double-blind, placebo-controlled study of the treatment of the painful shoulder. Arthritis Rheum. 1987; 30: 1040–1045
- White, R.H., Paull, D.M., and Fleming, K.W. Rotator cuff tendinitis: comparison of subacromial injection of a long acting corticosteroid versus oral indomethacin therapy. J Rheumatol. 1986; 13: 608–613
- Buchbinder, R., Green, S., and Youd, J.M. Corticosteroid injections for shoulder pain. Cochrane Database Syst Rev. 2003; : CD004016
- Villemure, C. and Bushnell, M.C. The effects of the steroid androstadienone and pleasant odorants on the mood and pain perception of men and women. Eur J Pain. 2007; 11: 181–191
- Shah, N. and Lewis, M. Shoulder adhesive capsulitis: systematic review of randomised trials using multiple corticosteroid injections. Br J Gen Pract. 2007; 57: 662–667
- Yoon, S.H., Lee, H.Y., Lee, H.J., and Kwack, K.S. Optimal dose of intra-articular corticosteroids for adhesive capsulitis: a randomized, triple-blind, placebo-controlled trial. Am J Sports Med. 2013; 41: 1133–1139
- Bulgen, D.Y., Binder, A.I., Hazleman, B.L., Dutton, J., and Roberts, S. Frozen shoulder: prospective clinical study with an evaluation of three treatment regimens. Ann Rheum Dis. 1984; 43: 353–360
- Shin, S.J. and Lee, S.Y. Efficacies of corticosteroid injection at different sites of the shoulder for the treatment of adhesive capsulitis. J Shoulder Elbow Surg. 2013; 22: 521–527
- Coombes, B.K., Bisset, L., and Vicenzino, B. Efficacy and safety of corticosteroid injections and other injections for management of tendinopathy: a systematic review of randomised controlled trials. Lancet. 2010;376: 1751–1767
- Blanchard, V., Barr, S., and Cerisola, F.L. The effectiveness of corticosteroid injections compared with physiotherapeutic interventions for adhesive capsulitis: a systematic review. Physiotherapy. 2010; 96: 95–107
