General practice

Comparison of physiotherapy, manipulation, and corticosteroid injection for treating shoulder complaints in general practice: randomised, single blind study

^a Department of General Practice, University of Groningen, Ant Deusinglaan 4, 9713 AW Groningen, Netherlands, ^b Department of Rehabilitation Medicine, University Hospital Groningen, 9700 RB, Groningen, Netherlands

Correspondence to: Dr J C Winters Nieuwe Schoolweg 2A, 9756 BB Glimmen, Netherlands.

	Abstract

Top Abstract Introduction Patients and methods Results Discussion References

Objective: To compare the efficacy of physiotherapy, manipulation, and corticosteroid injection for treating patients with shoulder complaints in general practice.
Design: Randomised, single blind study.
Setting: Seven general practices in the Netherlands.
Subjects: 198 patients with shoulder complaints, of whom 172 were divided, on the basis of physical examination, into two diagnostic groups: a shoulder girdle group (n = 58) and a synovial group (n = 114).
Interventions: Patients in the shoulder girdle group were randomised to manipulation or physiotherapy, and patients in the synovial group were randomised to corticosteroid injection, manipulation, or physiotherapy.
Main outcome measures: Duration of shoulder complaints analysed by survival analysis.
Results: In the shoulder girdle group duration of complaints was significantly shorter after manipulation compared with physiotherapy (P < 0.001). Also the number of patients reporting treatment failure was less with manipulation. In the synovial group duration of complaints was shortest after corticosteroid injection compared with manipulation and physiotherapy (P < 0.001). Drop out due to treatment failure was low in the injection group (17%) and high in the manipulation group (59%) and physiotherapy group (51%).
Conclusions: For treating shoulder girdle disorders, manipulation seems to be the preferred treatment. For the synovial disorders, corticosteroid injection seems the best treatment.

	Introduction

Top Abstract Introduction Patients and methods Results Discussion References

In the Netherlands most patients with shoulder complaints are diagnosed and treated by their general practitioner.¹ However, little research has been done to evaluate the effect of the different treatments given. The only trials of the efficacy of manipulation that we could find concerned treating the cervical spine. Among systematic reviews on the effect of non-steroidal anti-inflammatory drugs, corticosteroid injection, and physiotherapy on shoulder complaints, we found only three studies of treatment in general practice.^{2 3 4} Vecchio et al compared a injection of corticosteroid into the subacromial bursa with injection of local anaesthetic and found no difference in efficacy.⁵ Similarly, Jonquière found no difference in the effect of "classic" physiotherapy and that of "Cyriax" physio- therapy.⁶ However, Lacey et al did find that treatment with a non-steroidal anti-inflammatory drug was significantly better than placebo.⁷

In view of the large number of patients with shoulder complaints who are treated in general practice and the lack of studies evaluating different treatments, we set up a trial to find the most effective treatment of shoulder complaints in general practice. Thus, we compared the effects of physiotherapy, corticosteroid injection, and manipulation.

	Patients and methods

Top Abstract Introduction Patients and methods Results Discussion References

Patients
Between September 1994 and September 1995, all patients who consulted seven general practices in the Netherlands with shoulder complaints were included in our study unless one of the exclusion criteria applied. The study was approved by the ethics committee of the Medical Department of the University of Groningen. Before participating in the study, the patients had to give written informed consent.

Shoulder complaints were defined as pain localised in the region of the deltoid muscle, acromioclavicular joint, superior part of the trapezoid muscle, and scapula. Radiation of the pain in the arm could be present, and, besides the pain, the range of movement of the upper arm or shoulder girdle could be limited.

Exclusion criteria were treatment for shoulder complaints in the six months before consultation; bilateral shoulder complaints; presence of specific rheumatic disorders (polymyalgia rheumatica, rheumatoid arthritis, systemic lupus erythematosus, and fibromyalgia); shoulder complaints because of acute severe trauma such as fracture, dislocation, and cuff rupture (patients with a history of minor trauma were not excluded); presence of herniated cervical disc; presence of dementia or other psychiatric disorders; and refusal.

Allocation to treatment
On entry to the study, the patients' level of pain was established and they underwent a physical examination. On the basis of these the patients were allocated to three diagnostic groups: a synovial group, a shoulder girdle group, and a group with combinations of synovial and shoulder girdle disorders. For the first week, all patients were prescribed diclofenac sodium 50 mg thrice daily. At the end of the week, the patients' level of pain was measured again and the physical examinations were repeated by the general practitioners.

On the basis of this second diagnosis, patients were divided into two diagnostic groups: a shoulder girdle group and a synovial group (which also included the combination group because a previous study had shown that the course of complaints of the combination group and the synovial group was the same⁸). Randomisation to treatment took place separately in these two groups: patients in the synovial group were randomised to corticosteroid injection, manipulation, or physiotherapy, while those in the shoulder girdle group were randomised to manipulation or physiotherapy (injections could not be given in this group).

Assessment
Pain measurement—The severity of the shoulder complaints was assessed with the shoulder pain score, which is a six item questionnaire together with a 101 point numerical pain scale (for the total experienced pain).⁹ The six questions—pain at rest, pain during motion, pain during the night, sleeping problems because of pain, inability to lie on the affected side, and presence of radiated pain—were scored on a four point scale of severity. The score on the 101 point numerical pain scale was also converted to a four point scale in order to calculate the sum score of the shoulder pain score. The range was from 7 points (no pain) to 28 points (severe pain).⁹

Physical examination consisted of measuring the active and passive range of movement of the glenohumeral joint, cervical spine, and upper thoracic spine and palpating the muscle tendons on the head of the humerus, the acromioclavicular joint, and the upper ribs. The examinations on inclusion into the study and before randomisation to treatment were performed by the seven participating general practitioners. Follow up examinations were done by a physiotherapist. In order to limit variation between doctors, the researchers had several sessions practising the physical examination and diagnostic interpretation.

Diagnostic groups
The three diagnostic groups have been described in detail elsewhere.⁸

The synovial group consisted of patients with pain or limited movement in one or several directions of the glenohumeral joint. These complaints originated from disorders of the subacromial structures, the acromioclavicular joint, the glenohumeral joint, or combinations of these (the synovial structures).

The shoulder girdle group consisted of patients with pain and sometimes slightly limited range of active movement of the glenohumeral joint. These problems were not related to the synovial structures but, instead, probably originated from functional disorders of the cervical spine, upper thoracic spine, or the upper ribs (the shoulder girdle).

The combination group consisted of patients with pain and sometimes slightly limited range of active or passive movement of the glenohumeral joint together with pain or limited range of movement of the cervical spine, upper thoracic spine, or upper ribs. Both the synovial structures and the structures of the cervical spine, upper thoracic spine, or upper ribs could have caused these complaints.⁸

Treatment
Corticosteroid injection consisted of an injection of 1 ml of 40 mg/ml triamcinolone acetonide in combination with 9 ml of 10 mg/ml lignocaine. One to three injections were given by the participating doctors immediately after randomisation, one week later, and, if needed, after a further two weeks. In each treatment session two out of the three synovial structures (glenohumeral joint capsule, subacromial space, and acromioclavicular joint) were injected. We chose this multiple injection scheme because most of the patients in the synovial group had combinations of disorders of the synovial structures.⁸ Using a multiple injection scheme modified from that of Steinbroker et al,¹⁰ Roy et al had successfully treated frozen shoulder.¹¹ Our injection techniques were standardised: the intra-articular injection was given from the posterior side, the subacromial injection from the lateral side, and the acromioclavicular injection perpendicularly from the upper side of the joint.

Physiotherapy was given twice a week by local physiotherapists. They were instructed to use "classic" physiotherapy—such as exercise therapy, massage, and physical applications. No mobilisation techniques or manipulative techniques were allowed. This definition of physiotherapy was satisfactorily used by Koes et al in their study of treating low back pain.¹²

Manipulation consisted of mobilisation and manipulation of the cervical spine, upper thoracic spine, upper ribs (on the segmental level), acromioclavicular joint, and the glenohumeral joint once a week with a maximum of six treatment sessions. The manipulation was done by either the participating general practitioners or physiotherapists (graduates from the Eindhoven course for manipulative therapy). They were instructed in which techniques to use.

Follow up
After treatment had started, the patients weekly filled in the pain questionnaire. They were also asked to indicate if they felt "cured" or if the treatment failed. Feeling cured was defined as disappearance of shoulder complaints or a decrease of shoulder complaints to such an extent that they were no longer inconvenient, did not need treatment, or no longer interfered with normal working. In our previous study of the pain questionnaire we found that patients did not need to be totally free of pain to feel "cured."⁹ Treatment failed when a patient experienced no improvement or the condition deteriorated.

Follow up examinations were done by a physiotherapist at two, six, and 11 weeks after randomisation. If a patient felt cured or the treatment had failed a final examination was done as soon as possible. At the end of the study the physiotherapist contacted these patients to inquire about present complaints. If patients did have complaints, their level of pain was established and they could indicate whether they felt cured.

Outcome parameters and statistical analysis
Before the study began, power calculation showed that, with = 0.05 and a power of 80%, a difference of 0.8 standard deviation could be detected in treatment groups of 25 patients. Our aim was to achieve treatment groups of at least 30 patients.

To evaluate the effect of treatment, we analysed the duration of patients' complaints, treatment failures, and any complaints at the end of the study on an intention to treat basis. We analysed the duration of the complaints with a survival analysis (log rank test), also known as event history analysis. In this study the event we studied was patients' feeling "cured," and we corrected the calculations for patients who dropped out because of treatment failure. We evaluated the differences between group averages with analysis of variance or Student's t test and analysed the difference between group numbers with the ² test.

Assignment
The university's Department of Family Practice was in charge of the randomisation to treatment. For each diagnostic category, we had made a series of closed unnumbered envelopes which contained instructions of the treatment to be given. The participating general practitioners had to call a secretary and state the diagnostic category of each patient. The secretary in turn would draw an envelope to assign treatment.

Masking (blinding)
The follow up examinations after randomisation were done by a physiotherapist who was not informed about the patients' diagnosis and treatment.

	Results

Top Abstract Introduction Patients and methods Results Discussion References

A total of 198 patients enrolled in the study, and table 1 summarises their characteristics. A substantial proportion of the patients had previously experienced shoulder complaints, and almost 20% had had shoulder complaints for six months or more before consultation. About half of the patients had a synovial syndrome, a quarter had a shoulder girdle syndrome, and a quarter had a combination syndrome.

Twenty six patients dropped out of the study before randomisation. One patient dropped out because of family circumstances. The other 25 considered themselves to be cured after the week's treatment with non-steroidal anti-inflammatory drug: their mean pain scores had dropped from 17.8 to 8.9. These patients were generally similar to the total cohort enrolled, though they had a shorter history of shoulder complaints (table 1).

Figure 1 shows how the remaining 172 patients were randomised to treatment. The increased number of patients in the shoulder girdle group (from 46 patients to 58) was because of a diagnostic shift from the synovial or combination group towards the shoulder girdle group as a result of the treatment with non-steroidal anti-inflammatory drug.

View larger version (19K): [in this window] [in a new window]   Fig 1 Flow chart describing progress of patients through trial

Table 2 shows the characteristics of these groups. The shoulder girdle group was younger than the synovial group.

Shoulder girdle group
Figure 2 shows the survival analysis of the shoulder girdle group. Manipulation was superior to physiotherapy (P < 0.001): at five weeks after randomisation almost 70% of the patients in the manipulation group considered themselves to be cured compared with 10% of the physiotherapy group. Drop out because of treatment failure was significantly higher in the physiotherapy group (45% (13/29) of patients) than in the manipulation group (20% (6/29) of patients).

View larger version (15K): [in this window] [in a new window]   Fig 2 Duration of shoulder complaints in shoulder girdle group after randomisation to treatment

Table 3 shows the two treatment groups' pain scores at randomisation and the final pain scores (on being "cured" or at 11 weeks after randomisation). Both treatments significantly reduced the patients' pain scores. When we differentiated between patients who were "cured" and those who were not, we found that the reductions in the pain scores in both treatment groups were due to the "cured" patients. Of the patients who were "cured" before week 11 after randomisation, 15% (2/13) of patients in the physiotherapy group and 9% (2/22) of patients in the manipulation reported a recurrence of complaints by week 11 after randomisation.

Synovial group
Figure 3 shows the survival analysis of the three treatment groups in the synovial group. The corticosteroid injection group (average number of injections was 1.8) improved rapidly, while the physiotherapy group improved slowly and the manipulation group did only slightly better (P < 0.001): at five weeks after randomisation, 75% of patients in the injection group were "cured" compared with 20% in the physiotherapy group and 40% in the manipulation group. Drop out because of treatment failure was much lower in the injection group (17% (7/47)) than in the physiotherapy group (51% (18/35)) and manipulation group (59% (19/32)).

View larger version (18K): [in this window] [in a new window]   Fig 3 Duration of shoulder complaints in synovial group after randomisation to treatment

Table 3 shows that all three treatments significantly reduced the patients' pain scores. Again, the patients who were "cured" accounted for this reduction. Of the patients who were "cured" before 11 weeks after randomisation, a recurrence of complaints by week 11 was reported by 18% (7/39) of patients in the injection group 13% (2/15) in the physiotherapy group, and 8% (1/13) in the manipulation group.

	Discussion

Top Abstract Introduction Patients and methods Results Discussion References

Design of study
Our study design was based on the results of our earlier descriptive study.⁸ During that study it became evident that other diagnostic classifications, such as those by Cyriax¹³ and the National Guidelines for Shoulder Complaints of the Dutch College of General Practitioners,¹⁴ were not suitable for diagnosing shoulder complaints in general practice. Shoulder complaints seem to be often due to problems in various structures in and around the glenohumeral joint or the structures of the shoulder girdle.

Patients were prescribed a non-steroidal anti-inflammatory drug in the first week after enrollment in order to reduce moderate to severe pain to light to moderate pain. This allowed us to treat patients with physiotherapy and manipulation without having to give additional treatment for the pain. In our study 13% of the patients were "cured" after the non-steroidal anti-inflammatory drug treatment.

Despite the randomisation procedure, in the synovial group the patients allocated to manipulation were significantly younger than the patients allocated to the two other treatments, and the percentage of men in the physiotherapy group was significantly higher than in the other groups. In a separate regression analysis we concluded that sex did not have a significant influence on the duration of complaints but the age of patients did. Thus, the lower age of the patients given manipulation group could have influenced the better results that they showed in the first 6 weeks after randomisation compared with the physiotherapy group. However, the results of manipulation in the group were modest, especially when compared with the results of manipulation in the shoulder girdle group, which had the same average age.

Implications of results
To our knowledge, no other published study has described the positive effects of manipulation in treating shoulder complaints. The results of our study suggest that manipulation is to be preferred to physiotherapy for treating shoulder complaints originating from the shoulder girdle in general practice.

Of 22 comparative studies that investigated corticosteroid injection for treating shoulder complaints, only five describe success with injection.³ We consider our positive results with corticosteroid injection were helped by our setting in general practice (no patient selection by referral) and adequate selection of patients by diagnostic groups. We found injection to be the most effective treatment for shoulder complaints originating from the synovial structures in general practice (after a week's treatment with a non-steroidal anti-inflammatory drug). A slightly higher percentage of the "cured" patients in the injection group reported recurrence of complaints at the end of the study. However, in this group 80% of the patients were "cured" by the fifth week after randomisation so these patients had the longest period for symptoms to recur.

	Acknowledgements

We thank our colleague general practitioners Luit-Jan Lukkes, Mello Maaskant, Henk Spelde, Jan de Weerd, and Jan Woudhuizen for their willingness to participate in this study.

Funding: This study was conducted with a grant of the Ministry of Welfare, Health and Culture.

Conflict of interest: None.

	References

Top Abstract Introduction Patients and methods Results Discussion References

1. Lamberts H, Wood M, Hofmans-Okkes JM. The international classification of primary care in the European community. Oxford: Oxford University Press, 1993.
2. Winters JC, Sobel JS, Arendzen JH, Meyboom-de Jong B. NSAID's bij schouderaandoeningen [NSAID's and shoulder complaints]. Huisarts en Wetenschap 1995;38:339-41, 347.
3. Winters JC, Sobel JS, Arendzen JH, Meyboom-de Jong B. Het effect van corticosteroïdinjecties bij schouder-aandoeningen [The effect of corticosteroid injections on shoulder complaints]. Huisarts en Wetenschap 1995;38:164-8.
4. Van der Heijden GJMG, Bouter LM, Beckerman H, de Bie RA, Oostendorp RAB. De effectiviteit van fysiotherapy bij schouderklachten, een geblindeerd literatuuronderzoek [The effect of physiotherapy on shoulder complaints, a blinded literature research]. Nederlands Tijdschrift voor Fysiotherapie 1992;102:38-46.
5. Vecchio PC, Hazleman BL, King RH. A double-blind trial comparing subacromial methylprednisolone and lignocaine in acute rotator cuff tendinitis. Br J Rheumatol 1993;32:743-5. [Abstract/Free Full Text]
6. Jonquière M. De behandeling van schouder-en elleboogaandoeningen volgens de richtlijnen van Cyriax [The treatment of shoulder and elbow complaints according to the guidelines of Cyriax] [dissertation]. Rotterdam: Erasmus University, 1986: 121-35.
7. Lacey PH, Dodd GD, Shannon DJ. A double blind, placebo controlled study of piroxicam in the management of acute musculoskeletal disorders. Eur J Rheumatology Inflammation 1984;7:95-104.
8. Sobel JS, Winters JC, Arendzen JH, Groenier KH, Meyboom-de Jong B. Schouderklachten in de huisartspraktijk [Shoulder complaints in general practice]. Huisarts en Wetenschap 1995;38:342-7.
9. Winters JC, Sobel JS, Groenier KH, Arendzen JH, Meyboom-de Jong B. A shoulder pain score; a comprehensive questionnaire to assess pain in patients with shoulder complaints. Scan J Rehabil Med 1996;28:163-7.
10. Steinbroker O, Thomas G, Argyros MD. Frozen shoulder: treatment by local injections of depot corticosteroids. Arch Phys Med Rehabil 1974;55:209-12.
11. Roy S, Oldham R. Management of the frozen shoulder. Lancet 1976;i:1322-4.
12. Koes BW, Bouter LM, van Mameren H, Essers AHM, Verstegen GMR, Hofhuizen DM, et al. The effectiveness of manual therapy, physiotherapy and treatment by the general practitioner for non-specific back and neck complaints: a randomised clinical trial. Spine 1992;17:28-35. [Medline]
13. Cyriax J. Textbook of orthopaedic medicine. 11th ed, vol 1. Diagnosis of soft tissue lesions. London: Bailliere Tindall, 1984: 127-42, 143-58.
14. Bakker JF, de Jongh AC, Jonquière M, Mens J, Oosterhuis WW, Poppelaars A, et al. NHG Standaard Schouderklachten [National Guidelines for Shoulder Complaints of the Dutch College of General Practitioners]. Huisarts en Wetenschap 1990;33:196-202.

CiteULike    Complore    Connotea    Del.icio.us    Digg    Reddit    StumbleUpon    Technorati    What's this?

This article has been cited by other articles:

• Jewell, D. V, Riddle, D. L, Thacker, L. R (2009). Interventions Associated With an Increased or Decreased Likelihood of Pain Reduction and Improved Function in Patients With Adhesive Capsulitis: A Retrospective Cohort Study. ptjournal 89: 419-429 [Abstract] [Full text]  
• McKenna, C., Bojke, L., Manca, A., Adebajo, A., Dickson, J., Helliwell, P., Morton, V., Russell, I., Torgerson, D., Watson, J. (2009). Shoulder acute pain in primary health care: is retraining GPs effective? The SAPPHIRE randomized trial: a cost-effectiveness analysis. Rheumatology (Oxford) 48: 558-563 [Abstract] [Full text]  
• Watson, J., Helliwell, P., Morton, V., Adebajo, A., Dickson, J., Russell, I., Torgerson, D. (2008). Shoulder acute pain in primary healthcare: is retraining effective for GP principals? SAPPHIRE--a randomized controlled trial. Rheumatology (Oxford) 47: 1795-1802 [Abstract] [Full text]  
• Hambly, N., Fitzpatrick, P., MacMahon, P., Eustace, S. (2007). Rotator Cuff Impingement: Correlation Between Findings on MRI and Outcome After Fluoroscopically Guided Subacromial Bursography and Steroid Injection. Am. J. Roentgenol. 189: 1179-1184 [Abstract] [Full text]  
• Shrier, I, Macdonald, D, Uchacz, G (2006). A pilot study on the effects of pre-event manipulation on jump height and running velocity. Br. J. Sports. Med. 40: 947-949 [Abstract] [Full text]  
• Alvarez, C. M., Litchfield, R., Jackowski, D., Griffin, S., Kirkley, A. (2005). A Prospective, Double-Blind, Randomized Clinical Trial Comparing Subacromial Injection of Betamethasone and Xylocaine to Xylocaine Alone in Chronic Rotator Cuff Tendinosis. Am J Sports Med 33: 255-262 [Abstract] [Full text]  
• Bergman, G. J.D., Winters, J. C., Groenier, K. H., Pool, J. J.M., Jong, B. M.-d., Postema, K., van der Heijden, G. J.M.G. (2004). Manipulative Therapy in Addition to Usual Medical Care for Patients with Shoulder Dysfunction and Pain: A Randomized, Controlled Trial. ANN INTERN MED 141: 432-439 [Abstract] [Full text]  
• Kassimos, D G, Panayi, G, van der Windt, D A W M (2004). Differences in the management of shoulder pain between primary and secondary care in Europe: time for a consensus * Author's reply. Ann Rheum Dis 63: 111-112 [Full text]  
• McInerney, J J, Dias, J, Durham, S, Evans, A (2003). Randomised controlled trial of single, subacromial injection of methylprednisolone in patients with persistent, post-traumatic impingment of the shoulder. Emerg. Med. J. 20: 218-221 [Abstract] [Full text]  
• van der Windt, D A W M, Bouter, L M (2003). Physiotherapy or corticosteroid injection for shoulder pain?. Ann Rheum Dis 62: 385-387 [Full text]  
• Hay, E M, Thomas, E, Paterson, S M, Dziedzic, K, Croft, P R (2003). A pragmatic randomised controlled trial of local corticosteroid injection and physiotherapy for the treatment of new episodes of unilateral shoulder pain in primary care. Ann Rheum Dis 62: 394-399 [Abstract] [Full text]  
• van der Heijden, G. J M G, Leffers, P., Wolters, P. J M C, Verheijden, J. J D, van Mameren, H., Houben, J. P, Bouter, L. M, Knipschild, P. G (1999). No effect of bipolar interferential electrotherapy and pulsed ultrasound for soft tissue shoulder disorders: a randomised controlled trial. Ann Rheum Dis 58: 530-540 [Abstract] [Full text]  
• Winters, J. C, Jorritsma, W., Groenier, K. H, Sobel, J. S, Jong, B. M.-d., Arendzen, H. J (1999). Treatment of shoulder complaints in general practice: long term results of a randomised, single blind study comparing physiotherapy, manipulation, and corticosteroid injection. BMJ 318: 1395-1396 [Full text]  
• van der Windt, D A W M, Koes, B W, Devillé, W, Boeke, A J P, de Jong, B A, Bouter, L M (1998). Effectiveness of corticosteroid injections versus physiotherapy for treatment of painful stiff shoulder in primary care: randomised trial. BMJ 317: 1292-1296 [Abstract] [Full text]  
• Clarke, M., Chalmers, I. (1998). Discussion Sections in Reports of Controlled Trials Published in General Medical Journals: Islands in Search of Continents?. JAMA 280: 280-282 [Abstract] [Full text]  
• Symonds, G., Speed, C A, Orr, B., Winters, J. C (1997). Treating shoulder complaints in general practice. BMJ 315: 680-681 [Full text]  
• (1997). WHICH TREATMENT FOR SHOULDER COMPLAINTS?. JWatch General 1997: 3-3 [Full text]