Annals of the Rheumatic Diseases 1995; 54: 959-964

EXTENDED REPORTS

Shoulder disorders in general practice: incidence,patient characteristics, and management

Danielle AWM van der Windt, BartW Koes, Bareld A de Jong, Lex M Bouter

AbstractObjectives-To study the incidence andmanagement of intrinsic shoulder dis-orders in Dutch general practice, andto evaluate which patient characteristicsare associated with specific diagnosticcategories.Methods-In 11 general practices (35 150registered patients) all consultationsconcerning shoulder complaints wereregistered during a period of one year.Patients with an intrinsic shoulder dis-order who had not consulted their generalpractitioner for the complaint duringthe preceding year (incident cases) wereasked to participate in an observationalstudy. Participants completed a question-naire regarding the nature and severity oftheir complaints. The general prac-titioners recorded data on diagnosis andtherapy.Results-The cumulative incidence ofshoulder complaints in general practicewas estimated to be ll2/1000 patients/year(95% confidence limits 10-1 to 12.3).Rotator cuff tendinitis was the most fre-quently recorded disorder (29%). Therewere 349 incident cases enrolled in theobservational study. Patient character-istics showed small variations betweendifferent diagnostic categories. Age,duration ofsymptoms, precipitating causeand restriction ofmovement seemed to bediscriminating factors. Twenty two per-cent of all participants received injectionsduring the first consultation; most (85%)were diagnosed as having bursitis. Themajority of patients with tendinitis (53%)were referred for physiotherapy.Conclusion-With respect to diagnosisand treatment, the practitioners generallyappeared to follow the guidelines issuedby the Dutch College of General Prac-titioners. Although the patient character-istics of specific disorders showed somesimilarities with the clinical picturesdescribed in the literature, furtherresearch is required to demonstratewhether the proposed syndromes indeedconstitute separate disorders with adifferent underlying pathology, requiringdifferent treatment strategies.

(Ann Rheum Dis 1995; 54: 959-964)

Shoulder complaints are encountered fre-quently in primary health care. The incidence

in Dutch general practice has been estimatedat 12 to 25/1000/year. 1- The NationalMorbidity Surveys in England and Wales havereported a somewhat lower annual incidence of6-6/1000.5 A considerable number of episodesmay not be presented to the health careprofessionals; reports of point prevalencesin general populations range from 70 to260/1000.6-9A painful or stiff shoulder may be caused by

various diseases and conditions, includingneurological or vascular disorders, neoplasms,referred pain from internal organs, and dis-orders of the cervical spine (extrinsic causes).In most cases the complaints are of intrinsicorigin, caused usually by articular or peri-articular rheumatic conditions of the shoulderjoint. The majority of patients are treated inprimary health care, as intrinsic complaintsoften constitute a self limiting condition ofrelatively short duration (less than threemonths). The long term outcome is not alwaysfavourable: persisting pain or a limited range ofmotion in chronic conditions may last forseveral years.'S'3 In a community survey ofshoulder disorders in the elderly, 108 patientswere examined three years after the initialdiagnosis: 74% showed persisting signs of theircondition. 14The aetiology and pathogenesis of shoulder

disorders tend to remain enigmatic. Thecomplex anatomical and functional structureof the shoulder joint complicates identificationof the source of the lesion. This has resultedin much confusion and lack of consensusregarding the classification of shoulder dis-orders. Terms such as periarthritis humero-scapularis, or a painful stiff shoulder, do notrepresent clearly defined clinical syndromes.Diagnostic criteria may even vary for disordersmore straightforwardly labelled as rotator cufftendinitis or subacromial bursitis. As a result,information on the incidence and diseasecharacteristics of the various conditions of theshoulder joint is scarce, particularly regardingthose patients encountered in primary healthcare; the bulk of the medical literature onshoulder disorders has been written byorthopaedists and rheumatologists and refersto hospital based populations.

In 1990, the Dutch College of GeneralPractitioners issued clinical guidelines for thediagnosis and treatment of shoulder com-plaints.'5 In these guidelines, a classification ofshoulder complaints was introduced, basedlargely on the concepts of Cyriax,16 thatdescribes four intrinsic shoulder syndromes:

Institute for Researchin ExtramuralMedicine,Faculty ofMedicine,Vrije Universiteit,Amsterdam,The NetherlandsD AWM van der WindtB W KoesDepartment ofEpidemiology andBiostatisticsLM BouterDepartment ofRehabilitationMedicine,Academic MedicalCentre,University ofAmsterdam,The NetherlandsB A de JongCorrespondence to:Ms D AW M van der Windt,Institute for Research inExtramural Medicine,Faculty of Medicine,Vrije Universiteit,Van der Boechorststraat 7,1081 BT Amsterdam,The Netherlands.

Accepted for publication18 August 1995

959 on M

arch 28, 2020 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.54.12.959 on 1 D

ecember 1995. D

ownloaded from

van der Windt, Koes, de _tong, Bouter

capsular syndrome, acute bursitis, acro-mioclavicular syndrome, and subacromialsyndrome. We have used this system ofclassification in an observational study ofthe incidence and management of intrinsicshoulder disorders in primary health care in theNetherlands. An additional objective was toevaluate which patient characteristics wereassociated with each specific diagnosticcategory.

Patients and methodsEighteen general practitioners (11 practices),representing a population of 35 150 patientsparticipated in this observational study. Beforethe study, the general practitioners receivedextra training on the systematic examination ofthe cervical spine and shoulder joint, accordingto the concepts of Cyriax. 6 The clinicalguidelines issued by the Dutch College ofGeneral Practitioners15 were used to classifyshoulder complaints. Table 1 summarises theclinical guidelines for diagnosis and treatment.

INCIDENCE OF SHOULDER COMPLAINTS

During a period of 12 months, from April1993 to April 1994, the general practitionersregistered details of all consultations regardingshoulder complaints, including the age andgender of the patients, and the diagnosis.During the registration period the practitionerswere asked twice to estimate the percentage ofconsultations actually registered, in order toassess the number of missed records. Incidentcases were defined as those presented bypatients who had not consulted their generalpractitioner for shoulder complaints in thepreceding year.The cumulative incidence of shoulder com-

plaints (n/1000/year) was calculated, includingthe 95% confidence limits. Age and genderspecific incidences were calculated for a limitednumber of practices, as not every practicecould provide sufficient data on the age andgender of their practice population. Thecumulative incidence in each general practicewas subsequently adjusted for one month ofabsence (holidays, illness, etc) and for a

percentage of missed records, as estimated bythe practitioners themselves (mean proportionof missed records: 20%).

OBSERVATIONAL STUDY OF INTRINSIC

SHOULDER DISORDERSPatients with an incident complaint were

eligible for participation in the observationalstudy. Inclusion criteria for that study were:informed consent; age 18 years or older; abilityto complete a written questionnaire; intrinsicshoulder complaints, originating from withinthe shoulder joint, but no neurological orvascular disorders, neoplasms, referred painfrom internal organs, or systemic rheumaticconditions; and no fractures or luxations.For all participants, their general prac-

titioners recorded data on diagnostic tests andprocedures, medication prescribed, injections,referrals, and relevant comorbidity (systemicrheumatic conditions, diabetes mellitus, gastriccomplaints, ischaemic heart disease, disordersofthe gall bladder or thyroid gland, hemiplegia).The participants completed a written question-naire containing questions on demographicvariables, previous complaints, the nature ofcurrent complaints, the severity of complaintsat night and during the day (11 point ordinalscale: 0 = no complaints... 10 = very severe

complaints), and functional disability.Descriptive statistics were used to present

the frequencies of diagnoses, interventions andpatient characteristics.

ResultsINCIDENCE OF SHOULDER COMPLAINTS(TABLE 2)During a period of one year the 18 generalpractitioners recorded 754 consultations con-

cerning shoulder complaints in 472 patients;392 of the patients presented with an incidentcomplaint. The cumulative incidence of shouldercomplaints was calculated as 11 2/1000/year.The incidence varied considerably betweenpractices: from 6-5 to 20 3/1000/year. Afteradjustment for one month of absence andincomplete registration, the incidence was

estimated at 14 7/1000/year (table 2).

Table 1 Summary of clinical guidelines for diagnosis and treatment ofshoulder complaints (Dutch CoHlege of General Practitioners'5)

Syndrome Diagnostic criteria Guidelines for treatment

Capsular syndrome Restriction of lateral rotation, abduction, and medial rotation. 1. NSAIDs or local infiltration of a steroid or anaesthetic(capsulitis, arthrosis, etc.) Pain in C5 dermatome. 2. Passive mobilisation and exercise therapy

Acute bursitis Restriction of abduction. Severe pain in C5 dermatome. 1. Local injection of anaesthetic, steroid, or bothAcute onset, no evident preceding trauma. 2. Rest and NSAIDs or analgesics in less severe cases

Acromioclavicular syndrome Restriction of horizontal adduction. Pain in the area of the 1. NSAIDsacromioclavicular joint and/or C4 dermatome. 2. Local injection of an anaesthetic

3. Local injection of a steroid (and anaesthetic)4. Physiotherapy

Subacromial syndrome No restriction of passive movement. Pain in the C5 dermatome. BursitisRotator cuff tendinitis Painful arc during elevation. At least one positive resistance test. 1. Rest/NSAIDsChronic bursitis Bursitis: variable/little pain, normal power 2. Local injection of a steroid and/or anaestheticRotator cuff tears Tendinitis: pain, normal power Tendinitis

Cuff tears: little pain, loss of power 1. Rest/NSAIDs2. Physiotherapy3. Local injection of an anaesthetic

Cuff tears1. Rest/NSAIDs2. Local injection of an anaesthetic

Remainder (unclear clinical picture,fractures, luxations, myalgia, etc)

960

on March 28, 2020 by guest. P

rotected by copyright.http://ard.bm

j.com/

Ann R

heum D

is: first published as 10.1136/ard.54.12.959 on 1 Decem

ber 1995. Dow

nloaded from

Shoulder disorders in general practice

Table 2 Cumulative incidence ofshoulder complaints in 11 general practices(comprising 35 150 registered patients)

Incident Cumulative 95% confidencecases incidence limits

(n/l 000/year)

Total unadjusted incidence 392 11-2 10 1 to 12-3Total incidence adjusted for

incomplete reporting* 392 14-7 13-4 to 16-0Gender (unadjusted)tMale 90 8-4 6-7 to 10-2Female 117 11 1 91 to 13-1

Age categories (years) (unadjusted)t<25 7 1.0 0-3to 1-725-44 77 10-2 7-9 to 12-545-64 105 17-3 14-0 to 20-665-74 26 12-8 7 9 to 17-7>75 9 6-7 2-3to 11 1

Diagnosis (unadjusted)Capsular syndrome 84 2-4 1 9 to 2-9Acute bursitis 64 1-8 1-4 to 2-2Acromioclavicular syndrome 18 0 5 0 3 to 0 7Subacromial syndromef 174 5 0 4-3 to 5-7Unclear 39 1.1 0-8to 1-4Extrinsic causes 13 0 4 0-2 to 0-6

*The incidence in each practice was adjusted for one month of absence (holidays, illness, etc)and a percentage of missed records, as estimated by the general practitioners themselves.tIncidences based on a limited number of practices (n = 7) able to provide age and gender specificdata on their practice population.tSubacromial syndrome: 114 cases of rotator cuff tendinitis (29% of all patients), 45 cases ofchronic bursitis (12% of all patients).

The incidence of shoulder complaints wasgreater for women (11 1/1000/year) than formen (8 4/1000/year) and peaked in the agecategory 45-64 years (17 3/1000/year). Thesubacromial syndrome was the disorder diag-nosed most frequently, in particular rotatorcuff tendinitis (29%).

OBSERVATIONAL STUDY OF INTRINSIC

SHOULDER DISORDERSA total of 349 patients were enrolled in thefollow up study. There were 43 incident casesthat were not included for the followingreasons: nine refused to participate, 22 wereunable to complete written questionnaires forvarious reasons, and 12 were suspected ofhaving complaints that were extrinsic in origin.The response rate to the baseline questionnairewas 96% (335 patients). The population com-prised 28% housewives, 31% unskilled andlesser skilled labourers, 26% middle and highereducated personnel, and 15% who werestudents, unemployed, or retired. Comorbiditywas recorded for 42 patients (1 2%). Theconcomitant diseases most frequently recordedwere diabetes mellitus (15), ischaemic heartdisease (10), and systemic disorders of themusculoskeletal system (eight; mainly osteo-porosis). In addition to their shoulder com-plaints, 43% of patients reported a painful orstiff neck.

Table 3 presents the age and gender ofthe patients, with their history of shouldercomplaints, and diagnoses. Nearly 50% hadexperienced shoulder complaints previously(mean number of such episodes: 5-8 (SD 8 8)).Table 4 presents the information from the

patient questionnaires concerning the natureand severity of shoulder complaints. Preva-lence of patient characteristics is presented forall patients, and for the most frequentlyrecorded syndromes: capsular syndrome, acutebursitis, tendinitis, and chronic bursitis. There

was little difference between syndromes withrespect to gender, or to severity of complaints.The mean severity of complaints was ratedsomewhat higher during the day than at night(7-2 and 6&3, respectively, on the 11 pointordinal scale).Other patient characteristics showed some

variation between syndromes. A greater pro-portion of patients with capsular syndromewere aged 45 years or older. The longstandingnature of their complaint (more than onemonth at presentation) was reported morefrequently by patients with capsular syndrome(64%) or chronic bursitis (62%), comparedwith only 20% of those with acute bursitis.Overuse or strain was a precipitating causeaccording to a relatively large proportion ofpatients with subacromial syndrome, particu-larly rotator cuff tendinitis. Patients withcapsular syndrome relatively often reportedboth pain and stiffness as the predominantcomplaint (47%). Sleep disturbances wereprevalent, particularly in patients with capsularsyndrome or acute bursitis. Restriction ofmovement was a problem in all subgroups, butinternal rotation (difficulties reaching to thelower back) was relatively often reported bypatients with capsular syndrome (87%).

MANAGEMENT BY THE GENERAL PRACTITIONER(FIRST CONSULTATION)The general practitioners requested fewadditional diagnostic procedures: seven radio-graphs (2%) and one laboratory measurement(0-3%). Table 5 presents an analysis of thetreatments offered, both to all participants,and separately for the four most frequentlydiagnosed disorders: capsular syndrome, acutebursitis, tendinitis, and chronic bursitis.Treatment of shoulder complaints was often

initiated with the prescription of medication,mainly non-steroidal anti-inflammatory drugs.Local injection of a steroid, an anaesthetic, orboth, was the treatment received by 22% of allparticipants, most of whom were diagnosed as

Table 3 Gender, history ofshoulder complaints, anddiagnoses ofparticipants in the observational study(n = 349; mean (SD) age 49-6 (14-4) years). Responserate to the baseline questionnaire: 96% (335 patients)

Number %

GenderMale 146 44Female 189 56

History of shoulder complaintsNo 179 54Yes 154 46

Treatnent of previous complaints (n = 154)*No treatment 51 33Medication 27 18Physiotherapy 71 46Injections 52 34Surgery 4 3

Diagnosis (incident episode)Capsular syndrome 75 22Acute bursitis 56 17Acromioclavicular syndrome 14 4Subacromial syndr6met 159 -48Unclear 31 9

*Total frequency may exceed 100%, as more than onetreatment modality may have been offered to a patient.tSubacromial syndrome: 102 cases of rotator cuff tendinitis(30% of all patients), 42 cases of chronic bursitis (13% of allpatients), 15 cases of rotator cuff tears or mixed clinicalpicture.

961 on M

arch 28, 2020 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.54.12.959 on 1 D

ecember 1995. D

ownloaded from

van der Windt, Koes, de long, Bouter

Table 4 Patient characteristics ofshoulder disorders. Resultsfrom the patient questionnaires from all participants in theobservational study, andpresented separately for capsular syndrome (CS), acute bursitis (AcB), tendinitis (TD), andchronic bursitis (ChrB)

All (n 335) CS (n= 75) AcB (n 56) TD (n 102) ChrB (n= 42)(%/) (%) (%) (v.) (%)

Age a,45 years 62 91 57 48 74Female 56 55 66 57 64Duration of symptoms a 1 month 49 64 20 48 62

Precipitating causeUnknown 49 63 66 39 38Injury 12 9 2 9 17Strain/overuse usual activities 18 13 11 27 21Strain/overuse unusual activities 13 9 13 18 10

Predominant complaintStiffness 11 10 9 9 24Both pain and stiffness 32 47 29 27 37

Sleep disturbancesUnable to lie on afflicted shoulder 60 73 69 52 51Not getting to sleep/waking up 81 92 91 75 78

Severity of complaints*During the day 67 70 74 68 58At night 54 66 60 45 53

Difficulties reaching toBack of the head (elevation/abd.) 77 81 76 73 85Lower back (internal rotation) 73 87 75 67 73

*Scores a 7 on an ordinal 11 point scale, ranging from 0 (no complaints) to 10 (very severe complaints).

having acute or chronic bursitis. In our study,physiotherapy seemed to be the preferentialtreatment for rotator cuff tendinitis.

DiscussionThe cumulative incidence of shouldercomplaints in our study was estimated tobe 11 2/1000/year, which agrees with theproportion of 12-8/1000/year reported byMiedema.' Even after adjustment for in-complete registration (14 7/1000/year), theincidences are lower than those reported bytwo other Dutch morbidity registrations(25/1000/year2 and 18-21/1000/year4), buthigher than estimates from the NationalMorbidity Surveys in England and Wales(6-6/1000/year5). These differences in inci-dence may be explained by incompleteregistration by the general practitioner, differ-ences regarding the definition of an incidentepisode, or a variation in the proportion ofpatients reporting a shoulder complaint to theirphysician. In addition, diagnostic. criteria mayhave varied between studies. The practitionersin our study registered intrinsic disordersonly, whereas in other Dutch surveys everycomplaint concerning the shoulder region wasrecorded.The age and gender specific differences

demonstrated in our study are in accordancewith those revealed by other Dutch andEnglish surveys. The greatest proportions ofshoulder complaints are reported during the

Table S Treatment by the general practitioner*. First consultation for shoulder complaintsby participants in the observational studyDiagnosis Wait and see Medication Physiotherapy Injectiont

(%/) (%) (%/6) (%)Al patients (349) 28 38 30 22Capsular syndrome (76) 20 47 20 30Acute bursitis (60) 27 43 3 47Tendinitis (106) 27 32 53 5Chronic bursitis (42) 38 26 17 38

*Total frequency may exceed 100%, as more than one treatment modality may have been offeredto a patient.tLocal infiltration with steroid, anaesthetic, or both.

fifth to the seventh decades of life,' 2 5 with aslight majority of female patients (57-59%).' 2The incidence seems to level offor even declinein the older age categories. However, somecommunity surveys have reported a highprevalence of chronic shoulder conditions inthe elderly,8 9 14 suggesting that many elderlydo not seek medical attention for a persistingpainful or stiff shoulder.

Rotator cuff tendinitis (29% of all incidentcases) was the diagnosis most frequentlyrecorded, as was reported also by Chard et al.9Comparison of our data with those in otherreports is made difficult by the scarcity of data,and the fact that those available are basedpredominantly on hospital surveys. The studyof shoulder disorders is particularly com-plicated because of the lack of consensusregarding the diagnostic criteria of specificshoulder disorders. The ideas of Cyriax on thediagnosis and treatment of shoulder disordershave been applauded, but also severelycriticised.'7 1' Several authors have proposedalternative classification systems.'9 23The importance of a case history in the

assessment of shoulder complaints has beendemonstrated by Giirtner et al.24 In a pro-spective study of 65 patients, shouldercomplaints were classified into 24 diagnosticcategories on the basis of a detailed case historyalone. More than 50% of the diagnoses werein agreement with later findings of arthro-graphy, surgery, etc. In our study, patientcharacteristics showed some variation betweenspecific diagnostic categories, in particularwith respect to age, duration of symptoms,precipitating cause, and restriction of move-ment. The variations, though small, seem to becompatible with the clinical features describedby Cyriax'6 and other authors of frequentlycited papers on shoulder disorders. Forexample, our clinical picture of capsularsyndrome agreed with the description ofadhesive capsulitis by Murnaghan:25 patientsare often unable to sleep on the affectedshoulder, there is functional restriction of

962 on M

arch 28, 2020 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.54.12.959 on 1 D

ecember 1995. D

ownloaded from

Shoulder disorders in general practice

motion, often no precipitating event can berecalled, and occurrence is predominantduring the fifth to the seventh decades of life.However, it must be noted that a descriptionof shoulder syndromes was included in thetraining that preceded our study. A case historyof the participants is likely to have been partof the assessment of complaints by the generalpractitioners, and will have influenced theirdiagnoses. Consequently, there was probablysome cross contamination of data; those fromthe patient questionnaires were, to a certainextent, similar to the information available tothe general practitioner. What our results dorepresent is the distribution, in a consecutivegroup of primary care patients, of the Cyriaxshoulder syndromes as diagnosed by a group ofgeneral practitioners trained in this method ofclassification.As yet, there are no diagnostic tests or

procedures which provide decisive evidence asto the pathology of shoulder complaints.Imaging techniques can be very useful fordetecting rotator cuff tears, but are often oflittle help in diagnosing other soft tissue dis-orders.26 27 In our study, laboratory tests orradiographs were requested only infrequently,but this may be distorted by the fact thatpatients with complaints of suspected extrinsicorigin were excluded from the follow up study;additional diagnostic procedures may havebeen used more frequently for those patients.In other Dutch surveys, radiographs wererequested in approximately 8% of all cases, andlaboratory measurements in 1-2%. ' 2 For mostpatients, however, a detailed case history andclinical examination appear to be sufficientto permit a decision on management andprognosis.A classification of shoulder complaints

would be particularly useful if it implied conse-quences for treatment. The clinical guidelinesrecently introduced by the Dutch Collegeof General Practitioners contain tentativedirectives for treatment for each specificdiagnostic category (table 1). Our resultsindicate that, in general, treatment for shouldercomplaints was initiated according to theseguidelines, although considerable variation ofmethods was noted between practitioners.Treatment was, to some extent, similar to thatrecorded in other Dutch surveys, in which20%2 to 33%' of the patients were referred forphysiotherapy, and medication was prescribedin 43%2 to 50%.l However, injections weregiven more frequently in our sample: to 22%of the patients, compared with 1O/o2 and 13%./lThe majority of our participating generalpractitioners had received extra training in theuse of steroid injections. Other classificationsof shoulder disorders may imply otherdirectives for treatment. 19 20 28 29 Recentsystematic reviews have indicated thatconclusive evidence on the efficacy of non-steroidal anti-inflammatory drugs,30 physio-therapy,3' or steroid injections (Van derHeijden et al, in preparation) for shoulderdisorders is still lacking.The concepts of Cyriax are accepted by a

considerable number of physiotherapists and

physicians, but have seldom been subjected tofurther scrutiny. Future research shoulddemonstrate whether the proposed shouldersyndromes really constitute separate disorderswith differing underlying pathology. Theresults of our continuing observational studywill provide information on the prognosis ofthe shoulder syndromes, as all participants willreceive follow up questionnaires at fixedintervals during a period of one year. The nextquestions to be answered relate to whetherthe choice of treatment should depend ondiagnosis: which intervention is most effectivefor capsular syndrome, are injections indeedthe preferential treatment for bursitis, and isphysiotherapy the best option for rotator cufftendinitis?

The authors wish to thank Rob Scholten and Walter Deville forstatistical advice, Joy Bakker for data entry, Dirk Mul for hiscontribution to study design and conduct, and all participatinggeneral practitioners for their efforts during data collection. Thestudy has received a grant from the foundation 'De DrieLichten'.

1 Miedema H S. Reuma-onderzoek meerdere echelons(ROME): basisrapport. Leiden: Nederlands Instituut voorPraeventieve Gezondheidszorg TNO, 1994.

2 Lamberts H, Brouwer H J, Mohrs J. Reason for encounter-,episode- and process-oriented standard output from Transitionproject. Part I. Amsterdam: Department of GeneralPractice/Family medicine, University of Amsterdam,1991.

3 Velden J van den, Bakker D H de, Claessens A A M C,Schellevis F G. Een nationals studie naar ziekten enverrichtingen in de huisartspraktijk. Basisrapport: Morbiditeitin de huisartspraktijk. Utrecht: Netherlands Institute ofPrimary Health Care, 1991.

4 Lisdonk E H van de, Bosch W J H M van den,Huygen F J A, Lagro-Jansen A L M. Ziekten inde huisartspraktijk, ch 13. Utrecht: WetenschappelijkeUitgeverij Bunge, 1990; 289-316.

5 Croft P. Soft tissue rheumatism. In: Silman A J,Hochberg M C, eds. Epidemiology of the rheumatic diseases,ch 15. Oxford: Oxford Medical Publications, 1993;375-421.

6 Bjelle A. Epidemiology of shoulder problems. Bailliere'sClin Rheumatol 1989; 3: 437-51.

7 Bergenudd H, Lindgarde F, Nilsson B, Petersson C J.Shoulder pain in middle age. A study of prevalence andrelation to occupational work load and psychosocialfactors. Clin Orthop 1988; 231: 234-8.

8 Chakravarty K K, Webley M. Disorders of the shoulder: anoften unrecognised cause of disability in elderly people.BMJ 1990; 300: 848-9.

9 Chard M D, Hazleman R, Hazleman B L, King R H,Reiss B B. Shoulder disorders in the elderly (a communitysurvey). Arthritis Rheum 1991; 34: 766-9.

10 Reeves B. The natural history of the frozen shouldersyndrome. ScandJ Rheumatol 1975; 4: 193-6.

11 Binder A I, Bulgen D Y, Hazleman B L, Roberts S. Frozenshoulder: a long-term prospective study. Ann Rheum Dis1984; 43: 361-4.

12 Chard M D, Satelle L M, Hazleman B L. The long-termoutcome of rotator cuff tendinitis-a review study. Br JfRheumatol 1988; 27: 385-9.

13 Shaffer B, Tibone J E, Kerlan R K. Frozen shoulder.A long term follow-up. J Bone Joint Surg Am 1992; 74:738-46.

14 Vecchio P C, Kavanagh R T, Hazleman B L, King R H.Community survey of shoulder disorders in the elderly toassess the natural history and effects of treatment. AnnRheum Dis 1995; 54: 152-4.

15 Bakker J F, Jongh L de, Jonquiere M, et al. StandaardSchouderklachten. Huisarts en Wetenschap 1990; 33:196-202.

16 Cyriax J. Textbook oforthopaedic medicine, chs 9, 10. London:Baillire Tindal, 1981; 190-254.

17 Potter G E. Role of injections [letter]. Physiotherapy 1987;67: 1575.

18 Matthews J H. Bursitis and tendinitis [letter]. Am FamPhysician 1981; 24: 59-61.

19 Neer C S. Impingement lesions. Clin Orthop 1983; 173:70-7.

20 Uhthoff H K, Sarkar K. An algorithm for shoulder paincaused by soft-tissue disorders. Clin Orthop 1990; 254:121-7.

21 Uhthoff H K, Sarkar K. Classification and definition oftendinopathies. Clin Sports Med 1991; 10: 707-20.

22 Hedtmnann A, Fett H. The so-called periarthropathiahumeroscapularis. Classification and analysis of 1266cases. Z Orthop 1989; 127: 643-9.

23 Neviaser R J. Painful conditions affecting the shoulder. ClinOrthop 1983; 173: 63-9.

963 on M

arch 28, 2020 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.54.12.959 on 1 D

ecember 1995. D

ownloaded from

van der Windt, Koes, de j7ong, Bouter

24 Gartner J, Blauth W, Hahne H J. Importance of patienthistory in the tentative diagnosis of shoulder pain.ZOrthop 1991; 129: 322-5.

25 Murnaghan J P. Adhesive capsulitis of the shoulder: Currentconcepts and treatment. Orthopedics 1988; 11: 153-8.

26 Stiles R G, Otte M T. Imaging of the shoulder. Radiology1993; 188: 603-13.

27 Liou J T S, Wllson A J, Totty W G, Brown J J. Thenormal shoulder: common variations that simulatepathologic conditions at MR imaging. Radiology 1993;186: 435-41.

28 Frieman B G, Albert T J, Fenlin J M. Rotator cuff disease:a review of diagnosis, pathophysiology, and current trendsin treatment. Arch Phys Med Rehabil 1994; 75: 604-9.

29 Zuckerman J D, Mirabello S C, Newman D, Gallagher M,Cuomo F. The painful shoulder: Part II. Intrinsicdisorders and impingement syndrome. Am Fam Physician1991; 43: 497-512.

30 Windt D A W M van der, Heijden G J M G van der,Scholten R J P M, Koes B W, Bouter L M. The efficacyof non-steroidal anti-inflammatory drugs (NSAIDs) forshoulder complaints. A systematic review. 7 Clin Epidemiol1995;48: 691-704.

31 Heijden G J M G van der, Bouter L M, Beckerman H,Bie R A de, Oostendorp R A B. De effectiviteit vanfysiotherapie bij schouderklachten. Een geblindeerdliteratuuronderzoek. Ned Tijdschr Fysiother 1992; 102:38-46.

964 on M

arch 28, 2020 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.54.12.959 on 1 D

ecember 1995. D

ownloaded from