CLINICAL NOTE

Shoulder Pain as an Unusual Presentation of Pneumoniain a Stroke Patient: A Case ReportWannapha Petchkrua, MD, Stacy A. Harris, MD

ABSTRACT. Petchkrua W, Harris SA. Shoulder pain as anunusual presentation of pneumonia in a stroke patient: a casereport. Arch Phys Med Rehabil 2000;81:827-9.

Etiologies of shoulder pain in the hemiplegic population,such as glenohumeral subluxation, frozen shoulder, and reflexsympathetic dystrophy (RSD), have been described extensively.We present an 89-year-old woman with right hemiparesissecondary to ischemic lacunar infarction who developed sud-den onset of right shoulder pain on the fifth day of inpatientrehabilitation. The pain was severe, limiting range of motion(ROM) and participation in therapy. Extensive investigations torule out subluxation, fracture, connective tissue disease, RSD,and pulmonary embolism were negative. Ultimately, her shoul-der pain and decreased ROM completely resolved with antibi-otic treatment for right lower lobe pneumonia. We conclude thather symptoms were possibly referred pain from diaphragmaticirritation transmitted via right C4 sensory axons in the phrenicnerve, which shares the same dermatome as the right acromionarea. This case was an unusual presentation of pneumonia in anelderly woman with hemiplegia. We recommend that pneumo-nia be considered in the differential diagnoses of shoulder pain.

Key Words: Stroke; Hemiplegia; Shoulder pain; Pain;Rehabilitation.

r 2000 by the American Congress of Rehabilitation Medi-cine and the American Academy of Physical Medicine andRehabilitation

SHOULDER PAIN is probably the most frequent complica-tion of hemiplegia.1 The pain causes structural or functional

impairment and subsequently leads to disability and handicap.2

Griffin3 reviewed 118 articles detailing relevant causes, preven-tion, and treatment of hemiplegic shoulder pain (HSP) andreported a 38% to 70% incidence of HSP. Poduri4 reported anincidence of 41.4% (48/116) within 3 years poststroke. How-ever, the causes of the 48% (23/48) were not specified. Anotherliterature review by Totta and Beneck,5 although limited byfactors such as heterogeneity of stroke population, variablenature of stroke recovery, and difficulty in objectifying studyparameters, summarized 5 major causes of HSP as glenohu-meral subluxation (GHS), frozen shoulder, impingement syn-drome, rotator cuff lesions, brachial plexus neuropathy, andreflex sympathetic dystrophy (complex regional pain syndrometype I). There are other causes of hemiplegic shoulder pain thatmust be considered in the evaluation, such as fractures,

preexisting arthritis, tumor, cervical disc disease, heterotopicossification (HO), and radicular and referred pain.

Common comorbid diseases such as hypertension, hyperten-sive cardiovascular disease, heart disease, obesity, diabetesmellitus, and arthritis are found more frequently in persons withstroke than in stroke-free age matched controls (p , .05).6

Because the majority of persons with stroke are geriatric,6

referred pain of cardiac origin should always be considered.7

Considering all differential diagnoses, the evaluation plansuggested by Griffin3 includes detailed medical history, athorough diagnostic examination of neck and shoulder toidentify GHS, cervical spine pathologies, rotator cuff or bicepsbrachii tendinitis, plus additional tests of x-ray, magneticresonance imaging (MRI), arthrography, and electrodiagnosticstudy as considered necessary. If all tests are negative andsymptoms persist, the clinician should pursue less commonetiologies. We report an uncommon cause of shoulder pain in anelderly patient with stroke.

CASE REPORTThe patient was an 89-year-old woman who was admitted to

the hospital with a chief complaint of sudden right-sidedweakness and slurred speech. Her history was significant forosteoporosis, localized colon cancer requiring a partial colec-tomy, and elective cholecystectomy. Initial computed tomogra-phy of the head without contrast was unremarkable, with theexception of incidental bilateral calcification in the basalganglia. Echocardiography revealed normal left ventricularfunction without mural thrombus or atrioseptal defect.

On the second hospital stay, MRI of the brain with gadolin-ium revealed ischemic leukoencephalopathy with scatteredlacunar infarctions on the left side. A diagnosis of left internalcapsule white matter infarct was made. The patient wastransferred to our rehabilitation unit 3 days after admission forcomprehensive inpatient rehabilitation. At that point, she haddeveloped minimal right-sided spasticity (Ashworth I).

The patient performed well in therapy without any com-plaints until day 5, when she complained of the sudden onset ofsevere right shoulder pain with mild right-sided rib pain. Shedenied any associated shortness of breath, pleuritic pain, cough,or any gastrointestinal symptoms. Her vital signs were normalwith 97% oxygen saturation on room air. There was no edema,tenderness, or skin change of the right hand. Musculoskeletalexamination revealed normal cervical range of motion (ROM),no right shoulder subluxation, significant decreased ROM withtenderness on palpation at the right glenohumeral joint, andmild tenderness of all right anterior ribs. No new neurologicdeficits were detected. Continued evaluation included rightshoulder x-ray, which showed a questionable right acromiocla-vicular (AC) joint separation. Venous Doppler revealed noevidence of a deep vein thrombosis, electrocardiogram (ECG)was negative for acute change, and chest x-ray was read asnegative by a radiologist. We prescribed a right shoulder slingfor immobilization because of possible rightAC joint separation,

From Schwab Rehabilitation Hospital and Care Network/University of ChicagoHospitals, Chicago.

Submitted February 11, 1999. Accepted in revised form December 18, 1999.Presented in part as a poster at the Annual Assembly of the American Academy of

Physical Medicine & Rehabilitation, November 5-8, 1998, Seattle.No commercial party having a direct or indirect interest in the subject matter of this

article has or will confer a benefit upon the authors or upon the organization withwhich the authors are associated.

Reprints will not be available from the authors.0003-9993/00/8106-5460$3.00/0doi:10.1053/apmr.2000.7157

827

Arch Phys Med Rehabil Vol 81, June 2000

but it was discontinued after an orthopedist concluded that therewas no evidence of AC joint separation.

The chest x-ray was again reviewed, and because of question-able right heart border haziness, we started antibiotic treatmentwith clarithromycin. A rheumatologist was consulted becauseof the high erythrocyte sedimentation rate (ESR) (95) andrheumatoid factor (175). All blood tests to screen for connectivetissue disease (ANA, protein C, C3 complement, and serumprotein electrophoresis) were normal. Reflex sympathetic dys-trophy seemed less likely with mere shoulder pain presentation.

On the next day, the patient developed greenish sputum andher temperature was 38°C. White blood cell count was 11,600/µL with 85.9% neutrophils. A diagnosis of probablepneumonia was made. A treatment program of hydration, chestphysiotherapy with postural percussion and drainage of theright lung, oral expectorants, incentive spirometry, and antibiot-ics was begun. Two days later, a chest x-ray showed infiltrationand atelectasis involving the medial aspect of the right lowerlobe (fig 1). The shoulder pain resolved by the 7th day ofantibiotic and drainage treatment. Shoulder ROM returned tonormal.

DISCUSSIONGeriatric shoulder pain is common. Its differential diagnosis

was summarized by Zuckerman and Shapiro.8 Our patientposed a second challenge for her geriatric hemiplegic shoulderpain (GHSP). After reviewing the studies by Zuckerman andShapiro,8 Totta and Beneck,5 and Poduri,4 we considered adifferential diagnosis for GHSP (table 1). In our case, a detailedhistory and physical examination effectively narrowed thefocus to a few etiologies. Inflammatory arthritis was consideredbecause of the high ESR and rheumatoid factor, but ANA,protein C, C3 complement, and serum protein electrophoresiswere negative. The last three intrinsic causes (fracture, HO, andcalcific tendinitis) required a radiograph (bone scan if

HO presents early). The x-ray of the right shoulder showed noevidence of fracture and calcification, but was suspicious forright AC joint disruption; however, the result was normal,according to an orthopedist. Symptomatic HO was not likely asearly as 8 days after stroke.9 Hence, we did not pursue the triplephase bone scan.

For extrinsic disorders, the history and examination made itless likely to be spasticity, improper exercise, or handling of theparalyzed limb, or postural pain. The chest x-ray did not showevidence of lung cancer or metastatic tumors. The patient’sserum electrophoresis did not suggest multiple myeloma.Interestingly, the cervical spine pain was atypical, with tender-ness on palpation at right shoulder and mild tenderness of allright anterior ribs. Therefore, we planned further electrodiagnos-tic studies and cervical MRI if other evaluations did not confirma diagnosis.

Cardiac ischemia was another possibility because of her riskfor coronary artery disease (CAD). However, the pain locationand atypical characteristic of the pain (sharp and reproduciblewith chest wall palpation) with negative ECG made CAD amuch less likely explanation.

Stroke was a risk factor for deep vein thrombosis (DVT) andits major morbidity/mortality resulting from pulmonary embo-lism (PE). Often, the clinical signs and symptoms of PE arenonspecific.10 Therefore, DVT with PE remained in our differ-ential diagnosis. However, the patient had received low dosesubcutaneous heparin from the first day of admission to themedical unit, her calves were without swelling or tendernessbilaterally, and venous Doppler studies of lower extremitieswere negative for DVT on the day with symptomatic shoulderpain. Recently, a comprehensive literature review by Anand etal11 established the positive likelihood ratio and diagrammed analgorithm for diagnostic approach to patients with suspectedDVT. This patient was classified as moderate risk with normalvenous Doppler. Therefore, her positive likelihood ratio (95%confidence interval) would be merely .17 instead of 72 if

Table 1: Causes of Geriatric Hemiplegic Shoulder Pain

Intrinsic disordersGlenohumeral subluxationFrozen shoulderImpingement syndromeRotator cuff lesionsBrachial plexus neuropathyReflex sympathetic dystrophy (complex regional pain syndrome

type I)Preexisting degenerative/inflammatory arthritisBicipital tendinitisFracturesHeterotopic ossificationCalcific tendinitis

Extrinsic disordersImproper handling of the paralyzed armImproper exerciseSpasticityPostural pain from prolonged sittingCervical spine disordersNeoplastic disease, ie, multiple myeloma, Pancoast’s tumor,

and lymphomaPulmonary embolismReferred pain from cardiac, pulmonary, and intra-abdominal

organsIdiopathic

Modified and reprinted with permission.8

Fig 1. Anterior/posterior chest x-ray 2 days after symptoms started.Pneumonia was confirmed with a lateral film as well.

828 GERIATRIC HEMIPLEGIC SHOULDER PAIN, Petchkrua

Arch Phys Med Rehabil Vol 81, June 2000

positive venous Doppler. This algorithm recommended a repeatvenous Doppler in 1 week.

Referred pain from right upper quadrant abdominal organssuch as the liver and gall bladder was possible but the patientalready had a cholecystectomy. She did not have loss ofappetite, nausea, or vomiting.

Pneumonia has been estimated to occur in about 30% ofpatients with stroke.12 Common presentations of pneumoniainclude fever, chills, rigors, and cough.13 Nevertheless, Harperand Newton14 reported that the absence of classic signs andsymptoms was correlated with advanced age (.65) and cogni-tive and baseline functional impairment on admission.

After considering the little likelihood ratio for DVT and PE,plus the questionable haziness of the right heart border, atreatment program for presumed pneumonia was initiated.Within 7 days of pulmonary toilet and antibiotic treatment, thepatient was pain free and attained normal passive shoulderROM. We decided against a repeat venous Doppler, MRI, orelectrodiagnostic studies. Our assumption was that diaphrag-matic irritation from the right lower lobe pneumonia causedreferred pain transmitted via right C4 sensory axons in thephrenic nerve, which shares the same dermatome supplying theright acromion area.15

CONCLUSIONShoulder pain and decreased ROM in an elderly stroke

patient may be due to unexpected causes such as pneumonia.Ahighindex of suspicion andearly treatment may prevent morbidity.

Acknowledgments: The authors acknowledge the critical editingand suggestions of Jonathan L. Costa, MD, PhD, and StevenA. Stiens, MD.

References1. Ouwenaller CV, Laplace PM, Chantraine A. Painful shoulder in

hemiplegia. Arch Phys Med Rehabil 1986;67:23-6.2. Stiens SA, Haselkorn JK, Peters DJ, Goldstein B. Rehabilitation

intervention for patients with upper extremity dysfunction: chal-lenges of outcome evaluation. Am J Ind Med 1996;29:590-601.

3. Griffin JW. Hemiplegic shoulder pain. Phys Ther 1986;66:1884-93.

4. Poduri KR. Shoulder pain in stroke patients and its effect onrehabilitation. J Stroke Cerebrovasc Dis 1993;3:261-6.

5. Totta M, Beneck S. Shoulder dysfunction in stroke hemiplegia.Phys Med Rehabil Clin North Am 1991;2:627-41.

6. Gresham GE, Phillips TF, Wolf PA, McNamara PM, Kannel WB,Dawber TR. Epidemiologic profile of long term stroke disability:the Framingham study. Arch Phys Med Rehabil 1979;60:487-91.

7. Roth EJ. Medical complications encountered in stroke rehabilita-tion. Phys Med Rehabil Clin North Am 1991;2:563-78.

8. Zuckerman JD, Sharpiro I. Geriatric shoulder pain: commoncauses and their management. Geriatrics 1987;42:43-58.

9. Calliet R. Shoulder pain. 3rd ed. Philadelphia: FA Davis; 1991.10. Harvey RL, Green D. Deep vein thrombosis and pulmonary

embolism in stroke. Top Stroke Rehabil 1996;3:54-70.11. Anand SS, Wells PS, Hunt D, Brill-Edwards P, Cook D, Ginsberg

JS. Does this patient have deep vein thrombosis? JAMA 1998;279:1094-9.

12. Walker AE, Robins M, Weinfeld FD. Clinical findings. Thenational survey of stroke. Stroke 1981;12(1 Suppl):I13-37.

13. Musgrave T, Verghese A. Clinical features of pneumonia in theelderly. Semin Respir Infect 1990;5:269-75.

14. Harper C, Newton P. Clinical aspects of pneumonia in the elderlyveteran. J Am Geriatr Soc 1989;37:867-72.

15. Williams PL, Warwick R, Dyson M, Bannister LH, editors. Gray’sanatomy. Edinburgh: Churchill Livingston; 1989.

829GERIATRIC HEMIPLEGIC SHOULDER PAIN, Petchkrua

Arch Phys Med Rehabil Vol 81, June 2000