Bacterial septic arthritis is a serious health problem, associated with considerable morbidity.6,16 While most cases are limited to the large joints of the lower extremities, involvement of the upper extremity joints does occur, although it has a lower prevalence.8 Despite numerous studies pertaining to septic joints in the lower extremity, there are few studies on upper extremity septic arthritis.2,5,7,9,11,12,14
One study published in 1965 identified eight patients with septic arthritis of the glenohumeral joint.7 Subsequent studies reviewed a series of 7 to 23 patients with septic shoulder arthritis.2,5,9,11,12 Another study reviewed 28 patients with septic arthritis of the wrist in terms of range of motion (ROM), grip strength, and subjective complaints of discomfort and disability.14 The existing studies focus on the outcome of septic arthritis but offer limited discussion of diagnostic aids, even though some of the reported patients had delayed diagnosis and treatment.5,7,9 The white blood cell (WBC) count in the joint aspirate was reported by two studies and demonstrated considerable variability, with values ranging from 35,000 to 259,000/mm3 in one study5 and from 20,000 to 150,000/mm3 in the other.14 No study provided details on the differential count to evaluate its diagnostic usefulness.
There is also variability in the reported microbiology of septic arthritis in the upper extremity; Staphylococcus aureus was identified in 41% to 87% of cases, whereas Gram-negative organisms were identified in 12% to 22% of cases.5,9,11,14 Only one of these studies11 commented on empiric antibiotic therapy; the authors proposed empiric therapy should cover not only Staphylococcus aureus, but Gram-negative organisms as well in a subgroup of patients. Moreover, methicillin resistant Staphylococcus aureus (MRSA) is gradually emerging as a pathogen,4 and a recent study of septic arthritis of the shoulder reported an MRSA prevalence of 17%.2 Because of the infrequency of septic arthritis of the large upper extremity joints, additional information on the spectrum of organisms associated with these infections may assist the clinician in selecting appropriate initial empiric antibiotic therapy.
We determined the organisms associated with septic arthritis of the shoulder, elbow, and wrist. We also ascertained the ranges of values for the joint aspirate WBC, and finally the frequency of patient co-morbidities.
MATERIALS AND METHODS
This is a retrospective observational study of the diagnostic and microbiological features of septic arthritis of the shoulder, elbow, and wrist in patients admitted to the musculoskeletal infection ward at our facility from 1994 to 2004. We excluded children (< 18 years), because the vast majority of patients treated in our institution are adults. We excluded patients with an open joint, and patients who had any previous surgery to the affected joint in order to focus on hematogenous septic arthritis. We identified 52 adult patients with septic arthritis of the shoulder, elbow, or wrist and reviewed their medical records. Septic arthritis involved the elbow in 23 patients, the shoulder in 17, and the wrist in 12 patients. Forty-three patients (83%) were male and nine (17%) were female. The mean age was 44.3 years (range, 19-72 years). All patients presented with pain, swelling, and decreased ROM of the affected joint. The clinical diagnosis of septic arthritis was confirmed by aspiration/drainage of pus from the affected joint and/or by positive synovial fluid or blood cultures and/or prompt clinical response to drainage and antibiotic therapy.
All patients underwent a surgical drainage and irrigation procedure. Postoperatively, antibiotics were administered for 3 to 4 weeks.
For each patient we documented age, gender, comorbid conditions, history of trauma to the joint, history of intravenous drug use, laboratory and joint aspirate findings, and culture results.
Staphylococcus aureus was the most frequent pathogen, present in 76% (19/25) of the aspirates with positive cultures (Table 2); MRSA was identified in two (11%) of these 19 cases, resulting in an 8% (2/25) prevalence of MRSA in culture-positive cases. Streptococci species were isolated in six patients (24% of patients with positive cultures), and Gram-negative organisms in two patients (8% of patients with positive cultures). Cultures of the 25 aspirates were monomicrobial for 23 patients (92%) and polymicrobial (two organisms) for two patients (8%). Approximately half of cultures (27 of 52) had no growth at 48 hours, most likely because empirical antibiotics were started in 20 of these 27 patients (74%) before joint fluid aspiration and culture either in our emergency department or in another institution before referral to our center. In the patients with negative cultures the diagnosis of septic arthritis was based on the clinical findings and was confirmed by aspiration of pus in 10 patients and by positive peripheral blood culture in 2 patients (both grew Streptococcus pneumoniae on blood culture). The remaining 15 patients demonstrated elevated WBC count (54,000- 94,500 per mm3) and differential (80-96%), absence of crystals in the joint fluid, and a prompt clinical response to drainage and antibiotic therapy.
Forty-six of the 52 patients had WBC results of joint aspiration available for analysis. The mean WBC count was 136,300 per mm3 (range, 10,450-880,000 per mm3). Thirty-five of 46 aspirates (76%) had a WBC count greater than 50,000 per mm3. Evaluation of the WBC differential of the joint aspirate demonstrated the average value of polymorphonuclear leukocytes was 91% (range, 72- 100%). All aspirates had greater than 70% polymorpho- nuclear leukocytes and 91% of aspirates (42 of 46 aspirates) had greater than 85% polymorphonuclear leukocytes.
Twenty-three of the 25 patients with positive joint fluid cultures had WBC results of joint aspiration available for analysis. In this subgroup, the mean WBC count was 145,700 per mm3 (range, 11,200-495,000 per mm3). Eighteen of 23 aspirates (78%) had a WBC count greater than 50,000 per mm3. The mean proportion of polymorpho- nuclear leukocytes was 92% (range, 78-100%). All aspirates had greater than 70% polymorphonuclear leukocytes and 96% of aspirates (22 of 23 aspirates) had greater than 85% polymorphonuclear leukocytes.
The erythrocyte sedimentation rate (ESR) was available for 25 patients and was elevated in all of them, with a mean value of 76 mm/hour, ranging from 28 to 140 mm/hour (normal values are 0-20 mm/hour). The C-reactive protein (CRP) was available for 40 patients and was elevated in 38 (95%) of them. The mean CRP value was 15.2 mg/dL, ranging from 0.2 to 54.4 mg/dL (normal values are < 0.6 mg/dL). The peripheral white blood cell (WBC) count was available for 48 patients with an average of 12.8 × 109 per liter (range, 4.6-30.1 × 109 per liter). Twenty of these 48 patients (42%) had a WBC count of ≤ 10 × 109 per liter.
Seventeen of the 52 patients (33%) had at least one serious medical condition. Fourteen of our 52 patients (27%) reported a history of intravenous drug abuse (IVDA). Eleven of these IVDA patients were otherwise healthy, whereas three had medical comorbidities (one patient had HIV infection, one had hepatitis C, and one had diabetes mellitus and hepatitis C). Overall, 28 of 52 patients (54%) were compromised hosts due to either of the above comorbidities (Table 1). Twenty-six patients (50%) were unemployed and seven of them were homeless. The other 26 patients were employed in jobs requiring the use of their upper extremities such as construction and machine operation. Only 12 of the 52 patients (23%) reported trauma to the affected joint within the month before admission.
Septic arthritis of a large joint of the upper extremity is an uncommon clinical entity, as evidenced by identifying only 52 patients with this diagnosis treated over an 11-year period at our musculoskeletal infection ward. The limited literature on septic arthritis of the upper extremity is characterized by variability in the microbiology of these infections and has not assessed the diagnostic role of the joint fluid WBC count and differential.2,5,7,9,11,12,14 The current study identifies Staphylococcus aureus as the most common pathogen and reports the high sensitivity of elevated polymorphonuclear WBC in the joint fluid.
We note several limitations. This is a retrospective study and complete data were not available for all patients. Moreover, our study does not provide follow-up data to evaluate control of infection and functional outcome. Many of these infections occur in individuals who for one reason or another are difficult to follow. Due to the uncommon nature of these infections it may not be feasible for a single institution to conduct a prospective study which could ascertain the influence of confounding variables on outcome.
Staphylococcus aureus was the most common organism in our series, cultured in 76% of joint aspirates with positive cultures (Table 2). In the literature, Staphylococcus aureus was identified in 41% to 87% of cases, whereas Gram-negative organisms were identified in 12% to 22% of cases.5,9,11,14 A review of published cases of septic arthritis of the shoulder reported Staphylococcus aureus was present in 41% of cases, whereas Gram-negative organisms were identified in 35% of cases11; the authors suggested the addition of empiric antibiotic coverage for Gram-negative organisms in neonates, infants, elderly, and immunocompromised patients. No other study on septic arthritis of the upper extremity commented on empiric antibiotic therapy.
The preponderance of Gram positive cocci in our series confirms the need for empiric antibiotic coverage for Staphylococcus aureus. It should be noted methicillin resistant Staphylococcus aureus (MRSA) is gradually emerging as a pathogen.4 MRSA was identified in 8% of patients with positive cultures in our series. Ross15 reported that in the last 5 years 25% of septic arthritis cases treated in two medical centers were caused by MRSA. Community-acquired MRSA septic arthritis has been reported4 and a recent study of septic arthritis of the shoulder found an MRSA prevalence of 17%.2 The specific antibiotic may be chosen based on the antibiogram of the treating institution and vancomycin may be considered, especially if risk factors for an MRSA infection are present.1
Disseminated gonococcal infection is the most common systemic complication of acute gonorrhea and is also the most common cause of septic arthritis in patients younger than 30 years.3 However, in our series and in previous studies5,7,9,11,12,14 there were no patients with septic arthritis of the shoulder, elbow, or wrist from Neisseria gonorrhoeae. It should also be noted infections with Myco- bacterium tuberculosis or Mycobacterium avium complex were not identified as an important cause of septic arthritis in our cohort.
Early diagnosis is imperative for optimal treatment results.5,9 Studies have shown persistent pain and limitation of motion after delayed diagnosis.5,7,9 Gelberman et al5 reported delayed diagnosis of shoulder septic arthritis in 6 of 16 patients and attributed the delay to the low index of suspicion of the treating physicians.
Joint fluid analysis is a useful diagnostic tool for septic arthritis, but not without problems.15 The WBC count in the joint aspirate usually exceeds 50,000/mm3 but is not a very sensitive diagnostic tool.10,13,15 In the current series the WBC count was greater than 50,000 per mm3 in 76% of aspirates, which is in agreement with a recent study reporting a WBC count greater than 50,000/mm3 in 64% of patients with septic arthritis.10
Interestingly, previous studies on septic arthritis of the large upper extremity joints did not focus on the diagnostic usefulness of the joint aspirate. Leslie et al9 did not report any joint aspirate data in their series of septic shoulders. Rashkoff et al14 mentioned the absolute count ranged from 20,000 to 150,000 per mm3 in patients with septic wrists, but did not provide any data on the differential. Gelberman et al5 were the only authors to report that patients with septic arthritis of the shoulder had a joint fluid WBC differential with a strong predominance of polymorpho- nuclear leukocytes. However, they did not provide any quantitative data as an aid to the diagnosis.5 In our series, the differential was greater than 70% of polymorpho- nuclear leukocytes in all evaluated aspirates, and greater than 85% of polymorphonuclear leukocytes in 91% of aspirates. Assessment of the subgroup of aspirates with positive cultures demonstrated 96% of aspirates had greater than 85% polymorphonuclear leukocytes. Based on our findings, we believe elevation of polymorpho- nuclear leukocytes in the joint fluid may be an important diagnostic feature.
The ESR and the CRP values, when available, were elevated in all but one patient with a septic upper extremity joint. However, both are indicators of acute inflammation, and as such are nonspecific. The peripheral WBC count was within normal limits in 42% of patients (20 of 48 patients with WBC results available), which is consistent with previous studies. Leslie et al9 reported a normal peripheral WBC count in 38% of patients with septic shoulders. Therefore, the peripheral WBC count may be normal even in the presence of infection, making diagnosis difficult without aspiration of the joint.
The mean age of our patients was 44.3 years (range, 19 to 72 years) and comorbidities were present in 54% of patients (28 of 52). This is a younger and healthier cohort compared to previous studies, in which the mean ages of patients ranged from 57 to 65 years and comorbidities were present in more than 90% of patients.5,7,9,11 Leslie et al9 concluded septic arthritis of the shoulder is associated with aging of the population and increased survival of elderly patients with chronic, debilitating disease; the authors stated this infection rarely develops in young adults and healthy individuals of any age. Our patient cohort includes young and healthy patients as well, which emphasizes the importance of a high index of suspicion for the diagnosis of infection in all patients presenting with shoulder complaints, irrespective of age and medical status.
Staphylococcus aureus is the most common pathogen isolated in septic arthritis of the shoulder, elbow, or wrist, and should be targeted by empiric antibiotic therapy. The joint aspirate WBC count differential may be a useful diagnostic feature, especially if greater than 85%. A high index of suspicion is necessary for the diagnosis of septic arthritis in a large joint of the upper extremity, irrespective of patient age and medical status.
1. Brumfitt W, Hamilton-Miller J. Methicillin-resistant Staphylococcus aureus. N Engl J Med
Cleeman E, Auerbach JD, Klingenstein GG, Flatow EL. Septic arthritis of the glenohumeral joint: a review of 23 cases. J Surg Orthop Adv
3. Cucurull E, Espinoza LR. Gonococcal arthritis. Rheum Dis Clin North Am
4. Fridkin SK, Hageman JC, Morrison M, Sanza LT, Como-Sabetti K, Jernigan JA, Harriman K, Harrison LH, Lynfield R, Farley MM. Methicillin-resistant Staphylococcus aureus disease in three communities. N Engl J Med
5. Gelberman RH, Menon J, Austerlitz MS, Weisman MH. Pyogenic arthritis of the shoulder in adults. J Bone Joint Surg Am
. 1980;62: 550-.553.6.
Kaandorp CJ, Krijnen P, Moens HJ, Habbema JD, van Schaardenburg D. The outcome of bacterial arthritis: a prospective community-based study. Arthritis Rheum
7. Kelly PJ, Coventry MB, Martin WJ. Bacterial Arthritis of the Shoulder. Mayo Clin Proc
8. Kelly PJ, Martin WJ, Coventry MB. Bacterial (suppurative) arthritis in the adult. J Bone Joint Surg Am
9. Leslie BM, Harris JM3rd, Driscoll D. Septic arthritis of the shoulder in adults. J Bone Joint Surg Am
10. Li SF, Henderson J, Dickman E, Darzynkiewicz R. Laboratory tests in adults with monoarticular arthritis: can they rule out a septic joint? Acad Emerg Med
11. Lossos IS, Yossepowitch O, Kandel L, Yardeni D, Arber N. Septic arthritis of the glenohumeral joint. A report of 11 cases and review of the literature. Medicine (Baltimore)
12. Master R, Weisman MH, Armbuster TG, Slivka J, Resnick D, Goergen TG. Septic arthritis of the glenohumeral joint. Unique clinical and radiographic features and a favorable outcome. Arthritis Rheum
13. McCutchan HJ, Fisher RC. Synovial leukocytosis in infectious arthritis. Clin Orthop Relat Res
14. Rashkoff ES, Burkhalter WE, Mann RJ. Septic arthritis of the wrist. J Bone Joint Surg Am
15. Ross JJ. Septic arthritis. Infect Dis Clin North Am
. 2005;19: 799-817.
16. Weston VC, Jones AC, Bradbury N, Fawthrop F, Doherty M. Clinical features and outcome of septic arthritis in a single UK Health District 1982-1991. Ann Rheum Dis