In this study, we reviewed patients with a purulent secondary elbow infection in particular for their mortality, definitive therapy, outcome and findings in bacterial cultures.
While we observed no mortality related to secondary elbow joint infections, we describe severe courses of this disease requiring multiple interventions in several cases, in one case even an amputation of the forearm. Three patients needed > 10 procedures, since the infection recurred after the initial treatment, in one patient this led to a resection arthroplasty and chronic fistula with 25 procedures in total. We found a high rate of 28% of patients receiving radical salvage procedures such as resection arthroplasty of the joint as treatment of the infection, underlining the severity of a septic arthritis of the elbow joint. In line with these findings, we observed a high median impairment regarding the range of motion of the respective elbow. The leading organism in our study causing the elbow joint infection was Staphylococcus aureus.
In general, septic arthritis of non-rheumatic patients without a prosthesis is a rare disease. Weston et al. found infections of the elbow and shoulder joint in poly-articular septic arthritis to be associated with a higher mortality in their multivariate analysis . In contrast, mortality of a monoarticular infection of the elbow joint in this study was determined 5%, while the overall mortality rate was 11.5% in septic arthritis . At first glance, our data is not in line with these findings: In our collective, no patient died from a secondary infection to the elbow joint during the period under review. However, patients suffering from diseases that affect the immune system (due to their medication or the underlying condition) were excluded in our collective. We believe the exclusion of patients treated with immunosuppressive therapy or underlying systematic diseases (such as patients treated for rheumatoid arthritis) accounts for the differences in our results as survival rates of previously healthy patients might differ substantially from those suffering from immunodeficiency. Therefore, we consider our data not to be contradictive to the aforementioned study.
We could also show septic arthritis to be a possible complication of previous elective surgical procedures close to the joint as is common in the surgical treatment of a tennis elbow or chronic bursitis, for example. This is in line with a study of Moon et al., who also presented otherwise healthy patients affected by septic arthritis of the elbow joint .
To our knowledge, the only study reporting specifically about septic arthritis in the elbow joint was performed by Mehta et al. in 2006 . In their study, the authors focused on patients with hematogenous septic arthritis and therefore specifically excluded all patients with open joints or previous surgical procedures on the elbow, making a direct comparison of our collectives difficult. Our findings in bacterial cultures were however in accordance with the findings of bacterial cultures in septic arthritis described in literature and in particular Mehta et al’s study [4, 9].
Given the severity of septic arthritis of the elbow joint shown in our study, great care must be taken not to undertreat patients with infections to the elbow joint, since misinterpretation of early symptoms (pain, swelling, warmth at the affected area) and thus missing early diagnosis can have disastrous effects. When evaluating patients with septic arthritis, diagnostic accuracy of laboratory values such measurement of CRP or WBC alone is insufficient [11, 12]. Another powerful tool in cases with a suspicion of a septic arthritis could be the WBC and percentage of polymorphonuclear cells in joint aspirate [11, 12]. This parameter can be determined even if Gram stains or bacterial cultures are not (yet) available. Margaretten et al. propose a leukocyte count of greater than 50,000 cells /mm3 as a diagnostic predictor for septic arthritis [11, 13]. However, a low WBC in the joint aspirate cannot rule out a septic arthritis .
Infections to the joint can also be present in cases of seronegativity. Gupta et al. found the outcome in patients with seropositive septic arthritis comparable to those with a high clinical suspicion but seronegative septic arthritis .
Therefore, patients where septic arthritis is suspected should immediately be treated operatively with drainage of the purulent effusion, surgical debridement followed by antibiotic treatment [4, 15, 16]. However, evidence is missing in regard to the application regimen of antibiotic treatment: In general, intravenous application of antibiotics is recommended . Empiric antibiotic treatment should generally focus on Staphylococcus aureus. However, treatment strategy should be adapted according to local resistances and include treatment of methicillin-resistant Staphylococcus aureus (MRSA) or other (resistant) organisms if indications for resistance are observed.
In our practice, we use intravenous Cefazolin as calculative first-line treatment for septic arthritis if an indication for a broader antibiotic treatment is missing (Fig. 1).
As for the operative technique, there seems to be no clear indication of a superiority of arthrotomy over arthroscopic drainage or needle-aspiration in the literature . In any case, arthroscopic drainage could be shown to be safe in elbow infections with the advantage of a minimal-invasive procedure while at the same time allowing an assessment of the joint . More radical procedures as seen in our study were necessary in patients with life-threatening conditions call for an immediate and complete removal of the septic arthritis. Destruction of the elbow joint after open fractures as well as an osteomyelitis on the humerus or the radius/ulna can require a resection of the joint. Life-long antibiotic treatment as suppression therapy or creation of a fistula as a last resort for non-curable infections remain the exception but may unfortunately be necessary in individual cases. In our proposed staged protocol, we do not consider the time delay between the index surgery or trauma and revision surgery as indicative for a specific treatment algorithm. Operative therapy in these patients is highly individual and a staged protocol can only serve as orientation for such treatment. In general, we do not recommend radical treatments solely based on time delays of secondary joint infections but suggest basing such decision on the local extent of the infection. Therefore, if an infection is strictly limited to the elbow joint, we primarily perform an (arthroscopic) debridement and lavage of the joint.
Our study has several limitations: We performed a retrospective observational study of patients treated with secondary septic arthritis in our institution. Long-term outcome after dismissal of the patients could therefore not be described. Also, due to the structure of the German health care system, patients are usually referred to a maximum care hospital, from a primary care facility or by the emergency services when patients are in a poor general condition. Milder cases might be treated in primary care facilities such as general or regional hospitals or in outpatient medical-care centers with a less severe outcome and fewer comorbidities. Therefore, our results could be somewhat distorted in regard to the severity and the outcome. Our results might hence not be comparable to a potential larger collective also including milder cases of elbow infections. However, due to the infrequency of the disease, a prospective single-center study to describe influencing factors in this disease might not be feasible. A larger study in a multi-center design with inclusion of primary care facilities might overcome these limitations. Furthermore, cases with sole microbial or typical radiological signs were disregarded in this study, leading to a selection bias to more severe cases where turbid fluid was detected.