After reviewing patients with septic arthritis of the digit, we found that as an empiric antibiotic treatment in the referred patients, traditional narrow-spectrum agents seemed to be inappropriate due to the predominance of MRSA. The referred patients required more surgical intervention including flap surgery, prolonged duration of antibiotic therapy, and longer hospital stay than the primary patients.
Although septic arthritis is generally considered as a surgical emergency [2, 15], treatments are often delayed when it affects the digits [1, 3]. In addition, diagnosis is difficult in patients with native joints, and can be further impeded due to postoperative pain, particularly in patients after finger surgery [3]. In the present study, 18 (53%) of the referred patients were initially diagnosed with osteoarthritis at the previous hospital. Discerning the early stage of septic arthritis from noninfectious conditions might be difficult, especially considering the high incidence of osteoarthritis in the elderly. Moreover, even after the physician suspected an infectious condition, most cases were regarded as superficial wound infection, cellulitis, and early fingertip infections, except the 8 cases with surgical drainage [16]. Thus, the inaccurate diagnosis might be one of the main reasons that caused treatment delay, and we believe that suspecting septic arthritis can shorten the treatment delay, when old patients present with swollen joints of the digit.
In septic arthritis of the native finger joints, S. aureus was the most frequent isolated microorganism, followed by Streptococcal species; further, MRSA was detected in lesser number of cases than was MSSA [1, 7, 15, 17, 18]. Thus, conventional narrow-spectrum antibiotics such as penams and the first-generation cephalosporins are recommended in primary cases. However, only 14 referred patients in this study were treated with conventional narrow-spectrum antibiotics during primary treatment. While the second- or third-generation cephalosporins were indicated for those with an allergy to penicillin or those at the risk of atypical organism infections [5], of the referred patients only two reported a history of animal bites and none reported penicillin allergies. Considering that 21 gram-positive cocci were confirmed among the 35 microorganisms identified in the referred patients, this inappropriate use of second- or third-generation cephalosporins might be another cause of treatment failure. Conventional narrow-spectrum antibiotics appeared to be less effective in the referred patients than in the primary patients. Cefazolin was changed to other agents in 24 referred patients and this delay in the use of effective antibiotics led to extended durations of antibiotic therapy and hospital stay. Considering the high incidence of MRSA in the referred patients, we believe that physicians should suspect MRSA as a causative microorganism if gram-positive cocci are identified. Although the conventional narrow-spectrum antibiotics had several advantages [19], agents covering MRSA from the beginning of the treatment period might have shortened the duration of therapy for the referred patients.
In cases combined with osteomyelitis, we partially or completely debrided the cartilage. As more cartilage and subchondral debridement were performed in the referred patients, more of these patients showed irreversible osteochondral damage than the primary patients. Additionally, referred patients often showed soft tissue defects that could not be rectified by secondary healing. To address osteochondral damage and soft tissue defects, previous studies have recommended intended arthrodesis with shortening as an additional procedure [1, 3]. In our cases, 20.5% of the referred patients required additional procedure and we performed flap surgery instead of intended arthrodesis, to avoid infection after internal fixation. Although we did not perform arthrodesis, some joints were spontaneously united, and there was only one complaint of instability. Thus, we believe that this additional procedure should be considered as a feasible option especially for the referred patients, and flap surgery might be used instead of arthrodesis.
Whereas several previous reports have focused on primary patients, our study presents data on patients referred due to uncontrolled infection. Furthermore, all patients in this study underwent operations from a single surgeon in a tertiary clinic; thus, consistency in the surgical techniques and antimicrobial protocols were assured. Although this study has some strengths, it also has weaknesses. First, both groups were not similar in terms of age and delay from symptom onset. In the referred group, both older age and longer treatment delay might also affect the hospital course negatively. Second, suboptimal dosage or insufficient surgical drainage was not considered as inappropriate treatments. Thus, the number of effective treatments might be counted higher than the real number. Third, our result could not give a clear guideline about how many surgical procedures are required. Surgeons can choose either primary extensive resection or sequential gradual resection. We can only conclude the extent of debridement may be beyond the cartilage in most of the referred cases. Future studies should include large cohorts to match age and treatment delay and should focus on the extent of surgical debridement.