Treatment patterns and reoperation rates in periprosthetic joint infection identified 1 - 6 months after arthroplasty in Japan: A descriptive analysis using nationwide claims data.

OBJECTIVE

To describe the treatment patterns and reoperation rates after claims-defined postoperative periprosthetic joint infection (PJI) identified 1 - 6 months after total knee or hip arthroplasty using a Japanese nationwide claims database.

MATERIALS AND

METHODS

Patients who underwent total knee or hip arthroplasty between 2005 and 2017 were identified. Claims-defined postoperative PJI was operationally defined as infection-related claims accompanied by systemic antibacterial treatment in 1 - 6 months after arthroplasty. The agents targeting methicillin-resistant Staphylococcus aureus (MRSA) included vancomycin, daptomycin, linezolid, and teicoplanin. The primary outcome was reoperation after claims-defined postoperative PJI.

RESULTS

Among the 2,805 patients, 59 (2.1%) developed postoperative PJI 1 - 6 months after arthroplasty. Of these, 8 underwent immediate surgical intervention, and the remaining 51 patients were initially managed with antibiotic treatment without surgical intervention. Of these, 12 received anti-MRSA agents and 39 received non-MRSA agents. Reoperation was performed in 3 of 12 (25.0%) patients who received anti-MRSA agents and 6 of 39 (15.4%) patients who did not receive anti-MRSA agents. Among the patients treated with anti-MRSA agents, 11 of 12 (91.7%) required second- or third-line therapy. Diabetes mellitus was present in 66.7% and 83.3% of the reoperated patients who received and did not receive anti-MRSA agents, respectively; liver disease was observed in 66.7% and 16.7% of the patients, respectively.

CONCLUSION

This claims-based descriptive analysis indicated that treatment escalation and reoperation were observed after claims-defined postoperative PJI identified 1 - 6 months after arthroplasty. Because pathogen data were unavailable and receipt of anti-MRSA agents was used as a treatment-based classification, these findings should be interpreted as descriptive treatment-pattern data rather than microbiologically confirmed comparisons.