What Is the Incidence of and Outcomes After Debridement, Antibiotics, and Implant Retention (DAIR) for the Treatment of Periprosthetic Joint Infections in the AJRR Population?

Abstract

Background: Debridement, antibiotics, and implant retention (DAIR) is used to manage acute periprosthetic joint infections (PJIs) after total joint arthroplasty (TJA). Given the uncertain success of single or multiple DAIR attempts and possible long-term deleterious effects this treatment can create when trying to treat persistent infection, it is important to understand the frequency with which surgeons in the United States are attempting multiple debridements for PJI and whether those procedures are achieving the desired goal.

Question/purposes: In the context of the American Joint Replacement Registry (AJRR), we asked: (1) What proportion of patients who undergo DAIR have only one DAIR, and what percentage of those patients have more than one? (2) Of the patients who undergo one or more DAIR procedures, what is the proportion who progress to additional surgical procedures? (3) What is the cumulative incidence of medical or surgical endpoints related to infection on the affected leg (other than additional DAIR procedures)?

Methods: DAIR procedures to treat PJI, defined by ICD-9/10 and CPT (Current Procedural Technology) codes, reported to the AJRR from 2012 to 2020 were merged with Centers for Medicare and Medicaid Services (CMS) data from 2012 to 2020 to determine the incidence of patients aged 65 and older who underwent additional PJI-related procedures on the same joint. Linking to CMS ensures no loss to follow-up or patient migration to a non-AJRR site. As of 2021, the AJRR captures roughly 35% of all arthroplasty procedures performed in the United States. Of the total 2.2 million procedures in the AJRR, only 0.2% of the procedures were eligible based on our inclusion criteria. Additionally, 61% of the total population is Medicare eligible, and thus, these patients are linked to CMS. Of the 5029 DAIR attempts after a TKA, 46% (2318) were performed in female patients. Similarly, there were a total of 798 DAIR attempts after a THA, and 50% (398) were performed in female patients. For the purposes of decreasing confounding factors, bilateral THAs and TKAs were excluded from the study population. When querying for eligible procedures from 2012 to 2020, the patient population was limited to those 65 years and older, and a subsequent reoperation for infection had to be reported after a primary TJA. This limited the patient population as most infections reported to AJRR resulted in a revision, and we were searching for DAIRs. Although 5827 TJAs were identified as a primary TJA with a subsequent infectious event, more than 65% (3788) of that population did not have a reported event. The following conditions were queried as secondary outcomes after the first DAIR: sepsis, cellulitis, postoperative infection, endocarditis, amputation, knee fusion, resection, drainage, arthrotomy, and debridement. To answer our first and second study questions, we used frequency testing from the available AJRR data. Because of competing risks and issues with incomplete data, we used the cumulative incidence function to evaluate the outcomes specific to study question 3.

Results: Of the patients who underwent DAIR, 93% (5406 of 5827) had one DAIR and 8% (421 of 5827) had more than one. Among the DAIR population, at least 35% of TKAs and 38% of THAs were identified as having experienced an additional PJI-related event (an additional surgical procedure on the same joint, sustained an infectious endpoint in the linked CMS-AJRR dataset, or they had died). The cumulative incidence of developing a further medical or surgical condition related to the joint that had the initial DAIR were as follows: 48% (95% CI 42% to 54%) at 8 years after a DAIR following a TKA and 42% (95% CI 37% to 46%) at 4 years after a DAIR following a THA. The timepoints for TKA and THA are different because there are more longitudinal procedure data available for TKAs regarding DAIR procedures than for THAs.

Conclusion: In this study, we used data from the AJRR to assess the incidences of single and multiple DAIR attempts and additional surgical- and infection-related sequalae. Continued investigation is required to determine the fate of infected joints that undergo DAIR with regard to ultimate patient outcome. Future cross-sectional studies using large datasets are necessary to assess functional outcomes and determine the risk of persistent infection after DAIR more precisely.

Level of evidence: Level III, therapeutic study.