Abstract
Aim
Infection is one of the worst complications following total joint arthroplasty, which is often associated with significant morbidity. Currently, due to the global burden of multidrug-resistant Gram-negative bacteria (MDR-GNB) infections, few multicentre studies have described a microbiological shift from Gram-positive cocci (GPC) towards MDR-GNB PJI (prosthetic joint infection). Additionally, the emergence of MDR-GNB impacts the therapeutic options and may increase the rate of PJI treatment failure. The purpose of the present study was to describe the predisposing factors associated to failure of treatment in an orthopaedic reference hospital in Brazil from 2014 through 2019.
Method
Retrospective case-control analysis of patients treated for MDR-GNB PJI over a five-year period. Data were collected from medical, surgical and laboratory records. PJI were defined according the current MSIS criteria. MDR was defined as non-susceptibility to at least one agent in three or more antimicrobial categories. Patients with PJI with at least two positive tissue cultures for MDR-GNB were selected. The control group was patient with PJI caused by multisensitive organism (GNB or GPC). Absence of signs and symptoms of infection during the follow-up period was defined as cure. Definition of failure: death, need for another course of antibiotic, or the need for another surgical procedure to control the infectious site (relapse).
Results
A total of 104 patients were selected, 59 patients in the MDR-GNB PJI group and 44 in the control group. Two outcomes were compared: cure or failure. The overall 1-year survival rate was 65.3% with the median survival time being 207.08 days. In the MDR-BGN infection group the 1-year survival rate was 59.3% and the average time of survival was 141.14 days. In contrast, in the Control group the 1-year survival rate was 73.8% with an average survival time of 230.29 days (p = 0.023). HR: 2.447, IC 1.099–5.448. The independent variables in the multivariate analysis associated to treatment failure were MDR-BGN infection (p = 0.023) HR 2,447 IC 1,099 –5,448, revision surgery (p = 0.042) HR 2,027 IC: 2,027–4,061, presence of comorbidities (p = 0.048) HR 2,508 IC: 0,972- 6,469 and previous antimicrobial use in the last 3 months (p = 0.022). HR 2,132 IC: 1,096- 4,149.
Conclusions
GNB-MDR PJI increases approximately 2.5 times the chance of unfavourable outcome such as death and infectious relapse compared to infections with other multisensitive microorganism.
