# Quantification of Sonicated Implants from Patients with Osteoarticular Implant Infections

## TL;DR

This study evaluates bacterial counts from sonicated implants in patients with joint infections, finding that higher counts are linked to polymicrobial infections and suggesting a threshold for diagnosis.

## Contribution

The study introduces a clinically meaningful threshold for sonication-based diagnosis of implant infections and highlights variability in bacterial counts.

## Key findings

- Polymicrobial infections showed significantly higher bacterial counts than monomicrobial infections.
- A threshold of ≥1000 CFU/mL was found to be clinically meaningful for diagnosis.
- Acute PJI tended to have higher bacterial burdens than chronic PJI, though not significantly.

## Abstract

Background: Sonication of retrieved implants has emerged as a valuable diagnostic adjunct for Prosthetic Joint Infection (PJI), particularly in chronic infections or cases with prior antibiotic exposure. Quantitative culture of sonication fluid has been proposed to differentiate contamination from true infection; however, the diagnostic thresholds remain inconsistent across studies and may be influenced by methodological variability. Objectives: We aimed to evaluate bacterial counts obtained from the routine sonication of osteoarticular implants and assess their diagnostic performance across different infection types. Methods: A retrospective study was conducted (2011–2023) at a tertiary hospital. Implants from patients with PJI or Fracture-Related Infection (FRI), classified according to international criteria, were processed using a standardized sonication protocol, including centrifugation and inoculation onto multiple culture media. Quantitative results were expressed as CFU/mL. Bacterial counts were compared across infection types (acute PJI, chronic PJI, FRI), microbial characteristics, infection pattern, and affected joint using non-parametric tests. Results: A total of 457 sonicated implants were analyzed, including 316 PJI samples (26.3% acute; 73.7% chronic) and 141 FRI samples. The median bacterial count was 40,000 CFU/mL (IQR 1000–100,000). No significant differences were found between prosthetic and osteosynthesis implants. Polymicrobial infections showed significantly higher counts than monomicrobial infections (p < 0.005). No significant differences were observed according to Gram stain or joint site. Acute PJI tended to show higher bacterial burdens than chronic PJI, although not significantly (p = 0.052). Conclusions: Quantitative sonication yields substantial variability in bacterial loads, with higher counts in polymicrobial infections and a trend toward increased counts in acute PJI. A threshold of ≥1000 CFU/mL appeared to be clinically meaningful within our protocol. These findings support the diagnostic utility of quantitative sonication and underscore the need for protocol-specific thresholds.