# Diagnostic Utility of Synovial Cell Count Prior to Revision Compared to Re-Revision Arthroplasty

## TL;DR

This study examines how synovial cell count thresholds for diagnosing joint infections remain reliable even in patients with prior revision surgeries.

## Contribution

The study identifies that synovial cell count diagnostic accuracy remains high in re-revision arthroplasty cases, with minor threshold variations.

## Key findings

- No significant differences in cell counts were found between first revision and re-revision cases for septic and aseptic indications.
- The optimal synovial cell count cutoff for diagnosing infection was 2439.50 G/L with high sensitivity and specificity.
- The type of previous surgery in re-revisions significantly affects diagnostic accuracy.

## Abstract

Background/Objectives: The aim of this study is to investigate how the joint, the number and the type of prior revision surgeries influence the diagnostic thresholds for synovial cell count for patients who undergo their first total hip or knee arthroplasty revision compared to re-revisions, as different cutoffs might substantially influence treatment courses. Methods: In this retrospective single-center register analysis, data from 214 revised THAs (total hip arthroplasties) and TKAs (total knee arthroplasties) were collected, of which 103 (48.1%) have so far undergone at least one revision surgery. Diagnosis was based on the EBJIS criteria, and we identified 163 (76.2%) septic and 51 (23.8%) aseptic cases. Data on synovial cell count were collected and analyzed for their diagnostic accuracy and optimal cutoffs. For re-revisions, a covariate-adjusted ROC (receiver operating characteristic) for the joint, type of previous surgery and number of surgeries was created. Results: We found no significant differences in cell counts between patients before first revision compared to those undergoing re-revision for septic (p = 0.40) and aseptic indications (p = 0.84). The overall diagnostic accuracy was high for all re-revision cases, with a sensitivity of 0.86, specificity of 0.91, AUC (area under the curve) of 0.92, at an optimal cutoff value of 2439.50 G/L. As for re-revised hip joints, the optimal cutoffs were higher compared to knee joints (2439.5 G/L vs. 2626.5 G/L, hip AUC = 0.90, knee AUC = 0.93, p = 0.14). Furthermore, the AUCs for cell count differed significantly depending on the type of previous surgery in re-revision (p = 0.03). The covariate-adjusted analysis showed no significant differences compared to the unadjusted analysis. Conclusions: Cell count remains reliable for diagnosing periprosthetic joint infection in patients with prior revisions, with minor threshold variations from the EBJIS (European Bone and Joint Infection Society) criteria. While the type of preceding revision affects accuracy, the diagnostic value remains consistently high overall.