Aspirations Dashed: Conventional Synovial Fluid Analysis Is Superior to Synovial Fluid and Blood Neutrophil-to-Lymphocyte Ratios in Diagnosing Pediatric Septic Arthritis of the Hip and Knee

Abstract

Introduction: Recent research suggests that synovial fluid neutrophil-to-lymphocyte ratio (SF-NLR) is a superior diagnostic for pyogenic septic arthritis (SA) in adults compared with synovial fluid white blood cell count (SF-WBC) ≥50,000 cells/μL or ≥90% polymorphonuclear leukocytes (SF-%PMN). Other research also indicates that the neutrophil-to-lymphocyte ratio in the blood (B-NLR) may be of diagnostic significance. However, it is not known whether these findings extend to the pediatric population. Methods: Medical records at a large urban tertiary-care children’s hospital were queried for emergency department visits between 2012 and 2023, where synovial fluid (SF) analysis was performed to evaluate for SA of the hip or knee. Patients 18 years old and above were excluded. The “conventional composite test” (CCT) for SA was considered positive if SF analysis showed any of the following: (1) SF-WBC ≥50,000 cells/μL, (2) ≥90% PMNs, or (3) organisms reported on gram stain. Patients with aspirate and/or operating room (OR) cultures (or supplemental testing, ie, nucleic acid identification) revealing an offending organism were considered to have culture-positive septic arthritis (CPSA). The remaining patients were considered culture-negative (CN). Serum and SF test data were analyzed to assess their diagnostic utility in identifying CPSA. Receiver operating characteristic (ROC) curves were examined to compare the predictive value of SF-NLR and B-NLR versus conventional indicators of SA. Results: A total of 394 patients met the inclusion criteria. In all, 58.6% (n=231) were male, 67.5% (n=266) involved the knee, and 20.1% (n=79) had CPSA. Those with CPSA had higher ESR and CRP compared with CN patients (P<0.01). Bivariate testing did not show a difference in SF-NLR or B-NLR between those with CPSA and CN patients (P=0.93 and 0.37, respectively). The CCT showed 91% sensitivity and 35% specificity using conventional thresholds. ROC analysis showed that SF-WBC was superior to SF-NLR and B-NLR in the diagnosis of CPSA (AUC=0.71 vs. 0.50 and 0.53, respectively; both P<0.01). Among CCT (+) patients who ended up culture negative, Lyme testing was positive in 48.8% (100/205). Conclusion: In contrast to adults, SF-NLR and B-NLR were not found to be strong diagnostic indicators of SA of the hip or knee in pediatric patients. This may be because competing diagnoses in children come with systemic inflammatory responses similar to that seen in pyogenic SA, while noninfectious conditions that might represent the major alternate diagnoses in adults do not increase systemic inflammatory markers as significantly. Given the high incidence of Lyme disease seen among patients in this study, this topic should be further studied at pediatric centers outside Lyme-endemic areas to better understand the generalizability of these findings. Significance: Despite excitement regarding SF-NLR and B-NLR as diagnostics for adult SA, these criteria appear less useful in the diagnosis of pyogenic SA in pediatric patients in Lyme-endemic areas.