Affiliation:
1. Kutahya Health Sciences University
Abstract
Abstract
Objectives
In this research, it was aimed to evaluate the predictive value of blood parameters and ratios on knee prosthesis loosening in patients after knee arthroplasty.
Methods
A total of 221 patient files having knee prosthesis attempting to our clinic between 2018 and 2022 were retrospectively examined. Patients were divided into two groups as loosening (n = 108; 48.9%) and not loosening (n = 108; 48.9%) groups. Monocyte/HDL ratio (MHR), Albumin/CRP ratio (ACR), Eosinophile/lymphocyte ratio (ELN), MPV/PLT ratio (MPR), Monocyte/lymphocyte ratio (MLR), Platelet/lymphocyte ratio (PLR) and Hemoglobin/Platelet ratio (HPL) parameters of patients were calculated.
Results
HDL, albumin, hemoglobin, ACR, PLR and HPL levels were significantly higher in loosening patients (p < 0.05). CRP, neutrophile, lymphocyte, monocyte and MHR means were significantly higher in no-loosening patients (p < 0.05). Loosening was significantly correlated with gender (r = 0.141; p < 0.05), HDL (r = 0.474; p < 0.01), albumin (r = 0.349; p < 0.01), CRP (r=-0.466; p < 0.01), neutrophile (r=-0.167; p < 0.05), lymphocyte (r=-0.240; p < 0.01), monocyte (r=-0.380; p < 0.01), hemoglobin (r = 0.254; p < 0.01), MHR (r=-0.513; p < 0.01), ACR (r = 0.496; p < 0.01), PLR (r = 0.142; p < 0.05) and HPL (r = 0.204; p < 0.01). Binary logistic regression analysis results showed that MHR (B=-0,415; p < 0.01), ACR (B = 0,048; p < 0.01) and HPL (r = 30,640; p < 0.05) had significant predictive value for knee prosthesis at multivariate level. predictive value of MHR (AUC = 0.798; p < 0.01), ACR (AUC = 0.786; p < 0.01) and HPL (AU = 0.618; p < 0.01) were statistically significant. For MHR 6.2481 cutoff level, sensitivity was 72.2% and specificity was 10.6%. For ACR 8.6380 utoff level, sensitivity was 84.3% and specificity was 59.3%. For HPL 0.0506 cutoff level, sensitivity was 63.9% and specificity was 54.0%.
Conclusion
MHR, ACR and HPL may be used predictors of knee prosthesis after knee surgery.
Publisher
Research Square Platform LLC