The aim of this study was to identify laboratory predictors and develop a regression model to early detect a focal encephalitic form (EF) of tick-borne encephalitis (TBE). The study was involved 33 patients diagnosed with febrile form (FF) of acute TBE, 54 patients with meningeal form (MF), 14 patients with EF with good outcome and 6 patients with EF with fatal outcome (EFFO). The results of medical laboratory tests obtained in a day of hospitalization were analyzed. It was found that the significant predictive power for diagnosis of ЕF and EFFO demonstrated the following parameters: leukocytes levels above 11.9×106/L, segmented neutrophils levels in the blood above 73%, lymphocytes level in the blood below 10%, index of erythrocyte sedimentation rate below 20 mm/h, the blood glucose concentration above 6.5 mmol/L, the total bilirubin level above 10 μmol/L, the activity of aspartate aminotransferase above 25 IU/L, alanine aminotransferase above 24 IU/L, pleocytosis below 100 cells/μL, lymphocytes level in the cerebrospinal fluid below 49 cells/μL and hypochlorarchy is less than 118 mmol/L. We developed a model using logistic regression analysis that showed the significant predictive power for diagnosis of EF of TBE (AUC = 0.95). Thus, we demonstrated the possibility of the risk prediction of EF in patients with TBE using the regression model and diagnostic scoring system on the base of medical laboratory tests.
Keywords: tick-borne encephalitis, predictors, logistic regression, cerebrospinal fluid, blood, leukocytes, lymphocytes, glucose, aminotransferases.