CLINICAL PREDICTIONAL CRITERIA FOR ENCEPHALITIC FORM OF ACUTE TICK-BORNE ENCEPHALITIS

Abstract

Background. The aims of this study were to identify the main clinical predictors of mono(multi)focal encephalitis form (EF) of tick-borne encephalitis (TBE), develop binary logistic regression models, standardized questionnaire, and scoring system to differentiate between EF and meningeal form (MF) of this disease. Materials and methods. A retrospective clinical case-control study was carried out to analyze routine data of the case history files of 138 MF patients and 26 patients with EF of TBE. All the data obtained were recorded in a computer database that was analyzed statistically using Epi Info software, version 3.3.2 (USA) and Scorecard STATISTICA 12.0 (StatSoft, USA) applications for development of diagnostic models. To assess the predictive power of the scorecard model, the ROC (receiver operating characteristic) curve analysis was carried out. Results. The study enabled to identify the most significant predictors of physical and neurological examination for the early diagnosis of EF of TBE. These predictors included the following characteristics: the patient’s age was more than 41 years, pulse rate was more than 80 per min, systolic arterial pressure was over 130 mm Hg, dysfunction of cranial nerves, speech, consciousness, coordination of movements, and the disappearance of abdominal reflexes. Moreover, diabetes, hypertension, ischemic heart disease, and obesity were the additional predictors that were able to significantly increase the risk of a lethal outcome in the EF patients. Conclusion. As the results of this study three models were built and diagnostic scoring systems were developed to improve the management of the patients due to early prognosis for FEF of TBE.

Keywords

• клещевой энцефалит
• очаговая форма
• предикторы
• модель логистической регрессии
• ROC-анализ