[Bleeding Risk Score in Patients with Crimean-Congo Hemorrhagic Fever]. / Kirim Kongo Kanamali Atesi Hastalarinda Kanama Risk Skoru.

Bleeding is considered to be an indicator of poor prognosis in Crimean-Congo hemorrhagic fever (CCHF) disease. In the prehemorrhagic period, clinical signs are usually non-specific. The hemorrhagic period usually begins 3 to 5 days after the onset of the disease. The aim of this study was to build a risk score to predict bleeding status in CCHF patients with clinical and laboratory findings. This methodological study was carried out in one of the largest centers which is located in the east part of Turkey and CCHF-endemic region between April 2014-October 2019 with 450 CCHF patients' data. Risk score was created with univariate and multivariate logistic regression analyzes with the data of 80% of the patients, and the diagnostic power of the created score was determined by ROC analysis. The data of the remaining 20% were used as the verification data set and the created score was tested by ROC analysis. The patients had a mean age of 47.83 ± 17.46 years (median 48; min-max: 16-90 years) and 209 (59.7%) were male. Hemorrhage was detected in 93 patients (26.6%). Of the hemorrhagic patients, 83 (23.7%) had multiple hemorrhage sites. In univariate analyzes, time between the onset of symptom and admission to the hospital (≥4 days), aspartate aminotransferase (≥228 U/L), alanine aminotransferase (≥143.5 U/L), lactate dehydrogenase (≥641 U/L), creatine kinase (≥227 U/L), white blood cell count (≤1810 × 106/L), platelet count (≤38385 × 106/L), activated partial thromboplastin time (≥38.5 s) and fibrinogen value (≤227 mg/dl) were found to be an independent risk factor for bleeding. As a result of multivariate analysis, the time between the onset of symptoms and admission to the hospital, white blood cell count, platelet count, aspartate aminotransferase, creatine kinase and fibrinogen values were included in the risk scoring. The area under the curve of the generated score is 0.875; sensitivity was 80.6% and specificity was 80.5%. Platelet count responsible for hemostasis, affected in terms of number and function in CCHF disease, LDH, AST and aPTT values used as indicators of liver functions that are the target of the virus can be used with high diagnostic prediction for bleeding. However, the predictive power of the generated score on bleeding is higher than the effect of each variable alone. In addition, it can be easily calculated during patient follow-up and can guide the treatment process.