Esin Firuzan

Risk Factors for Early Bacterial Infections in Liver Transplantation

OBJECTIVE Our aim was to determine perioperative risk factors for early bacterial infection after liver transplantation. METHODS Retrospectively examining medical records using Centers for Disease Control and Prevention (CDC) definitions to identify nosocomial infections, we analyzed data on 367 adult patients. RESULTS The incidence of infection was 37.3% (n = 137): namely, surgical site (n = 74; 20.2%) [corrected], blood stream (n = 64; 17.4%), pulmonary (n = 49; 13.4%), urinary system (n = 26; 7.1%). Significant risk factors within the first 30 days were as follows: deceased donor, Model for End-Stage Liver Disease (MELD) >20, albumin level <2.8 g/dL, intraoperative erythrocyte transfusion >6 U, intraoperative fresh frozen plasma >12 U, bilioenteric anastomosis, postoperative intensive care unit stay >6 days, and postoperative length of stay >21 days. Significant risk factors detected within the first 90 days were as follows: MELD >20, preoperative length of stay >7 days, reoperation, postoperative length of intensive care unit stay >6 days, and postoperative length of stay >21 days. Variability was observed in risk factors according to localization of infection. As a result, except for MELD, type of donor, and biliary anastomosis, the others are preventable factors for early bacterial infection. In addition, the same risk factors showed variability according to the site of infection.

New Block Bootstrap Methods: Sufficient and/or Ordered

ABSTRACT In this study, we propose sufficient time series bootstrap methods that achieve better results than conventional non-overlapping block bootstrap, but with less computing time and lower standard errors of estimation. Also, we propose using a new technique using ordered bootstrapped blocks, to better preserve the dependency structure of the original data. The performance of the proposed methods are compared in a simulation study for MA(2) and AR(2) processes and in an example. The results show that our methods are good competitors that often exhibit improved performance over the conventional block methods.

COMPARISON OF SINGLE AND MODIFIED EXPONENTIAL SMOOTHING METHODS IN THE PRESENCE OF A STRUCTURAL BREAK

Zaman serisi analizlerinde serilerin modellenmesinin, aralarindaki iliskilerin incelenmesinin yaninda temel amac gelecege yonelik ongorumleme yapmaktir. Literaturde en yaygin kullanilan yontemlerden biri ussel duzeltme yontemleridir. Serilerinin veri uretim sureclerinde, finansal krizler, dogal afetler gibi bircok nedenden dolayi kalici yapisal degisimler meydana gelebilmektedir. Bu degisimler model parametrelerini degistirebildigi gibi analiz sonuclarina da etki etmektedirler. Bu calismadaki temel amac, seride yapisal kirilmalar oldugunda basit ussel duzeltme (SES) ile yeni gelistirilmis olan Modifiye Ussel Duzeltme (MSES)(2016) yontemlerinin tahminleme performanslarini karsilastirmaktir. Hata teriminin (MAE) ortalama ve varyansi orneklemin buyuklugunden, kirilmanin siddetinden ve konumundan etkilenmektedir. Veri setindeki kirilmalar model tahmini olumsuz etkilemektedir. MSES yonteminin kullanilmasinda olasi kirilmalarin buyuklukleri ve konumlari dikkate alinmalidir.