Search

Statinler hakkında bir sunum.

Background: The aim of this study was to determine mortality risk by calculating Score for Neonatal Acute Physiology and Perinatal Extension II (SNAP-PE-II) and Clinical Risk Index for Babies (CRIB) score, and evaluate prediction of the effects of antenatal corticosteroid and surfactant treatment on mortality. Methods: This multicenter study was conducted simultaneously in five different centers in four different provinces in Southern Turkey between July 2012 and July 2013. A total of 1668 inborn subjects hospitalized in the neonatal intensive care unit within the first 12 h of delivery, and meeting the selection criteria, were included in the study, and CRIB and SNAP-PE-II were used to determine mortality. Results: The SNAP-PE-II scoring system was applied to all patients, and the CRIB scoring system was used for 310 newborns with gestational age

Purpose: Spinal anesthesia for caesarean delivery is often associated with hypotension. This study was conducted to evaluate the effects of rapid crystalloid (Lactated Ringer’s solution; LRS) or colloid (hydroxyethyl starch; HES) cohydration with a second intravenous access line on the incidence of hypotension and ephedrine requirement during spinal anesthesia for cesarean section. Methods: We studied 90 women with uncomplicated pregnancies undergoing elective cesarean section under spinal anesthesia. Intravenous access was established in all patients with two peripheral intravenous lines, the first being used for the baseline volume infusion. Immediately after induction of spinal anesthesia, LRS (Group L) or HES (Group C) infusions were started at the maximal possible rate via the second line in groups L and C respectively. In the third group (Group E) , patients received lactated Ringer’s solution at a ‘keep vein open’ rate to maintain the doubleblind nature. The incidence of hypotension, ephedrine requirements, total amount of volume and side effects were recorded. Results: The incidence of hypotension was significantly greater in group E than in groups L and C, and greater in group L than in group C (p

SUMMARY: Kılıç M, Altıntaş DU, Yılmaz M, Bingöl-Karakoç G, Burgut R, Güneşer-Kendirli S. Evaluation of efficacy of immunotherapy in children with asthma monosensitized to Alternaria. Turk J Pediatr 2011; 53: 285-294. In this study, we aimed to evaluate the efficacy of specific immunotherapy (SIT) in children monosensitized to Alternaria. Sixteen children with bronchial asthma monosensitized to Alternaria were enrolled in the study. Patients were divided into two groups as the immunotherapy group (Group I; 9 patients) and control group (Group II; 7 patients). A significant reduction in bronchial responsiveness to methacholine and Alternaria was found in Group I after one year of SIT (p=0.03, p=0.006) in comparison to controls. Specific IgE levels were decreased in the immunotherapy group (p=0.001). Following allergen provocation, a rise in sputum eosinophil count was found to be lower in the SIT group compared to controls after one year (p=0.011), and sputum eosinophil cationic protein (ECP) levels did not change in the SIT group, while there was a statistically significant increase in controls. Our results demonstrated that SIT with Alternaria caused clear changes in airway responsiveness and serum-specific IgE levels. However, further longterm studies in large series should be carried out for clinical documentation of the efficacy of SIT in the treatment of children with Alternaria allergy.

BACKGROUND The main objective of this study was to explore the patterns of road traffic crashes (RTCs) among drivers in Qatar and to examine the contributing factors involved. METHODS This study was a cross-sectional survey conducted from February to June 2009 at the Primary Health Care Centers in the State of Qatar. A random sample of 1675 Qatari drivers were approached, and 1228 of them responded and agreed to participate in this study, with a response rate of 74.6%. Face to face interview was conducted by well-trained research assistants based on a questionnaire covering sociodemographic information, driving history, type of vehicle, driver behavior, details of crashes, and accident pattern. RESULTS The study revealed that 26.6% of the studied Qatari drivers were involved in RTCs. 69.4% of the drivers with a history of RTCs were male. Drivers in the age group 25-34 years showed a higher risk of having a crash (31.2%). Drivers of cars (45.6%) and 4WD/SUVs (23.5%) were more frequently involved in RTCs. 23.5% of drivers who had RTC did not always wear their seat belt while driving. 37.9% of the drivers with RTC were engaged in eating and drinking and 41.9% were using their mobile phones while driving. More than half of the drivers with RTCs had traffic violations (57.2%), with 25.7% exceeding the speed limit. A highly significant difference was observed between the two groups (drivers with and without RTC) in terms of presence of traffic violation (, AMAÇ Bu çalışmanın başlıca amacı, Katar’daki sürücüler arasındaki karayolu trafik kazalarının (KTK) paternini araştırmak ve KTK ile ilişkili faktörleri incelemektir. GEREÇ VE YÖNTEM Bu çalışma, Şubat-Haziran 2009 tarihleri arasında Katar Devleti Sağlık Merkezlerinde yürütülen kesitsel çalışmadır. Rastgele yöntemle seçilen 1675 Katarlı sürücüden oluşan grupla çalışıldı. Sürücülerin 1228’i %74,6’lık bir yanıt oranıyla soruları yanıtlayıp bu çalışmaya katılmayı kabul etti. İyi eğitimli araştırma asistanları tarafından yüz yüze görüşülerek; sosyodemografik bilgi, sürücülük öyküsü, araç tipi, sürücü tutumu, kazaların ve kaza şeklinin ayrıntılarını kapsayan bir anket yapıldı. BULGULAR Bulgular şöyleydi: Çalışma, Katarlı sürücülerin %26,6’sının KTK’ya karıştığını gösterdi. KTK öyküsü bulunan sürücülerin %69,4’ü erkek idi. 25-34 yaş grubundaki sürücüler, daha yüksek kaza riskini (%31,2) gösterdi. Arazi taşıtı ve spor amaçlı taşıt özelliğe sahip araçlarla (%23,5) sürücüleri (%45,6), KTK’dan daha fazla etkilendi. Kaza geçiren sürücülerin %23,5’i, araç sürerken her zaman emniyet kemeri kullanmıyodu. Sürücülerin yarısından daha azı, araç sürerken yeme ve içme (%37,9) ve cep telefonu kullanmakta (%41,9) idi. Sürücülerin yarısından daha fazlası, trafik kuralı ihlali (%57,2) ve hız sınırını aşma (%25,7) ile trafik kazasına karıştı. Her iki grup (trafik kazası geçiren ve geçirmeyen sürücüler) arasında trafik kuralı ihlali (

Aim: To determine the frequency of nosocomial infections (NIs) and the antimicrobial resistance of pathogens isolated from patients admitted to 4 intensive care units (ICUs) at a university hospital. Materials and Methods: The patients were prospectively followed over 2 years in medical, surgical, and reanimation ICUs. The isolates were collected from patients with NI determined by the Center for Diseases Control and Prevention criteria. The identification was performed and susceptibility to antibiotics was determined in an automated system Sceptor (Becton Dickinson) as described by the CCLS. Results: Among the 3962 patients in the ICUs, total NI incidence was 6.9%, with the highest rate in the reanimation ICU (14.7%) and the lowest rate in the surgical ICU (2.5%). In all, 492 NI episodes were diagnosed in 272 patients. The incidence of NIs was 13.1 per 1000 patient days. The NIs were urinary tract infections (32%), primary blood stream infections (24%), pneumonia (20%), surgical-site infections (13%), and other infections (11%). Pneumonia was the most common NI in the reanimation ICU (54%). The pathogens were 39% gram-positive bacteria, 52% gram-negative bacteria, and 9% Candida species. S. aureus (18%) was the most frequently isolated bacteria, followed by P. aeruginosa (16%), Acinetobacter spp. (10%), coagulase negative staphylococcus (CNS) (10%), Klebsiella (9%), E. coli (9%), and Enterococcus spp. (9%). Resistance to methicillin was 90% in S. aureus isolates and was 95% in the CNS isolates; no resistance was detected to vancomycin in S. aureus and CNS isolates. Eight percent of Enterococcus isolates were resistant to vancomycin. Against the gramnegative bacteria the carbapenems were most active, followed by amikacin and ciprofloxacin. Five percent of P. aeruginosa and 15% of Acinetobacter spp. were resistant to all antibiotics. The most active antimicrobial agents against Acinetobacter spp. were imipenem (73%) and tobramycin (46%); against Pseudomonas spp. were ticarcillin/clavulanate (56%), piperacillin (44%), and imipenem (33%); against Klebsiella spp. was imipenem (92%); against E. coli were imipenem (94%), amikacin (94%), and ciprofloxacin (74%); and against Enterobacter spp. were ciprofloxacin (89%) and imipenem (78%). Conclusions: The high rates of resistant pathogens responsible for NI in ICUs suggest that infection preventive procedures should be implemented and antimicrobial agents must be used appropriately., Amaç: Bir üniversite hastanesinde dört yoğun bakım ünitesinde yatan hastalarda nozokomiyal infeksiyonların sıklığı ile izole edilen patojenlerde antimikrobiyallere direnci belirleme. Yöntem ve Gereç: Medikal, cerrahi ve reanimasyon yoğun bakım ünitelerinde yatan hastalar prospektif olarak izlendi. İzolatlar CDC kritelerine göre belirlenen nozokomiyal infeksiyonlardan toplandı. İzolatların tanımlanması ve antibiyotiklere duyarlılığı Sceptor (Becton Dickinson) otomatize sistemle CCLS kriterlerine göre yapıldı. Bulgular: YBÜ’lerinde yatan 3962 hastada, nozokomiyal infeksiyon insidansı % 6,9; en yüksek oran reanimasyon YBÜ’nde (% 14,7), en düşük oran cerrahi YBÜ’nde (% 2,5) olarak belirlendi. 272 hastada 492 Nİ epizodu belirlendi. Nİ insidansı 1000 hasta gününe 13,1 olarak saptandı. Nİ oranları, üriner sistem (% 32), primer bakteriyemi (% 24), pnömoni (% 20), cerrahi alan (% 13) ve diğer infeksiyonlar % 11 bulundu. Pnömoni Reanimasyon YBÜ’nde en sık saptanan (% 54) Nİ idi. Etkenlerin % 39’u gram pozitif, % 52’si gram negatif bakteri, % 9’u Candida türleri idi. En sık izole edilen bakteri S. aureus (% 18); bunu P. aeruginosa (% 16), Acinetobacter spp. (% 10), koagülaz negatif stafilokok (KNS) (% 10), Klebsiella (% 9), E. coli (% 9), ve Enterococcus spp. (% 9) izlemekteydi. S. aureus türlerinde metisilin direnci % 90, KNS larda % 95, S. aureus ve KNS larda vankomisine direnç saptanmadı. Enterokokların % 8’i vankomisine dirençli idi. Gram negatiflere en etkili antibiyotik karbapenemi, amikasin ve siprofloksasin izlemekteydi. P. aeruginosa suşlarının % 5’i ile Acinetobacter türlerinin % 15’i tüm antibiyotiklere dirençli bulundu. Acinetobacter türlerine en etkin antibiyotikler imipenem (% 73), tobramisin (% 46); Pseudomonas türlerine tikarsilin/klavulanat (% 56), piperasilin (% 44), imipenem (% 33); Klebsiella türlerine imipenem (% 92), E. coli suşlarına imipenem (% 94), amikasin (% 94), siprofloksasin (% 74); Enterobacter türlerine siprofloksasin (% 89), imipenem (% 78) idi.

Computerized tomography and magnetic resonance imaging allow the accurate diagnosis in stroke and distinction of ischemic from hemorrhagic lesions. However, clinical diagnosis is still critical where neuroimaging techniques are not available, especially to establish first‐aid measures in a stroke patient. In this prospective study of 300 patients with stroke, the diagnosis of ischemic and hemorrhagic strokes was made as an informal bedside diagnosis prior to neuroimaging. The accuracy of bedside diagnosis was also compared with the results of the Allen and Siriraj stroke scores. Then, the reliability of these scores and our informal bedside diagnosis, along with the final diagnosis, were compared with each other. Our informal bedside diagnosis was correct in 250 of the 300 patients (83.3%). The diagnostic sensitivity and positive predictive value (PPV) of bedside diagnosis for ischemic stroke were 87.8% and 86.0% and for intracerebral hemorrhage 75.7% and 78.5%, respectively. Sensitivity and PPV for every cut‐off value of the Allen and Siriraj stroke scores were less than that of sensitivity and PPV of informal bedside diagnosis. When the receiver operating curves obtained from the Allen and Siriraj stroke scores were compared, the Allen stroke score was found to be a better predictor in the final diagnosis. Our data suggest that an informal bedside diagnoses is as good as diagnoses made on certain intracerebral hemorrhages and on certain ischemic strokes by the Siriraj and Allen stroke scores; when compared, the Allen score seems to be better than the Siriraj stroke score. Hence, the use of both scores is recommended in epidemiologic studies as a screening tool as suggested by previous studies employing other diagnostic tools in clinics.

A study was performed on 176 patients with benign cervical lesions. Human papillomavirus (HPV) structural antigens which were stained with the immunoperoxidase staining were sought for in cervical smears and in cervical tissues excised by the loop electrosurgical excision procedure (LEEP). HPV infection was found in 16.4% of the patients. Furthermore, cytologic screening on Papanicolaou (PAP) smears, colposcopy, and histologic examinations of whole transformation zones excised by the LEEP were performed. The diagnosis of HPV was based on the positive immunoperoxidase staining of either the exfoliated cells or the cervical tissue. HPV was detected on 5.6% of the PAP smears. The reliability of cytologic colposcopic, and histologic diagnosis of HPV was investigated.

We studied the occurrence of anti-hepatitis C (HCV) antibodies in patients with malignant disease (53), in patients undergoing haemodialysis (56), and in blood donors (204) as healthy population controls. The study was carried out using the second-generation EIA test. Anti-HCV positivity was 23.2% in haemodialysis (HD) patients, 0.5% in blood donors, and 0% of the patients with malignant disease (MD). There was no association between anti-HCV positivity and the results of AST, ALT and HBsAg tests in patients and controls. But there was significant association of blood transfusion frequency and duration of HD with anti-HCV positivity in patients undergoing HD, and conversely an absence of this association in patients with MD. However, two of the anti-HCV-positive HD patients did not have any blood transfusion history. One HCV-positive blood donor had a history of surgical operation. Nosocomial transmission of HCV infection has replaced blood transfusion as the main risk factor in HD patients, and preventive measures should be performed in this direction to control infection.

Fifty-four early stage gastric cancer patients were investigated for the presence of AFP and CEA. In their tumor specimens 9 patients showed AFP, 14 patients showed CEA and 9 patients showed AFP and CEA positivity by immunoperoxidase staining. Mean age was younger in AFP (+) patients (44.35 ± 8.4) than others (53.5 ± 8.0, p=0.0006). There was a significant correlation between the AFP and CEA positivity and stage of the disease (p=0.001, p=0.001). Lymphatic and/or venous invasion were more prominent in tumors with positive AFP staining (p=0.04) but serosal invasion was more frequent in CEA (+) tumors (p=0.05). The incidence of poorly differentiated tumors was high in AFP (+) tumors (p=0.034). Disease free survival (DFS) and overall survival (OS) were significantly shorter in patients with high serum AFP and CEA levels (p=0.0001). As conclusion AFP and CEA content of the tumor is a very important parameter in the management and prognosis of gastric cancer patients.

Objective- To assess smoking- related be- haviour, beliefs, and attitudes of students at different medical schools in Turkey, in comparison to those of students in nonmedical schools.

The prevalence of IgG antibodies to C. trachomatis was determined in 185 infertile women who underwent laparoscopy, and 110 pregnant women with no known infertility problems. In addition, chlamydial antigens were evaluated by EIA in cervical samples taken from all subjects. Subjects with tubal infertility had the highest prevalence of chlamydial antigen and antibody (P less than 0.01). While the percentage of subjects with antigen and antibody positive was 11.6 per cent, those with antigen negative and antibody positive averaged 55.8 per cent among women with tubal related infertility. The results of our study provide additional support to the concept that infertility of tubal etiology often is a sequela of a previous chlamydial infection.

Background: In recent years the use of surrogate end points ( S ) has become an interesting issue. In clinical trials, it is important to get treatment outcomes as early as possible. For this reason there is a need for surrogate endpoints ( S ) which are measured earlier than the true endpoint ( T ). However, before a surrogate endpoint can be used it must be validated. For a candidate surrogate endpoint, for example time to recurrence, the validation result may change dramatically between clinical trials. The aim of this study is to show how the validation criterion ( R 2 trial ) proposed by Buyse et al. are influenced by the magnitude of treatment effect with an application using real data. Methods: The criterion R 2 trial proposed by Buyse et al. (2000) is applied to the four data sets from colon cancer clinical trials (C-01, C-02, C-03 and C-04). Each clinical trial is analyzed separately for treatment effect on survival (true endpoint) and recurrence free survival (surrogate endpoint) and this analysis is done also for each center in each trial. Results are used for standard validation analysis. The centers were grouped by the Wald statistic in 3 equal groups. Results: Validation criteria R 2 trial were 0.641 95% CI (0.432–0.782), 0.223 95% CI (0.008–0.503), 0.761 95% CI (0.550–0.872) and 0.560 95% CI (0.404–0.687) for C-01, C-02, C-03 and C-04 respectively. The R 2 trial criteria changed by the Wald statistics observed for the centers used in the validation process. Higher the Wald statistic groups are higher the R 2 trial values observed. Conclusion: The recurrence free survival is not a good surrogate for overall survival in clinical trials with non significant treatment effects and moderate for significant treatment effects. This shows that the level of significance of treatment effect should be taken into account in validation process of surrogate endpoints.

TEZ12152 Tez (Yüksek Lisans) -- Çukurova Üniversitesi, Adana, 2016. Kaynakça (s. 65-69) var. XI, 77 s. :_res. (bzs. rnk.), tablo ;_29 cm. Tanı testleri, hasta ve sağlıklı bireylerin oluşturduğu heterojen bir kitlede bireyin gerçek durumunu ortaya çıkarmak amacıyla kullanılır. Doğruluğu kesin olarak kanıtlanmış altın standart testler ile bireylere "hasta" ya da "sağlıklı" tanısı konulmaktadır. Bazı altın standart testlerdeki uygulama zorlukları, test maliyetlerinin yüksek olması ve çeşitli hastalıklarda testin girişimsel olması nedeniyle her şüpheli durumda kullanımı mümkün olmayabilmektedir. Bu nedenle sağlık alanında altın standart testlere alternatif olabilecek tanı testleri geliştirilmeye çalışılır. Altın standart test ile gerçek durumu belirlenen bireylere alternatif yeni test uygulanarak, bu yeni tanı testinin sınıflama gücünü gösteren doğruluk ölçütleri elde edilir. Altın standart varlığında ve yokluğunda bu doğruluk ölçütleri farklı yaklaşımlar kullanılarak elde edilmektedir. Bu yaklaşımlardan birisi de araştırmacılar tarafından değişik alanlarda yaygın olarak uygulanmaya başlanan Bayesci yaklaşımdır. Bayesci yaklaşım araştırmacıya ilgilendiği parametre tahminlerinin elde edilmesinde önsel bilgiyi (prior) kullanma olanağı sağlamaktadır. Son günlerde klinikteki yararı yadsınamayacak olasılık hesabı nomogram ve benzeri araçlarda kullanılmaktadır. Bu tezde, Bayesci yaklaşım ile tanı testlerine ait duyarlılık ve seçicilik gibi doğruluk ölçütleri elde edilmiş ve yaklaşımın katkısı araştırılmıştır. Doğruluk ölçütlerinin elde edilmesi aşamasında; Bayesci çıkarsama, Gibbs örneklemesi, Markov Zinciri Monte Carlo (MCMC) yöntemleri kullanılmıştır. Bu amaçla üniversitemizde yapılan Helicobacter Pylori ile ilgili çalışmadan elde edilen veriler WinBUGS ve/veya R paket programı yardımıyla analiz edilmiştir. Bayesci çıkarsama yapıldığında hastalık (veya durum) ile ilgili prevalans bilgisi dikkate alındığından tanı testleri daha güvenilir sonuçlar verebilmektedir. Diagnostic tests are used for defining the disease status of person in a heterogeneous population which consists of healthy and diseased people. With the help of the gold standard tests one can be diagnosed as “diseased” or “healthy”, however these tests may not be used in each uncertain situation due to the difficulty on practice, high costs and being interventional in some diseases. For this reason in many fields of medicine, diagnostic tests are developed to be used in place of gold standard tests. Measures of accuracy that show the power of classification of the newly developed alternative tests are obtained by using information provided by the gold standard test. In the absence/presence of a gold standard test these measures of accuracy can be determined using different approaches, one of which is Bayesian approach commonly used by the researchers in many different fields. The Bayesian approach also provides a way to include expert prior knowledge concerning parameters of interest. Nowadays the calculation of probability on clinical science is using on nomograms etc. and the contribution is undeniable. The aim of this thesis is to evaluate diagnostic tests by using Bayesian approach whether or not there is a gold standard, to assess the contribution of Bayesian approach, to analyze the datasets for practical application with WinBUGS and/or R package programme. Datasets will be analyzed with WinBUGS package programme. As statistical methods, Bayesian inference, Gibbs sampling and Markov Chain Monte Carlo will be employed. Diagnostic test for a disease give more reliable results when there is a Bayesian inference that including prevalence information. Bu çalışma Ç.Ü. Bilimsel Araştırma Projeleri Birimi tarafından desteklenmiştir. Proje No: TYL-2014-2716.

TEZ12176 Tez (Yüksek Lisans) -- Çukurova Üniversitesi, Adana, 2016. Kaynakça (s. 88-90) var. xii, 124 s. :_res. (bzs. rnk.), tablo ;_29 cm. Giriş: Risk tahmin modelleri, hastalık ya da ölüm gibi klinik sonuçların tahmin edilmesinde kullanılan ve risk faktörlerini içeren bir regresyon tekniğidir. Uygulama alanındaki gelişmelerle ve teknolojik ilerlemelerle beraber yeni risk faktörlerinin bulunması kaçınılmaz bir durum haline gelmiştir. Bulunan bu yeni risk faktörlerinin modele ilavesi ve performans artışının değerlendirilmesinde merkezi bir rol üstlenen alıcı işlem karakteristikleri eğrisi (ROC) bazı sınırlamalar içermektedir. Bu sınırlamaların başında, tahmini regresyon katsayıları arasındaki değişimin ölçülmesinde sonuçların oldukça tutucu özelliğe sahip olması gelmektedir. Bundan dolayı yeni yöntem arayışları başlamış ve 2008 yılında Pencina ve arkadaşları tarafından yeniden sınıflama tablolarına dayanan ve yeni faktörün risk modeline katkısını açıklayan iki metot geliştirilmiştir. Bu iki yöntem net yeniden sınıflama iyileştirmesi (NRI) ve entegre ayırım iyileştirmesi (IDI)’dir. Çalışmada, iç içe tasarıma sahip modellerin performansının değerlendirilmesine imkan sağlayan ?AUC, NRI ve IDI tekniklerinin birbirlerine göre avantaj ve dezavantajlarını araştırmak amaçlanmıştır. Gereç ve Yöntem: Çukurova Üniversitesi Tıp Fakültesi Balcalı Hastanesi’nin Yeni Doğan Yoğun Bakım Ünitesine, 2013-2015 Temmuz yılları arasında yatırılan 864 bebeğin ölüm olasılıklarını tahmin etmek için SNAP-II ve SNAPPE-II risk tahmin modelleri kullanılmıştır. Bebeğin sürfaktan ve annenin antenatal kortikosteroid (celestone) tedavisini alıp almaması yeni değişken olarak eklenmiştir. Simülasyon çalışmasında ise anlamsız değişken, uygun (fit) modele eklenmiş ve farklı örneklem büyüklüklerinde (100, 250, 500 ve 1000), farklı prevalans değerlerinde (0.10, 0.25 ve 0.50) 1000 tekrarlı simüle edilmiş verilerle yöntemlerin HP oranları belirlenmiştir. Bulgular: SNAP-II risk tahmin modeline hem celestone hem de sürfaktan bilgisinin eklenmesi ile elde edilen modele bağlı performans artışı anlamlı bulunmuştur. Bununla beraber SNAPPE-II risk tahmin modeline celestone bilgisinin eklenmesi ile elde edilen performans artışı sadece NRI sürekli istatistiği tarafından anlamlı bulunmuştur. Simülasyon çalışması ile NRI sürekli istatistiğinin HP oranı en yüksek iken ?AUC ve IDI istatistiğinin HP oranının en düşük olduğu belirlenmiştir. Bununla beraber NRI kategorik istatistiği düşük prevalanslarda yüksek HP oranına sahip iken prevalansın artışına bağlı olarak HP oranı azalmaktadır. Sonuç: ?AUC istatistiği oldukça tutucu sonuçlar vererek H0 hipotezini kabul etmeye en yatkın yaklaşım olarak belirlenirken NRI sürekli istatistiği tam tersi şeklinde davranmaktadır. IDI istatistiğinin HP oranındaki düşüklük ve anlamlı değişken senaryosunda performans artışını anlamlı bulması göz önüne alındığında kullanımı en uygun olabilecek yöntem olarak belirlenmiştir. Introduction: Risk prediction model is a regression method that is used for predicting clinical outcomes, such as disease or death, and including risk factors. Along with the developments in application fields and technological progress, discovery of new risk factors has become inevitable. Receiver Operating Characteristic (ROC) curve, that has a major role in including these new risk factors into the model and evaluating the performance of the model, has some limitations. The most important limitation of ROC is that the difference between predictive regression coefficients is highly conservative. To be used as a solution to this limitation, Pencina et al. (2008) proposed two methods which are based on reclassification tables and explaining the new factor’s contribution to the risk model. These two methods are called the Net Reclassification Improvement (NRI) and Integrated Discrimination Improvement (IDI). In this study, the aim is to investigate the performance and compare the advantages and disadvantages of ?AUC, NRI and IDI techniques that are nested models. Material and Method: In order to predict the death probabilities of 864 babies hospitalized in the Newborn Intensive Care Unit in Çukurova University Medical Faculty Balcalı Hospital between the years 2013 and 2015, SNAP-II and SNAPPE-II risk prediction models are used. Surfactant treatment for the baby and corticosteroid (celestone) treatment for the mother are added into the model as a new variable. In the simulation study, the insignificant variable is added to the fitted model and then several datasets are generated using different sample sizes (100, 250, 500 and 1000) and different prevalence values (0.10, 0.25 and 0.50) with 1000 replications. These datasets are used to calculate False Positive Ratios (FPR) of the ?AUC, NRI and IDI methods. Results: Improvement in the performance of the model obtained by including both celestone and surfactant variables into the SNAP-II risk prediction model is decided to be significant. Nevertheless, Improvement in the performance of the model obtained by including celestone into the SNAPPE-II risk prediction model is founded significant only by continuous NRI. The simulation study showed that the FPR of the continuous NRI is the highest while FPR of the ?AUC and IDI is the lowest. Moreover, it is concluded that categorical NRI has high FPR values at low prevalence levels and depending on the increase in prevalence values, FPR values tend to decrease. Conclusion: Whereas ?AUC statistic is decided to be the approach that has the most tendency to accept H0 by yielding very conservative results, the continuous NRI behaves quite the opposite. IDI statistic is decided to be the most appropriate method since its FPR values are lower than the other two methods and it concludes that the performance increase is significant in the real dataset.

Genetik Epidemiyolojide İstatistiksel Yaklaşımlar: Ailesel Kümelenmede Bir UygulamaGenetik ve çevrenin, hastalıkların sürecine katkısı keşfedildiğinden bu yana kanser hastalıkları da dahil olmak üzere, koroner arter hastalıkları, alerji ve psikiyatrik hastalıklar gibi bir çok hastalığın etiyolojisi daha karmaşık hale gelmiştir. Genetik ile ilgili araştırmalarda ilk basamak ailesel kümelenmenin ortaya konmasıdır. Bu bağlamda genetik epidemiyoloji, hastalık etiyolojisinde genetik faktörlerin oynadığı rolü araştırmada, klasik epidemiyolojik tasarım ve yöntemlerinin yerini alan, yeni bir çalışma alanıdır. Bu alanda istatistik yöntemler günümüzde giderek çeşitlenmektedir. Bu çalışma, hematolojik kanserli(Hem CA), solid kanserli (Solid CA) ve kanser olmayan(Kontrol) bireylerde ailesel kümelenmenin araştırılması için farklı istatistik yöntemlerin değerlendirilmesi amacıyla gerçekleştirilmiştir. Bu gruplarda ailesel kümelenmenin varlığını ortaya koyan 4 farklı istatistiksel yöntem sonuçları verilmiştir, bu yöntemler; İkili lojistik regresyon, Regressive lojistik regresyon, Genelleştirilmiş tahmin denklemleri (Generalized Estimation Equation-GEE) ve Tam çıkarsama (Exact Inference). Modellemede, SPSS 20.0 ve R paket programı kullanılmıştır. Bütün gruplar için alt kırılım, demografik bilgiler ve aile bilgileri karşılaştırılmıştır. Üç grubun birbiri ile karşılaştırılmasında cinsiyet, yaş, eğitim ve medeni durum anlamlı bulunmuştur(p

PubMedID: 17621481 The aim of this retrospective study was to determine the risk factors associated with intractability to therapy in childhood epilepsy. Fifty children with intractable epilepsy as evidenced by at least 1 epileptic fit per month were included in the study group, whereas the control group consisted of children who did not experience any recurrent seizure for at least 1 year at the time of the study. A ?2 test was used to evaluate the relationship between the test variables for the 2 groups, and the estimated relative risk (odds ratio) for each variable was calculated. The risk factors were subsequently determined by logistic multiple regression analysis. Univariate analysis showed that mental retardation, neurological abnormality, neuroradiological abnormality, perinatal anoxia, neonatal convulsion, presence of status epilepticus, and symptomatic etiology were significant risk factors for the development of refractory epilepsy (P < .05). For multivariate logistic regression analysis, age at seizure onset, status epilepticus, mixed type of seizures, and history of frequent seizures (more than once a month) were all found to be significant and independent risk factors for refractory epilepsy, and the number of drugs used in the study group was significantly higher than that in the control group (P < .05). In line with these findings, it was concluded that children who present with epilepsy and have these risk factors should be referred to a center where epileptic surgery is carried out without delay. © 2007 Sage Publications.

Aim: Non-small cell lung carcinoma is the leading cause of cancer related to death in the world. Squamous cell lung carcinoma (SqCLC) is the second most frequent histological subtype of lung carcinomas. Recently, growth factors, growth factor receptors, and signal transduction system-related gene amplifications and mutations are extensively under investigation to estimate the prognosis and to develop individualized therapies in SqCLC. In this study, besides the signal transduction molecule phosphatidyl inositol-3-phosphate kinase (IP3K) p110α, we explored the expressions of fibroblast growth factor 2 (FGF2) and receptor-1 (FGFR1) in tumor tissue and also their clinical and prognostic significance in patients with early/advanced SqCLC. Materials and methods: From 2005 to 2013, 129 patients (23 early, 106 advanced disease) with a histopathological SqCLC diagnosis were selected from the hospital files of Cukurova University Medical Faculty for this study. Two independent pathologists evaluated FGFR1, FGF2, and PI3K (p110α) expressions in both tumor and stromal tissues from 99 of the patients with sufficient tissue samples, using immunohistochemistry. Considering survival analysis separately for patients with both early and advanced stage diseases, the relationship between the clinical features of the patients and expressions were evaluated by univariate and multivariate analyses. Results: FGFR1 expression was found to be low in 59 (60%) patients and high in 40 (40%) patients. For FGF2; 12 (12%) patients had high, 87 (88%) patients had low expression and for IP3K; 31 (32%) patients had high and 66 (68%) patients had low expressions. In univariate analysis, overall survival (OS) was significantly associated with stage of the disease and the performance status of the patient (P

Klasik sağ kalım analizlerinde genellikle tek bir başarısızlık (ölüm, hastalığın, nüksetmesi vb.) nedeni araştırılmaktadır. Ancak çoğu kez, başarısızlığa birden fazla faktör etki etmekte ve bu faktörlerden birisi öne çıkarak başarısızlığa neden olursa, sağ kalım analizlerinin yarışan riskler dikkate alınarak yapılması gerekmektedir. Sağ kalım analizlerinde yarışan risk verileri mevcutken Kaplan-Meier (KM) yöntemi uygulandığında yarışan risklerin göz ardı edildiği görülmektedir. Yarışan risk varlığında Kaplan Meier yaklaşımını uygulamanın hatalı sonuçlar ve yorumlar sunacağını ortaya koymak amacıyla bu çalışmada yarışan riskler analizi (CR) ile KM yöntemlerinin sağ kalım olasılıklarının farklarının ortalama değerlerini hem gerçek veri setinde hem de simüle veri setinde incelenerek değerlendirilmiştir. Simülasyon üretim kısmında farklı dağılımlarda ve farklı örneklem büyüklüklerinde senaryo sonuçları sunulmuş ve bulgular değerlendirilmiştir. Farklı senaryolar oluşturulurken weibull dağılımından ve üstel dağılımdan yararlanılmış, örneklem büyüklüğü n=100, 150, 250, 500, 1000 alınmıştır. Sonuçlar Kümülatif ölüm olasılığı grafikleri de sunularak değerlendirilmiştir. Yarışan risklerin varlığında eş değişkenleri incelerken Grays test istatistiği uygulanmıştır. Yapılan hem gerçek veri seti üzerinde ki uygulama hem de simülasyon sonuçları yarışan risklerin varlığında genel sağ kalım analizleri kullanmak yerine yarışan riskler analizi uygulamanın doğru olacağını sunmuştur. Weibull dağılımından üretilen sonuçlarda iki yöntemin sağ kalım olasılıklarının farklarının ortalama sonuçları farklı çıkarken örneklem sayısı etkilememiştir. Üstel dağılımda bu sonuçlar sansürleme oranına bağlı olarak değiştiği ortaya konulmuştur. Sansürleme oranı arttıkça farkların azaldığı, örneklem sayısının büyümesi ile farkların arttığı gözlemlenmiştir Sonuçlar R yazılım programı kullanılarak yapılmış ve simülasyon çalışması için gerekli kodlar yazılmış simülasyonlar yapılmıştır.Anahtar sözcükler: R programı, Yarışan riskler, Üstel Dağılım, Kümülatif Ölüm Olasılığı, Grays modeli Evaluation of Competing Risks Based On Both Dependent , Independent Real And Simulated Data By Using Self Developed R ProgramClassical survival analysis methods commonly evaluate single cause of failure (die, relapse, etc.). However, failures may be of several distinct causes and the observation of any one of them prevents the observation in others. In this situation, survival analysis must be done by considering competing risks. In survival analysis when several competing risks exist, Kaplan Meier (KM) method ignores the competing risks (CR) aspect of the data. In order to show KM approach gives insufficient results and inferences when there is competing risks, we studied mean differences in survival probability which are both analyzed by CR and KM approaches separately in real data set and simulated data sets. During creation of simulated data, we used different sample sizes (n=100, 150, 250, 500, 1000.) in two different distributions (weibull and exponentials distributions) so that different scenarios are presented to give corresponding outcomes and inferences. Outcomes and graphs for Cumulative hazard function are presented. We used Grays test statistics to study covariate effects when competing risks exists. Both real and simulated data sets shows that when there is competing risks competing risks approach gives more appropriate results than classical survival analysis. When weibull distributions are used mean differences in survival probabilities of two methods are different but not effected by sample sizes. But when exponentials distributions are used mean differences in survival probabilities depend on censoring rate and sample sizes. When censoring rates gets higher we observed that mean differences in survival probabilities of two methods is decreased. As the sample sizes gets higher we see that mean differences in survival probabilities of two methods is increased.R statistical software is used to simulate and analyze data sets.Keywords: R program, Competing Risks, Exponential Distribution, Cumulative Hazard Function, Gray's models 105

Verification bias may occur when the test results of not all subjects are verified by using a gold standard. The correction for this bias can be made using different approaches depending on whether missing gold standard test results are random or not. Some of these approaches with binary test and gold standard results include the correction method by Begg and Greenes, lower and upper limits for diagnostic measurements by Zhou, logistic regression method, multiple imputation method, and neural networks. In this study, all these approaches are compared by employing a real and simulated data under different conditions. © 2013 © 2013 Taylor & Francis. Firat University Scientific Research Projects Management Unit This research was supported by the Scientific Research Project Unit of Cukurova University, Grant No. TF2007D10.

Agreement among raters is an important issue in medicine, as well as in education and psychology. The agreement among two raters on a nominal or ordinal rating scale has been investigated in many articles. The multi-rater case with normally distributed ratings has also been explored at length. However, there is a lack of research on multiple raters using an ordinal rating scale. In this simulation study, several methods were compared with analyze rater agreement. The special case that was focused on was the multi-rater case using a bounded ordinal rating scale. The proposed methods for agreement were compared within different settings. Three main ordinal data simulation settings were used (normal, skewed and shifted data). In addition, the proposed methods were applied to a real data set from dermatology. The simulation results showed that the Kendall's W and mean gamma highly overestimated the agreement in data sets with shifts in data. ICC4 for bounded data should be avoided in agreement studies with rating scales

Trichophyton tonsurans tinea gladiatorum is an emerging epidemic among combat-sport athletes across the globe. In the present study, we investigated the prevalence of symptomatic and asymptomatic dermatophytic infections among wrestlers in the National Greco-Roman Championship in Turkey. In total, 194 wrestlers from 32 provinces and 72 clubs were examined for scalp, trunk, groin, and toe web dermatophytic infections. We also administered a questionnaire to obtain information on the participants' lifestyles, wrestling characteristics, and risk factors for dermatophytic infections. The hairbrush method was used for scalp and trunk sampling, whereas a cotton swab was used for groin, toe web, and mat sampling. Three wrestling mats in the gymnasium were surveyed for dermatophytes using the touch preparation method. A total of 17 (8.8%) wrestlers harbored dermatophytes, and 22 strains were isolated: 13 (59.1%) T. tonsurans and 9 (40.9%) T. rubrum. These isolates were found on the scalp (8), trunk (2), forearm (1), hand (1), groin (3), and feet (7). In addition, we recovered 8 dermatophyte strains from the 150 mat samples (5.3%): T. rubrum in 6 samples (75%) and T. tonsurans in two samples (25%). T. tonsurans was only recovered from 11 out of 194 (5.7%) wrestlers. Scalp carriage represents the predominant (72.7%) clinical picture of a T. tonsurans infection in these Greco-Roman wrestlers in Turkey. © 2011 Springer Science+Business Media B.V. Acknowledgments The present study was partly supported by the Turkish Wrestling Federation. We gratefully acknowledge Mr. Önder Yaks¸i, the General Coordinator of the Turkish Wrestling Federation, who kindly helped in the organization of this study, making the present study possible. In addition, we appreciate and give our sincere thanks to two anonymous reviewers for their critical comments on earlier drafts of this paper.

PubMedID: 8384659 This trial was carried out to assess the response rate and survival benefit achieved, if any, by substitution of etoposide for doxorubicin and addition of methotrexate in combination with cyclophosphamide and vincristine in the treatment of 113 patients with small cell lung carcinoma (SCLC). Cyclophosphamide, etoposide, vincristine, methotrexate (CEV-M) yielded a response rate of 75% in localized disease (LD) and 63% in extensive disease (ED), versus 61% in LD and 52% in ED in the cyclophosphamide, doxorubicin, vincristine (CAV) arm. There was also a significant survival benefit for the responders in favor of CEV-M (21.7 ± 3.8 months of median survival compared to 13.6 ± 2.8 months in CAV arm) in patients with LD.

The effectiveness of empiric therapy was evaluated in 186 cancer patients who experienced 417 febrile and neutropenic episodes during the last 4 years. Ninety-eight patients experiencing 244 episodes were treated with empiric antibiotic therapy with mezlocillin and tobramycin. The response rate to the empiric treatment was 88% for bacterial infections and 58% for unexplained fever. Fluconazole 200 mg IV daily was added to the empiric antibiotic therapy in the subsequent 173 episodes experienced by 90 patients, of whom 12 had also been evaluated in the previous group. The addition of fluconazole improved the response rates to 90% in bacterial infections and to 86% in unexplained fever. In proven fungal infections in neutropenic patients, empiric antibiotic therapy combined with higher doses of fluconazole seemed beneficial. This study suggests that fluconazole 200 mg IV daily is an effective adjunct to empiric antibiotic therapy, but higher doses of fluconazole are preferable in proven fungal infections. © 1992 Excerpta Medica, Inc., an Elsevier U.S. Holdings, Inc. Company. All rights reserved.

PubMedID: 21293929 In this study, we investigated the epidemiological characteristics of VVC among pregnant women. We conducted a prospective survey among 372 pregnant women to investigate the prevalence, clinical forms, etiological agents, and predisposing factors of VVC. In addition, we determined the relationship between vaginal and rectal flora by simultaneously obtaining one high vaginal swab and one rectal swab from each patient using sterile cotton-tipped swabs. Furthermore, we compared the recovery and identification performances of chromID Candida agar to Sabouraud dextrose agar with gentamicin and chloramphenicol. Clinically and mycologically confirmed cases of VVC were detected in 139 (37. 4%) and vaginal colonization described in 42 (11. 3%) of 372 pregnant women. Rectal cultures were also positive in 98 of the 139 (70. 5%) VVC cases. Candida albicans and C. glabrata were identified in vaginal samples in 58. 0 versus 19. 0% and from rectal samples in 49. 0 versus 13. 5%, respectively. Increases in gestational week and gravidae were identified to be statistically significant in patients with acute VVC (AVVC) and symptomatic recurrent VVC (RVVC), and asymptomatic RVVC (P = 0. 04 and P = 0. 03, respectively). In the laboratory diagnosis of VVC, specifically tailored chromogenic media are reliable tools for both the recovery and rapid identification of common Candida spp., particularly C. albicans, as well as for the detection of polyfungal populations in vaginal samples (P > 0. 05). In addition, rectal colonization is a common finding in cases of AVVC and symptomatic-RVVC cases and corresponds well with the presence of the same yeast species in the vagina. © 2011 Springer Science+Business Media B.V. TF2010BAP1 Acknowledgments This study was supported by the Research Fund of Cukurova University (Project No: TF2010BAP1). We gratefully acknowledge all of the patients who agreed to participate in this study. We also greatly appreciate to Selim Mısırlıoglu, MD; Erdem Akter, MD; Enver Naci Tuncer, MD; and Tuba Yuksel, MSc for their kind assistance in the laboratory workup.

Vulvovaginal candidiasis (VVC), particularly the recurrent form, remains an intractable problem for clinicians, microbiologists, and patients. It is essential to confi rm the clinical diagnosis by mycological methods and avoid empirical therapy. The recovery of yeast in fungal culture, such as on Sabouraud dextrose agar, remains the gold standard for diagnosis. In this investigation, we examined 474 participants, including 122 (25.7%) with acute VVC cases, 249 (52.5%) who had recurrent VVC (RVVC) cases, and 103 (21.7%) healthy controls. We also administered a questionnaire to obtain information on patient lifestyle and medical, gynecological, and sexual history. In addition, we compared the performance of chromID Candida agar (CAN2) to CHROMagar Candida (CAC) and Sabouraud dextrose agar with gentamicin and chloramphenicol (SGC2). The yeasts were identifi ed by conventional methods including the germ tube test, microscopic morphology on cornmeal–Tween 80 agar, and the commercial API 20C AUX system. We detected yeasts in 60 of 122 (49.2%) patients with acute VVC cases, 110 of 249 (44.2%) with RVVC cases, and in 35 of 103 (34%) healthy controls ( P 0.07). A total of 205 samples were found to be positive for fungi (43.2%), of which 176 (85.9%) were monofungal, and 29 (14.1%) were polyfungal. In addition, 198 of these samples (96.6%) were positive on CAN2, 195 (95.1%) on CAC, 189 (92.2%) on SGC2, and 183 (89.3%) samples on all three ( P 0.17). The 234 yeast isolates recovered were C. albicans ( n 118), C. glabrata ( n 82), C. kefyr ( n 11), C. krusei ( n 9), C. lipolytica ( n 3), C. colliculosa ( n 2), C. parapsilosis ( n 2), C. pelliculosa ( n 2), C. tropicalis ( n 2), and other species of Candida ( n 3). Of the 29 polyfungal populations, 28 (96.6%) were detected in CAN2, 25 in (86.2%) CAC, and 25 (86.2%) on both ( P 0.35). Notably, we detected the high predominance of C. albicansC. glabrata (86.2%) in polyfungal populations. Briefl y, the detection of C. albicans after 24 h of incubation was easier on CAN2 (64.4%) than on CAC (25.4%). This study showed that CAN2 is a rapid and reliable medium for immediate identifi cation of C. albicans and for detecting polyfungal populations in vaginal specimens. We observed that the use of antibiotics, intrauterine devices, as well as, perineal laceration, short anovaginal distance ( 3 cm), and genital epilation in common areas are predisposing factors for RVVC ( P 0.001). In addition, we detected that the use of menstrual pad, using an (IUD), and having a history of childbirth increased the risk of both acute and recurrent VVC ( P 0.01), whereas the use of a daily pad and walking daily signifi cantly decreased the risk of both acute and recurrent VVC ( P 0.01).

TEZ9543 Tez (Doktora) -- Çukurova Üniversitesi, Adana, 2010. Kaynakça (s. 144-147) var. xii, 148 s. : tablo ; 29 cm. Bir tanı testinin doğruluk ölçütlerinin değerlendirilmesinde yansız tahmin ediciler tercih edilir. Ancak, testin doğruluk ölçütlerinin belirlenmesinde yanlılıklar ile karşılaşılabilir. Bu yanlılıklardan sıklıkla görülenleri; Doğrulama yanlılığı ve Kesin olmayan altın standart yanlılığıdır. Bu tezde bu yanlılıkların düzeltilmesinde kullanılan yaklaşımlar incelenmiş ve bu yaklaşımların sorun yaşadığı durumlarda alternatif olabilecek yaklaşımlar önerilmiştir. Doğrulama yanlılığı çalışmaya alınan tüm bireylere altın standart testin uygulanmaması sonucunda ortaya çıkar. Tanı testinin ikili ölçüm olması durumunda yanlılığın düzeltilmesinde; altın standart uygulanmayan bireylerin sonuçları eksik ölçüm olarak kabul edilirse, eksik ölçüm mekanizmasına göre, Begg&Greenes, Zhou sınırları, Lojistik Regresyon, Çoklu İmputasyon gibi yaklaşımlar kullanılır. Bu tezde bu yaklaşımlara Yapay Sinir Ağları yaklaşımı da alternatif olarak eklenmiş ve yaklaşımlar çeşitli karşılaştırma başlıkları altında üretilmiş ve gerçek veri setleri kullanılarak karşılaştırılmıştır. Kesin olmayan altın standart yanlılığı ise, altın standart olarak kabul edilen bir testin olmaması durumunda bu test yerine daha düşük sınıflama başarısı olan bir testin, kesin olmayan altın standart testinin, kullanılması ile ortaya çıkar. Yanlılığın düzeltilmesinde tanı testinin ölçüm ölçeğine göre Tutarsızlık Çözücü, Bileşke Referans Standardı, Zhou’nun Parametrik olmayan ROC eğrisi tahmini, Bayes yaklaşımı gibi yaklaşımlar kullanılır. Bu tezde bu yaklaşımlar ile bunlara alternatif olarak önerilen Gizli Sınıf Analizi yaklaşımı, üretilmiş ve gerçek veri setleri kullanılarak karşılaştırılmıştır. Üretilmiş veri seti ile yapılan tüm karşılaştırmalarda yaklaşımların sonuçlarını etkileyebilecek: değişen veri seti büyüklüğü, değişen hastalık prevelansı, değişen doğruluk ölçütü gibi durumlar göz önüne alınmıştır. Doğrulama yanlılığını düzeltmede; kesin tanı doğrulaması yapılmamış bireylerin rastgele seçilmiş olması durumunda Begg&Greenes yaklaşımı, rastgele olmaması durumunda ise Lojistik Regresyon ve Yapay Sinir Ağları yaklaşımlarının uygun yaklaşımlar olabileceği görülmüştür. Kesin olmayan altın standart yanlılığını düzeltmede; tanı testinin ikili ölçüm olması durumunda uygun prevelans aralığı ve yüksek doğruluk ölçütünde Gizli Sınıf Analizi; tanı testinin sıralı ölçüm olması durumunda yine uygun prevelans aralığı, yeterli sayıda kategori ve test varlığında Gizli Sınıf Analizi; tanı testinin sürekli ölçüm olması halinde ise küçük örneklem, düşük eğri altında kalan alan ve karıştırıcı etkili değişken varlığı dışındaki durumlarda Bayes yaklaşımı, bu yaklaşımın sorun yaşadığı durumlarda ise Gizli Sınıf Analizi önerilebilir yaklaşımlardır. In the evaluation of accuracy for a diagnostic test, unbiased estimators for accuracy measures are preferred. However, some bias in measuring diagnostic accuracy of the test may occur. The most common biases among encountered are verification bias and imperfect gold standard bias. In this thesis, proposed approaches for correcting these biases were examined and alternative approaches were proposed for the situations involving problems for the current approaches. Verification bias arises when not all patients in the study are taken the gold standard test. When the diagnostic test is binary, depending on the missing mechanism, Begg&Greenes, Upper and Lower Bound by Zhou, Logistic Regression orMultiple Imputation approaches are used for correcting this bias. In this thesis, a new approach, called Artificial Neural Networks, was alternatively proposed and compared with other approaches in different comparison schemes using simulated and real data sets. Imperfect gold standard bias arises when using an imperfect gold Standard test instead of gold standard test. Depending on the measurement scale of new diagnostic test, Discrepant Resolution, Composite Reference Standard, Non Parametric AUROC Estimation by Zhou and Bayesian Approaches are used for correcting this bias. In this thesis, a new approach, called Latent Class Analysis, was alternatively proposed and compared with other approaches in different comparison schemes using simulated and real data sets. In the simulations, in order to measure the correct performance of the approaches and to determine deficiencies involving, data were generated with a variety of experimental conditions; changing sample size, disease prevalence, and diagnostic accuracy of new test. When the missing mechanism is MAR, Begg&Greenes approach and when the missing mechanism is NMAR, Logistic Regression or Artificial Neural Networks approaches are preferable for correcting verification bias. When the measurement scale of the new test is binary and disease prevalence and diagnostic accuracy values are well-defined, Latent Class Analysi approach; when the measurement scale is ordinal and disease prevalence and the number of test and test categories are appropriate, Latent Class Analysis approach; and finally when the measurement scale is interval or ratio and with large sample, high AUROC and without confounding variable, Bayesian Approach and for other situations Latent Class Analysis approach are the best approaches for correcting imperfect gold standard bias. Bu çalışma Ç.Ü. Bilimsel Araştırma Projeleri Birimi tarafından desteklenmiştir. Proje No: TF2007D10.

PubMedID: 19660236 Background: Juvenile myoclonic epilepsy (JME) is a well-defined idiopathic generalized epileptic syndrome, and diagnostic criteria for JME are to have a normal brain imaging and clinical evidence of typical epileptic seizures. The aim of this study is to evaluate electrophysiological and neuroimaging findings of JME and determine their relationship with prognosis. Methods: Thirty-two patients (23 women and nine men) with a mean age of 22 (16-37) years were included in this study. Interictal electroencephalography (EEG), magnetic resonance imaging (MRI) and single photon emission computed tomography (SPECT) were carried out in all patients. Results: Analysis of premedication EEGs revealed primary generalized pattern activity in 75% (n=24) and focal abnormalities in 18.75% (n=6). MRI was abnormal in seven (21.88%) patients (two with arachnoid cyst, two with mild cerebral atrophy, two with ventricular enlargement and one with single gliotic lesion), and SPECT imaging detected hypoperfusion in 15 (47%) patients. Hypoperfusion was mostly found on the parietal lobe. Conclusion: We found that, after medication, only 6.25% of EEGs had primary generalized pattern activity (p>0.0001); nevertheless, the prognosis was good in patients who had typical EEG findings (p=0.106). The prognosis of patients with MRI abnormalities was grave (p=0.023). Twenty percent of the patients who had SPECT abnormalities were seizure free, and 80% of them had been partially controlled (p=0.059). There were no correlations between MRI abnormalities, EEG and SPECT findings. © 2010 W. S. Maney & Son Ltd.

Tütün kullanımı dünyada ve Türkiye'de önlenebilir hastalık, sakatlık ve ölümlerin en önemli nedenidir. Sigaranın yol açtığı zararlara içmeyenlerde maruz kalmaktadır. Ülkemizde sigara tüketiminin kontrol altına alınarak, özellikle gençlerin korunması amacıyla, 2006-2010 yıllarını kapsayacak şekilde bir ?Ulusal Tütün Kontrol Programı? hazırlanmıştır. Bu program dahilinde sigara içiciliğin sorgulanması, sigara dumanına maruziyet ve tütün kontrol yöntemleri ile ilgili bilgi, tutum ve davranışların saptanması konusunda verilerin toplanması için ulusal ölçekli araştırmalar yapılması gerekmektedir.Bu çalışmada Adana İl Merkezinde Sigara Kullanımı, Sigaraya ve Dumana Maruziyet ve Tütün Kontrol Yöntemleri ile İlgili Bilgi, Tutum ve Davranışlar araştırılmıştır. Araştırma yöntemi olarak 'telefon ile anket' yöntemini kullanılmıştır. Görüşmeler sırasında anket sorularını yöneltmek ve verileri toplamak için `Bilgisayar Destekli Telefon Görüşmesi` sistemi kullanılmıştır. Görüşmeler Çukurova Üniversitesi Telefon ve İnternet İle Saha Araştırmaları Merkezi'nde (TİSAM) gerçekleştirilmiştir.Çalışmada toplam 586 kişiyle görüşme yapılmıştır. Ankete katılım oranı %79,7' dir. Araştırmaya katılan kişilerin yaş ortalaması 38,72±13,10 olarak bulundu.. Örneklemin sigara içme prevalansı %22,7 olarak bulundu. Sigara içme oranı erkeklerde %37,7 ve kadınlarda %14,8 bulundu. Görüşmecilerin % 25,1 ` i 1-10 yıl arası, % 15,3'ü 11-20 yıl arası, %10,9'u ise 20 yıldan fazla süredir sigara içtiğini bildirdi. Sigara içmeyen görüşmecilerin %39,1' i evlerinde, %52,2' si işyerlerinde sigara dumanına maruz kaldığı bulunduÇalışmamız göstermiştir ki; geniş bir coğrafyada fazla sayıda kişiye ulaşılması planlanan saha araştırmalarında telefonla görüşme yöntemini kullanmak maliyet ve zaman kazandıracak uygun bir yöntem olabilir.Anahtar kelimeler:Adana, sigara içiciliği, telefonla anket, maruziyet, prevalans. The tobacco use is one of the most important reason of morbidity, disability and mortality in Turkey and the world. Non-smokers are also exposed to hazards of tobacco smoke. In our country the National Tobacco Control Program is prepared and recommended to be put into force between the year of 2006-2010, to reduce tobacco consumption especialy to protect the younger population from harmful exposure of tobacco. There is need to collect data on smoke and conduct follow up studies to determine the trend in tobacco use and behavior, attitude and knowledge towards; second hand tobacco smoke and putting tobacco control program into practice within the national tobacco control program.In this study we investigated knowledge, attitude and behavior about smoking, second hand smoking and National Tobacco Control Program of people in Adana city toward. Telephone survey method was used to collect data. A `Computer Assisted Telephone Interviewing` system which record answers simultaneously was used during the interviews. The Computer Assisted Telephone Interviewing System used for sampling and collecting data was based on `TİSAMSoft` software. The study was conducted at `Telephone and Internet Research Center of Cukurova University`.We completed 586 interviews. The acceptance rate was 79.7%. Mean age was 38.72±13.10. The overall smoking prevalance was 23.4% and 37.7% for male and 14.8% for female population. The length of tobacco use was as follows: 25.1% 1-10 years, 15.3% 11-20 years and 10.9% longer than 20 years. Our study showed: that 39.1% and 52.2% of non smokers were exposed to second hand tobacco smoke at home and work respectivelyOur study showed that in the field studies which were designed to conduct on a large area and to reach large number of people, the telephone interview method has advantages for cost and time.Keywords: Adana, smoking, telephone interview, exposure, prevalance. 82

Tıp alanında yapılan çalışmalarda ölçümlerin tekrar elde edilebilirliği; deney hatasını küçültmede ve araştırma sonuçlarının benzer araştırma sonuçlarıyla kıyaslanmasında önemli bir rol oynamaktadır. Uyumluluk ölçüsü; bir hasta hakkında, iki klinisyenin değerlendirmelerinin tutarlılığının veya bir klinisyenin farklı zamanlardaki değerlendirmelerinin tutarlılığının ölçütüdür.Uyumluluk ölçütleri farklı değerlendiriciler veya aynı değerlendiricilerin farklı zamanlarda tekrar ölçümlerindeki uyumluluğunun belirlenmesinde; ölçümlerin kategorik ve sürekli oluşuna göre farklı uyumluluk ölçütleri kullanılması gerekmektedir. Bunlar temel uyumluluk ve genel uyumluluk indeksleri altında toplanmış olup bu çalışmada genel uyumluluk oranı,p0, özel uyumluluk oranı, ps, Kappa, Ağırlıklandırılmış Kappa, Kendall W, Goodman ve Kruskal ?, Kendall ?, Sınıf içi korelasyon katsayısı ve Krippendorff alfa literatürden alınan örnek veriler kullanılarak incelenmiştir. Ayrıca Çukurova Üniversitesi Tıp Fakültesi Hastanesi Cilt Hastalıkları Anabilim Dalında yapılan bir çalışmada elde edilen verilerin alt grubu farklı uyumluluk ölçütlerini değerlendirmede ve karşılaştırmada kullanılmıştır.Literatürdeki çalışmalardan elde edilen veriler için alternatif uyumluluk ölçütleri uygulanmış ve karşılaştırılmıştır. Bu karşılaştırmalar değerlendirici sayısına ve değişken tipine göre yapılmıştır.Çalışmada bulgular üç başlık altında değerlendirilebilir: Ordinal değerlendirme yapan çoklu değerlendirici için Kendall W' nin ölçümler yerine dizin değerlerini kullandığı için bazı durumlarda uyumu doğru yansıtamamaktadır. Bu nedenle Krippendorff alfa değerininde hesaplanması ve Kendall W değeri ile karşılaştırması doğru olacaktır. Nominal, ordinal veya sürekli ölçekte çoklu değerlendiricili eksik veri durumunda; Krippendorff alfa eksik değerlendirmesi olan bireyleri de hesaba kattığından diğer ölçütlerden daha avantajlıdır. Ayrıca ilişki katsayılarından Goodman ve Kruskal ?, Kendall ?, Pearson ve Spearman korelasyon katsayıları bazı durumlarda uyumluluğu doğru yansıtmadığından dikkatli kullanılmalıdır. Reproducibility of the measurements in medical studies has an important role both in reducing the error of experiment and in comparison of the results with results of similar research. Measurement of agreement is the criterion of the consistency between two clinicians? ratings of a patient or consistency of a clinician?s ratings of a patient over a number of occasionsAssesment of the agreement for different raters or repeated measurements by the same raters, depending on whether measurements are categorical or continuous require use of different measures. This study summarizes the measures of basic and general agreement indices such as p0, ps, kappa, weighted kappa, Fleiss kappa, Kendall?s W, Goodman and Kruskal?s ?, Kendall?s ?, ICC and Krippendorff?s alpha and evaluates them by using data from literature. Additionally data from a study conducted in the department of dermatology at the medical school of Cukurova University is used in the evaluation and comparison of different measures of agreement.Alternative measures of agreement were applied on and compared for data from literature. Comparisons of these measures are made depending on the number of raters and types of variables.The study concludes with three main findings: Kendall W does not reflect true agreement in some cases, because it uses rank of measurements instead of the originals. Thus calculating Krippendorff alfa and comparing it with Kendall W are recommended. For ratings made on nominal, ordinal or continuous scale for multiple raters with missing data, Krippendorff alfa has the advantage that it uses information from individuals with missing ratings. And lastly, the association measures Goodman and Kruskal ?, Kendall ?, Pearson and Spearman correlation coefficient that should be used with care becuause they may not reflect true agreement. 77

PubMedID: 18626314 The clinical ABCD criteria are still recommended to both physicians and laymen when checking moles. The aim of this study was to determine the level of interrater reliability and therefore objectivity in rating for one of these criteria, namely border irregularity. Five professors, five residents, five nurses, and 10 students rated a set of 54 clinical images of pigmented skin lesions for border irregularity. After a descriptive presentation, rating was again carried out on another set of 54 images. In all groups, the agreement was moderate or substantial before the presentation and increased after the presentation. An almost perfect agreement was achieved by the professors after the presentation. Although both experience and receiving information could increase the level of interrater reliability, the disagreement was usually sufficient to suggest subjectivity in rating for border irregularity. © 2008 Wolters Kluwer Health.

We reviewed the clinical, radiological, surgical, and histopathological features of patients with meningiomas to identify factors that can predict tumor recurrence after ‘‘microscopic total removal,’’ to improve preoperative surgical planning, and to help determine the need for close radiological observation at shorter intervals or the need for radiotherapy as an adjuvant treatment in the early postoperative period. Clinical data, magnetic resonance imaging studies, angiographic data, operative reports, and histopathological findings were examined retrospectively in 137 patients with a meningioma treated microsurgically and with no evidence of residual tumor on postoperative MR images. Based on univariate analysis, tumor size, a mushroom shape, proximity to major sinuses, edema, osteolysis, cortical penetration, signal intensity on T2-weighted MRIs, pial-cortical arterial supply, presence of a brain-tumor interface in surgery, Simpson’s criteria, and histopathological classification were significant predictors for recurrence. However, age, gender, location of tumor, dural tail, calcification, signal intensity on T1-weighted images, and histopathologic subtypes in the benign group were not significant predictors. By Cox regression analysis the most important variables related to the time to recurrence were mushroom shape, osteolysis, dural tail, and proximity to major sinuses. Aggressive surgical therapy with wider dural removal should be considered in the presence of the preoperative predictors of a recurrence. Close radiological observation at shorter intervals or radiotherapy should be considered as adjuvant therapy in high-risk patients based on surgical findings predicting recurrence related to the brain-tumor interface, Simpson’s criteria, and histopathological findings in the early postoperative period.

The aim of this study is to determine the prognosis of patients with seizure onset from 1 to 24 months of age in respect to epilepsy, developmental outcome, and neurological status. It also aims to determine predictive factors regarding an unfavorable prognosis. Seventy-five patients were retrospectively analyzed. Univariate analysis revealed the following findings: (1) mental retardation at initial admission, abnormal neurological finding, infantile spasm, use of more than 1 antiepileptic drug, epileptic activity on electroencephalography (EEG) of neonatal seizure, and perinatal anoxia were significant risk factors with regard to developmental outcome; (2) mental retardation at initial admission, abnormal neurological finding, infantile spasm, use of more than 1 antiepileptic drug, epileptic activity on EEG, symptomatic etiology, history of neonatal seizure, and perinatal anoxia were significant risk factors regarding neurological status; and (3) mental retardation at initial admission, neurological abnormality, infantile spasm, use of more than 1 antiepileptic drug, epileptic activity on EEG, status epilepticus, symptomatic etiology, seizure frequency of more than once per week, history of perinatal anoxia, and neonatal seizure were significant risk factors regarding epilepsy prognosis. In addition, multivariate analysis revealed that neurological abnormality and use of more than 1 antiepileptic drug were significant for developmental outcome, that epileptic activity on EEG and use of more than 1 antiepileptic drug were significant for neurological status, and that perinatal anoxia, infantile spasm, and status epilepticus were significant for epilepsy prognosis. These findings suggest that neurological abnormality, use of more than 1 antiepileptic drug, infantile spasm, status epilepticus, and perinatal anoxia are unfavorable predictive risk factors regarding the prognosis of patients with seizures that have an onset from 1 to 24 months of age.

Ülkemizde beslenme ile ilgili araştırmalarda, Amerika veya Avrupa'da güvenilirlik ve geçerlilik çalışmaları yapılmış beslenme anketleri tercih edilmektedir. Fakat beslenme durumunun belirlenmesi çalışmalarında bu beslenme anketlerinin kullanılabilmesi için bu soruların geçerlilik ve güvenilirliğinin ülkemizde araştırılması gerekmektedir. Bu tez ile ülkemizde uygulanabilecek ``Beslenme Sıklığı Anketinin'' geliştirilmesi ve bu anketin geçerlilik ve güvenilirliğinin değerlendirilmesi amaçlanmaktadır.Bu tezde Kayseri ili Ahmet Gündeş Sağlık Ocağı 4 nolu bölgesinde yaşayan, 15?49 yaş arası kişiler arasından seçilen 100 kişiden, beslenme sıklığı anketini ve 24 saatlik beslenme kayıtlarını doldurmaları istenmiştir. Bu işlem anketin güvenirlilik çalışması için 2006 yılı Mart ve Kasım aylarında iki defa yapılmıştır. Beslenme anketindeki genel özellikler, yemek pişirme şekilleri ve kullanılan yağlarla ilgili soruların uyumu için Mc Nemar test istatistiği, Kappa istatistiği ve Gama katsayısı kullanılmıştır. ?Wilcoxon? testi ile gıdaların tüketim miktarları karşılaştırılmıştır. Korelasyon için Spearman rank korelasyon katsayısı kullanılmıştır.İki dönemde yapılan anketler arasındaki uyum incelendiğinde; sigara içme, kilo değişimi, diyet uygulanması, uyku, stres durumları ve fiziksel aktivitelerle ilgili soruların yanıtlarının genellikle uyumlu olduğu görülmüştür. Kullanılan yağlar ve yemek pişirme şekillerinin uyum yüzdelerine baktığımızda, genellikle bu sorulara verilen yanıtlarında uyumlu olduğu görülmüştür. Yemek pişirme şekilleri, yemek pişirmede kullanılan yağlar ve salata yapmada kullanılan yağlar ile ilgili soruların diğer seçeneklerinin sıklıkları ile ilgili yanıtlarında ve diğer aktiviteleri yapma sıklığı sorusunun yanıtında uyum % 40'ın altında bulunmuştur. Korelasyon katsayılarına göre, iki beslenme sıklığı anketi arasındaki tüketimler için iyi derecede uyumlu (0,7 ve üzeri) bulunan gıdalar; Süt, içli köfte, balık, revani, irmik helvası, işkembe çorbası ve greyfurttur. Balık çorbası, diğer deniz ürünleri ve maydanozun tüketimleri için iki anket arasında anlamlı korelasyon saptanamamıştır. Ankette bulunan diğer 79 gıdanın korelasyonları orta derecede (0,5?0,7) bulunmuştur.Sonuç olarak, uyarlanmış/geliştirilmiş olan beslenme sıklığı anketinin güvenilirliği orta ölçüdedir. İsteksizlik ve rasgele işaretlemelerden kaynaklanan hataların düzeltilmesi, katılımın arttırılması ve yüz yüze görüşme yönteminin kullanılması beslenme sıklığı anketlerinin güvenilirliğini daha da artıracak faktörlerdir.Anahtar kelimeler: Beslenme sıklığı anketi, geçerlilik, güvenilirlik. Nutritional surveys in Turkey are generally conducted based on questionnaires developed and used in similar studies in America or European countries. Before use, the validity and reliability of those questionnaires were investigated in those countries. In order to use such a questionnaire in our country, the validity and reliability studies should be done. Therefore the purpose of this thesis is to adapt / develop ?food frequency questionnaire? for our country and to evaluate the validity and reliability of such questionnaire.In this thesis, food frequency questionnaire and 24-hour food records were filled by 100 participants, who were selected in 15-49 aged people of 4th region of Ahmet Gündeş Village Clinic of Kayseri in March, 2006. In order to make reliability study, the questionnaire and 24-hour food records were filled second time by the same persons in November, 2006. In the analysis of questions related to general characteristics, modes for cooking and oils used in preparing meals, etc. McNemar test, Kappa statistics, Gamma coefficients were used. The food consumptions from two questionnaires were compared using Wilcoxon test. For correlations, Spearman Rank Correlation Coefficients were used.When the consistency of two questionnaires filled in March and November were examined, answers to questions related to general characteristics such as smoking, weight change, diet, sleeping period, stress, physical activities etc. were highly correlated. Similarly, answers to questions related to oils used in preparing meals and modes for cooking were highly correlated, too. However in answers to ?Other? choices of questions related to modes of cooking and oils used in preparing meals and salad, the agreement were low (less than 40 percent). Based on correlation coefficients, for each of following foods; meatballs stuffed with cracked wheat, fish, sweet semolina pastry, semolina helva, tripe soup, grapefruit the agreement was high between two questionnaires. On the other hand, fish soup, other sea food and parsley consumption correlations were not significant. Agreement in the remaining 79 foods in two questionnaires was moderate.As a conclusion, reliability of adapted / developed new food frequency questionnaire was moderate. The reliability of questionnaire could be improved by correcting errors and bias resulting from unwillingness and careless approach by participants, increasing the number of participant and face to face interviews.Key Words : Food frequency questionnaire, validity, reliability. 101

ÖZETKlinik Denemelerde, Çeşitli Vekil Değişkenlerin Gerçek Sonlanım Değişkenleri YerineKullanılabilme Özelliklerinin Yeni Geliştirilmiş Kriterlere Göre AraştırılmasıKanser veya kronik hastalıkların tedavisi ile ilgili klinik denemelerde incelenenyeni tedavilerin faydaları varsa bir an önce kullanılması gerekmektedir. Yaşam süresiyerine onunla ilişkili olan ve tedavinin etkinliği hakkında daha erken bilgi veren birVekil Sonlanım (VS) değişkeni kullanılabilirse, etkinliği belirlenen yeni tedavi dahaerken kullanıma sunulabilecektir.Bir VS değişkeninin Gerçek Sonlanım (GS) değişkeni yerine kullanılabilmesi içinVS değişkeninin geçerliliği belirlenmelidir. Prentice 1989'da VS değişkenini ve geçerlilikölçütlerini tanımlamıştır. Daha sonra sırasıyla açıklanan oran (PE), relatif etki (RE) vedüzeltilmiş ilişki (DI) ölçütleri geliştirilmiştir. Sonra Buyse et.al. (1998) RE ve DIölçütlerini çok merkezli bir düzende R-kare(Trial) ve R-kare(Individual) ölçütlerinegenelleştirmiştir.Bu tezde NSABP kliniksel deneme protokollerinden (C-01, C-02, C-03 ve C-04)elde edilen veriler kullanılarak, VS değişkeni olabilecek kesikli ve (sansürlü) süreklideğişkenlerin geçerliliğinin belirlenmesi için geliştirilmiş olan ölçütlerin incelenmesiamaçlanmıştır.Geçerlilik ölçütlerinin güven aralıklarının (GA) belirlenmesinde, Delta ve Fielleryöntemlerine ek olarak Bootstrap yöntemi uygulanarak karşılaştırılmışlardır. Copulayaklaşımında R-kare(Individual) ölçütünün tahmin edilmesinde ?Likelihood? veparametrik olmayan yöntemler kullanılmıştır. NSABP verileri kullanılarak, nüks venüksüz sağ kalımın (NSK) GS için VS özelliği incelenmiştir.Tek bir araştırmadan elde edilen geçerlilik ölçütlerinden PE ve RE, tedavilerarasında büyük farklar olmadıkça, kullanılmalarının uygun olmadığı gözlenmiştir.Tüm protokollerde NSK, VS olarak incelenmiştir. Ancak sadece C-03 veri setindeVS olarak (R-kare(Trial)=0.761) geçerliliği gösterilebilmiştir. NSK için R-kare(Trial) ileR-kare(Individual) geçerlilik ölçütlerinin incelenmesi sonucunda NSK'ın orta ölçekli birVS değişkeni olduğu sonucuna varılmıştır.Kolon kanseri ile ilgili literatürden elde edilen verilerin analizi sonucundahastalıksız sağ kalım süresinin deneme bazında (R-kare(Trial)=0.904) iyi-cok iyi VSözelliği gösterdiği gözlenmiştir.Sonuç olarak a-geçerlilik değerlendirilmelerinde meta-analitik yaklaşımdakullanılan R-kare(Trial) ve R-kare(Individual) ölçütlerinin diğer ölçütlere tercihedilmesi b-VS değişkenlerinin geçerliliği test edilirken, mümkün oldukça tedavilerininetkinliği ispatlanmış klinik deneme verilerinin kullanılması önerilir.Anahtar kelimeler: Vekil sonlanım, Geçerlilik, Copula, Sansürlü Değişken, Kanser. ABSTRACTThe Validation of Different Surrogate Endpoints in Clinical Trials with SomeNewly Proposed Criteria.New treatments for cancer or other chronic diseases which are tested in clinicaltrials should be given approval as soon as they show effectiveness. If an earliermeasurable surrogate for survival can be used, the treatment under research can beapproved earlier if it?s effective. A new surrogate needs to be validated before it can beused instead of the true endpoint. Prentice (1989), proposed criteria for the validation ofsurrogate endpoints. Later, other criteria like Proportion explained (PE), Relative Effect(RE) and Adjusted Association (AA) were proposed. Buyse et.al.(1998), additionallyproposed two new criteria R-sqr(Trial) and R-sqr(Indiv) which link RE and AA to themeta-analytic approach.This Thesis aims to analyze the criterias proposed for validation of discrete,continuous and censored time variables from NSABP trial protocols (C-01, C-02, C-03and C-04) as surrogate endpoints. Additionally, Bootstrap confidence intervals for theproposed validation criteria?s were compared with those obtained from Fieller andDelta methods. In the Copula approach, likelihood and nonparametric methods wereapplied for the estimation of R-sqr(indiv) criteria.Data sets from NSABP trials were used for the validation of ?Relapse FreeSurvival? (RFS) as a surrogate for ?Overall Survival? (OS). In the one trial case forwhich PE, RE and AA were proposed, it was observed that these criteria are not veryuseful for the validation of surrogate endpoints, unless the treatment effect is highlysignificant.The validation of RFS as a surrogate for OS was tested for all protocols (C-01 toC-04). But RFS could be validated only for data set C-03 where R-sqr(Trial) was 0.761.By inspecting R-sqr(Trial) and R-sqr(Indiv) RFS was ranked as a medium surrogatefor OS.Using data from published clinical trials on colon cancer, Disease Free Survival(DFS) was ranked as a good to very good (R-sqr(Trial)=0.904) surrogate for OS, at leastat the trial level.As a conclusion, we propose a-the validation criteria?s R-sqr(Trial) and R-sqr(Indiv) in the meta-analytic approach is to be preferred to the other measures forvalidity of Surrogate and b- when determining the validity of variables as surrogate endpoint, data from clinical trial in which treatment effect is highly significant should beused.Key words: Surrogate Endpoint, Validation, Copula, Censored Variable, Cancer 138

To investigate the relationship between atopy and hepatitis A virus (HAV) and hepatitis B virus (HBV) infections, we studied 42 children who had had HAV infection (Group I), 28 children who had had HBV infection (Group II), and 31 children who were seronegative for both HAV and HBV infection (Control group). Serological tests for HAV and HBV infections (anti-HAV IgG, HBsAg, anti-HBc IgG) and allergic skin tests and specific IgE investigations for the detection of atopy were carried out. In this study, there was no significant divergence in the socio-demographic characteristics among the three groups. The rates of specific IgE positivity in children in the HAV seropositive group (11.9%) and in children in the HBV seropositive group (17.8%) were lower than in the control group (35.4%) (p=0.03 and p=0.22, respectively). Also, the number of children with respiratory allergic diseases (allergic rhinitis and/or asthma) both in the HAV seropositive group and in the HBV seropositive group were significantly lower than in the control group (p

PubMedID: 16044453 The PRAME (preferentially expressed antigen of melanoma) gene has been shown to be expressed in high levels in some solid tumors and hemopoietic neoplasias but not or only weakly expressed in normal tissues. It encodes an antigen recognized by autologous cytolytic T lymphocytes. PRAME is a good candidate for tumor immunotherapy and is a useful marker gene for detection of minimal residual disease (MRD). In this study, PRAME mRNA using real-time RT-PCR was studied in 74 adult cases with acute leukemia-68 had de-novo acute leukemia, 3 had chronic myeloid leukemia-blastic crisis (CML-BC), and 3 had myelodysplastic/myeloproliferative syndrome-blastic transformation (MDS/MPD-BT)-and the results were compared with 30 age-matched healthy volunteers. Nineteen of 74 cases with leukemia expressed PRAME, while only 2 controls showed weak expression. The prevalence of PRAME expression in AML and ALL cases was 30% and 17%, respectively. We did not find any important correlation between PRAME expression and clinical characteristics, such as age, sex, organomegaly/lymphadenopathy, Hb, WBC count, platelet count, LDH level, alkaline phosphatase, albumin, cell-surface antigens, response to therapy, or progression-free and overall survival. PRAME was monitored in 15 cases during remission and/or relapse. There was a good correlation between PRAME mRNA and hematological remission and/or relapse. Interestingly, PRAME was very high in one case with AML but was not found 3 months after allogeneic transplantation. PRAME mRNA is observed in about one-third of AML cases; it may be a useful marker to detect MRD, and it may also be a good predictor for the timing of donor lymphocyte infusions (DLI) in the post-transplant period in cases of molecular relapse. © 2005 Wiley-Liss, Inc.

PubMedID: 15725319 Background: Invasive procedures such as bone marrow aspiration in children may be painful and cause anxiety. We investigated the efficacy and safety of remifentanil (R) alfentanil (A), remifentanil-midazolam (RM), and alfentanil-midazolam (AM) combinations in providing analgesia and sedation for bone marrow aspiration. Methods: Eighty children undergoing a diagnostic bone marrow aspiration whose ages ranged from 5-16 years (mean 9.20 ± 3.00 years) were enrolled in this study. The patients were randomly assigned to one of 4 treatment groups. Vital signs, sedation and pain scores, somatic responses (sweating and tears) were recorded before, during bone marrow aspiration and after 5 and 15 min of the procedure. Results: There were no statistical differences between sedation and the CHEOPS scores of the four groups during and after the procedures. The VAS scores were significantly higher for group A compared with groups R and RM during the procedure (P < 0.008). There were no differences between the VAS scores in group AM compared with groups R and RM (P > 0.008). There were also no statistical differences among the VAS scores of four groups after the procedure (P > 0.008). All patients had adequate sedation and analgesia. None of the patients in the study had deep sedation, hypotension, bradycardia, hypoxemia, or respiratory depression. Conclusions: Remifentanil, alfentanil, remifentanil-midazolam, and alfentanil-midazolam combinations are effective in children and can be used safely in bone marrow aspiration which is a brief but painful procedure performed in the pediatric patient group. © 2004 Blackwell Publishing Ltd.

PubMedID: 15027046 Objective. This study was undertaken to evaluate the possible association between low levels of serum cholesterol and depression in the elderly. Background. The alteration of cholesterol content of synoptosomal membrane in response to low serum cholesterol levels has been shown to decrease the serotonin receptors in depressed patients. Previous studies suggest that low levels of serum cholesterol may be associated with the increased risk of depression in the elderly. Subjects and methods. This was a cross-sectional study where 189 subjects over 65 years old of both sexes were enrolled. Serum total cholesterol, HDL-cholesterol (HDL-C), LDL-cholesterol (LDL-C), and triglycerides were measured. Cognitive functions were evaluated with mini mental state examination survey (MMSES) and depression was assessed with Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I). Results. Of the 189 subjects, 42 were affected by depression. Low serum cholesterol level (cut-off ? 160 mg/dl) as well as the levels of HDL-C, LDL-C and triglycerides were not associated with depression in older men or women. Conclusions. In the present cross sectional study, there was no association between depression and low serum cholesterol levels after adjusted for confounding factors. Further studies are needed to clarify this suggestion with larger number of patients. Copyright © 2004 John Wiley & Sons, Ltd.

ÖZET PROGNOZDA VE TANIDA LOJİSTİK REGRESYON VE NEURAL NETWORK YAKLAŞIMI: EPİDEMİYOLOJİK VERİ SETLERİNDE KARŞILAŞTIRMALI UYGULAMALAR Hastalık tanısı koymak (diagnoz) (veya hastalıkta prognozu etkileyen faktörleri saptamak) bir çok tıbbi araştırmanın amacı olmuştur. Bu alanda yapılan çalışmalarda çeşitli bir çok istatistiksel yönteme başvurulmuştur. Yanıt değişkeni ikili (binary) bir değişken ise lojistik regresyon modelleri veya tekrarlı bölme (recursive partitioning) yöntemleri, eğer yanıt değişkeni birden fazla seçeneği içeriyorsa çoklu lojistik regresyon modelleri standart yöntemler arasındadır. Bununla birlikte son zamanlarda Neural Networks yaklaşımı da bu yöntemlere eklenmiştir. Neural Networks yaklaşımı genellikle bir tahmin veya sınıflandırma ya da kontrol problemlerinde kullanılmaktadır. Bu tezde iki yöntem, performansları yönünden farklı büyüklük ve özellikteki örneklemlerde, karşılaştırılıp, hangi durumda hangi yöntemin kullanılması ile ilgili karar verilmeye çalışılmıştır. Bu karşılaştırmalar Lojistik Regresyon, Neural Networks ve Neural Networks' ün farklı bir uygulaması olan Genetik Algoritmalı Neural Networks sonuçlarını içermektedir. Performansın doğru ölçülebilmesi için farklı özelliklerde oluşturulan örneklemler eğitim ve test alt örneklemlerine bölünmüştür. Örneklemler büyüklüğüne göre büyük, orta ve küçük olmak üzere, astımlı öğrenci oranına göre 0.1, 0.2, 0.3, 0.4 ve 0.5 olmak üzere toplam 8 örneklemden oluşmuştur. Yöntemlerin karşılaştırılmasında, elde edilen Receiver Operating Characteristic (ROC) eğrileri altında kalan alanlar ve duyarlılık-seçicilik değerleri kullanılmıştır. Büyüklük farklılıklarına göre en yüksek alan değerleri, büyük veri setinin eğitim ve test örneklemlerinde 0.848 ve 0.825 ile Lojistik Regresyona, orta büyüklükteki veri setinin eğitim ve test örneklemlerinde 0.882 ve 0.735 ile Neural Networks'e ve küçük veri setinin eğitim örnekleminde 1.000 ile Genetik Algoritmalı Neural Networks'e, test örnekleminde ise 0.902 ile Lojistik Regresyona aittir. Ancak bu değerlerin güven aralıkları birbirlerini içerdiği için yöntemler arasında fark bulunamamıştır. Oran farklılıklarına göre ise tüm eğitim örneklemlerinde Genetik Algoritma yüksek değeri almıştır. Astımlı oranı 0.3 ve 0.3'ten yukarıda olan 3 örneklemdeki Genetik Algoritma eğitim sonuçları, güven aralıkları incelendiğinde, Lojistik Regresyon sonuçlarından daha iyidir. Tüm test örneklemlerinde ise Lojistik Regresyon daha yüksek alan değerleri almasına rağmen güven aralıklarına bakıldığında yöntemler arasında fark bulunamamıştır. Bu tez, eldeki verilere en iyi uyarlanmış model seçimi konusundaki kararın, aynı amaç için kullanılabilecek tüm analizlerin sonucuna bakılmadan verilemeyeceğini göstermiştir. Anahtar Sözcükler : Duyarlılık - Seçicilik, Genetik Algoritma, Lojistik Regresyon, Neural Networks, ROC Eğrileri. ıx ABSTRACT COMPARATIVE STUDIES OF LOGISTIC REGRESSION AND NEURAL NETWORK IN THE DIAGNOSIS AND THE PROGNOSIS: APPLICATIONS WITH EPIDEMIOLOGIC DATA Several medical studies are conducted with a purpose for diagnosis and prognosis for many diseases. In these studies, several statistical approaches are used such as logistic regression model with a binary response, recursive partitioning, multiple logistic regression model when the response is more than binary (multi response). In addition to these approaches, a new method, Neural Networks have been recently used by researchers. In general, prediction (forecasting) or classification or controlling is the main use of this methodology. In this thesis, two methodologies, Logistic Regression and Neural Networks, were compared with using randomly chosen samples that are different in size and in properties. In these comparisons ordinary Logistic Regression, Neural Networks and Neural Networks with Genetic Algorithm were used. Samples were divided into two as training and test sets. Samples were used with different sizes as large (n = 3300), medium (n = 1000) and small (n = 100) and with different properties as having different proportions of students with asthma (p = 0.1, 0.2, 0.3, 0.4 or 0.5). The comparisons among three methods were primarily based on analysis of the receiver operating characteristic (ROC) curves as well as a number of scalar performance measures pertaining to the classification contingency tables. For all samples none of methods outperformed one another when confidence interval for ROC curves areas were used as comparison. When using sample large in size, the largest ROC curve areas were obtained from Logistic Regression with 0.848 in training and 0.825 in testing sets. When using sample medium in size, Neural Networks had the largest area with 0.882 and 0.735 in training and testing sets respectively. In the case of small sample, Neural Networks with Genetic Algorithm had the area 1.000 in the training set but in the test set, Logistic Regression outperformed the others with the area 0.902. For samples with different properties, in all training Neural Networks with Genetic Algorithm had taken the lead. Moreover, samples in which asthma prevalence is greater than or equal to 0.3, it can be said that Neural Networks with Genetic Algorithm is superior to the others according to the confidence interval of areas. On the other hand, although Logistic Regression got the better results in the test set, three methods were similar. In this thesis, it is concluded that the decision as to choosing the model that fit data best can not be finalized before all analysis with possible competing methods are performed. Keywords: Genetic Algorithm, Logistic Regression, Neural Networks, ROC Curves, Sensitivity and Specificity 103

PubMedID: 15224361 The aim of this study is to detect the possible role of hepatitis C Virus (HCV) in lymphomagenesis. HCV-RNA and anti-HCV antibodies were studied in tissue and serum samples taken from patients with non-Hodgkin's Lymphoma (NHL). The prevalence of HCV, the clinical presentation of these cases, and association with histologic subtypes were determined. RT-PCR was used to detect the HCV-RNA in serum and tissue samples. The anti-HCV antibodies were tested with microparticle enzyme immunoassay. Immunohistochemistry with the ABC method was used to detect the HCV core protein in HCV-RNA+ cases. RNA could be detected in 30 of 35 cases, and other tests were performed in these 30 samples. HCV-RNA was detected in 11 tissue samples (11/30, 37%). HCV core protein was studied in 10 of 11 HCV-RNA+ cases, and 1-3% nuclear staining was found in only 2 samples. Serologically, HCV-RNA was detected in 7 of 30 samples (23.3%) and anti-HCV antibody was detected in 3 of 30 samples (10%). Detection of HCV-RNA in 37% of the lymphoma tissue samples suggests that HCV may have a role or is a contributing factor in the pathogenesis of lymphoma. The very low HCV core protein in lymphoma tissues may be due to the low viral load in lymphoid tissues and/or higher sensitivity of the PCR method. Detection of anti-HCV antibody in only three cases may be associated with undetectable levels of antibodies due to the immune deficiency in cases with NHL. © 2004 Wiley-Liss, Inc.

PubMedID: 15008944 The objective of this study was to determine prognostic factors related to death from adult tetanus. Fifty-three cases of tetanus, 25 females and 28 males, were treated in Çukurova University Hospital during 1994-2000. The mean age was 46.6 years. Forty-one (77.7%) patients came from rural areas. Most (64.1%) cases had minor trauma, but 19 (35.8%) had deep injuries. The mean incubation period was 11.5 days. Mortality was high (52.8%), caused by cardiac or respiratory failure or complications, and was related to the length of the incubation period. In cases with an incubation period ?7 days, the mortality rate was 75% (p 0.07). Mortality was significantly associated with generalised tetanus (p < 0.05), fever of ?40 °C, tachycardia of >120 beats/min (p < 0.05), post-operative tetanus (p 0.03), and the absence of post-traumatic tetanus vaccination (p 0.068). Patients who were given tetanus human immunoglobulin or tetanus antiserum (p > 0.05) had similar outcomes. Patients who were given penicillin had a mortality rate similar to patients who were given metronidazole (p 0.15). The mortality rate was higher (92%) in patients with severe tetanus than in patients with moderate disease (53%). By multivariate analysis, the time to mortality caused by tetanus, and also the mortality rate, were both related significantly to age and tachycardia. © 2004 Copyright by the European Society of Clinical Microbiology and Infectious Diseases.

PubMedID: 14508793 Hepatitis C virus (HCV) is an RNA virus in the Flaviviridae family. It displays lymphotropism in addition to hepatotropism and extrahepatic manifestations are very well known. There are many studies showing an association between HCV infection and non-Hodgkin's lymphomas (NHL). In this study the evidence for HCV infection was studied in cases with NHL. To this end, anti-HCV antibody and HCV-RNA were screened in serum samples of cases with NHL using third-generation ELISA and RT-PCR. Anti-HCV antibody was studied in 223 patients and was found to be positive in 18 cases (8.1%). Anti-HCV antibody positivity was compared with our blood bank/blood donor population. There was an important increased risk of HCV infection - the common odds ratio was 34.56 and corrected odds ratio was 19.07. HCV-RNA was studied in 67 of 223 serum samples. HCV-RNA was found to be positive in 21 of 67 samples (31.3%). When compared with clinico-demographic parameters for anti-HCV and HCV-RNA, including age, nodal status, and grade (in evaluable cases), except age in cases with or without HCV-RNA, we did not find an important correlation with HCV status and clinical findings (P = 0.155; 0.442; 0.288 for anti-HCV and 0.027; 0,558; 0.126, respectively). These results suggest that HCV infection may be an important risk factor for lymphomagenesis and HCV-RNA is more useful for the detection of HCV infection in these immunosuppressed cases. Simultaneous detection of anti-HCV and HCV-RNA will be more informative in this population. © 2003 Wiley-Liss, Inc.

Saha Araştırmalarında `Discrete Multivariate ` Yöntemlerin Uygulanabilirliği Araştırması VI- ÖZ Bu çalışmada, üniversite öğrencilerinde sigara içmeyi etkileyen faktörlerin saptanmasında `univariate` ve `multivariate` yöntemlerin uygulanabilirliğini göstermek ve sonuçlarının karşılaştırılması amaçlanmıştır. Türkiye'nin 7 coğrafik bölgesinde bulunan toplam 23 üniversiteden, her bölgeyi temsilen 7 üniversite rastgele seçilerek araştırmaya alınmıştır. Kasan 1990 ile Şubat 1991 tarihleri arasında, bu üniversitelerde okuyan öğrencilerin 5834'üne anket uygulanmıştır. Anket; üp fakültelerinde okuyan öğrencilerde 1. 4. ve 6. sınıflara, tıp fakültesi dışındaki fakültelerde okuyan öğrencilerde ise 1. ve 4. sınıflara uygulanmıştır. Sigara içen ve hiç sigara içmeyen gruplar karşılaştırıldığında; öğrencilerde hergün sigara içmeyi etkileyen birinci faktör `en yalan arkadaşın sigara içme durumudur`. Lojistik regresyon analizi ile, arkadaşı hergün sigara içen öğrencilerin hergün sigara içme riski, arkadaşı hiç sigara içmemiş olanlara kıyasla 8.26 kat fazla olduğu bulunurken univariate yöntemle 13.43 kat bulunmuştur. Arkadaşı ara sıra sigara içen bir öğrencinin hergün sigara içme riski; arkadaşı hiç içmemiş olan bir öğrenciye kıyasla lojistik regresyonla 2.71 kat, univariate yöntemle 3.40 kat daha az bulunmuştur. Bu oran arkadaşı sigarayı bırakmış olan bir öğrenci için lojistik regresyonla 2.30 kat, univariate yöntemle 2.34 kat fazla bulunmuştur. İkinci risk aktörü `öğrencinin alkol kullanımıdır` Hiç alkol almayan öğrenciye kıyasla; haftada bir kezden fazla alkol alan öğrencinin hergün sigara içme riski lojistik regresyonla 7.98 kat, univariate yöntemle 12.66 kat ; ayda bir kezden fazla alkol alan öğrencinin hergün sigara içme riski lojistik regresyon ile 4.99, univariate yöntemle 7.53 kat ; altı ayda bir kez alkol alan öğrencinin hergün sigara içme riski lojistik regresyonla 3.65 kat, univariate yöntemle 5.04 kat daha fazla bulunmuştur. Üçüncü risk faktörü olarak `yaş`, dördüncü risk faktörü olarak `sigaranın zararlı olduğu düşüncesine katılım derecesi` ve sırayla; `cinsiyet`, `kiminle veya nerede kaldığı`, `okuduğu fakülte`, `ablasının, abisinin, annesinin sigara alışkanlığı`, `yaptığı aylık harcama`, ve `babasının eğitimi` hergün sigara içmeyi etkileyen diğer risk faktörleri olarak bulunmuştur. Polychotomous lojistijk regresyon(PLR) analizi sonuçlan dichotomous(DLR) analiz sonuçlarına benzemekte ancak bazı önemli farklılıkları vardır. PLR'da babanın sigara içme durumu 6. önemli değişken olurken DLR'da üç modelden hiç birine girememiştir. Bu nedenle odds değerlerinde bazı farklılıklar görülebilmektedir. PLR'da değişkenlerin önemlilik sırası; arkadaşın sigara içme durumu, alkol kullanımı, sigaranın zararlı olduğuna katılım derecesi, cinsiyeti, kaldığı yer, babanın sigara içme durumu, abinin sigara içme durumu, okuduğu fakülte, ablanın sigara içme durumu ve yaş şeklinde dir. DLR ve PLR'da verilen değişkenlerin önemlilik sıralamasını univariate analizlere dayanarak vermemiz pek mümkün olmayacaktır çünkü univariate yöntemde değişkenlerin birbirleri ile olan etkileşimleri göz önüne alınamamaktadır. Yukarda görüldüğü gibi univariate ve multivariate sonuçlar birbirinden çok farklı olabilmektedir. Model yaklaşımı ikiden fazla değişkenin aynı anda analizini mümkün kıldığından modelleme ile elde edilen odds değerleri univariate yönteme kıyasla daha güvenilirdir. `Univariate` analizler için SPSS ve EPİ6 paket programlan, `Multivariate` analizler için BMDP paket programının lojistik regresyon biriminden yararlanıldı. Anahtar sözcükler: Öğrenci, tıp, sigara, multivariate, lojistik V The applicability of multivariate analysis on survey data VII- Abstract Objective- To apply univariate and multivariate analysis methods in assessing risk factors related with smoking in university students and to compare these results. Design- First and final year medical and non medical (science) students from seven universities located in seven different geographical areas in Turkey replied anonymously to an adapted World Health Organization questionnaire in Turkish. Subjects-First, fourth and sixth (final) year medical students (totaI=3646) and first and last year non medical students (total=2188) from different universities. Results- The results of dichotomous logistic analysis and univariate analysis of daily smokers and never smokers are as follows: The first factor found associated with daily smoking was the smoking behavior of the best friend. The risk of being a daily smoker was 8.26 (logistic regression), 13.43 (univariate) times higher for students whose best friend was an daily smoker than for a never smoker best friend. The risk of being a daily smoker was about 2.71 (logistic regression), 3.40 (univariate) times higher for an occasional smoker best friend and 2.30 (logistic regression), 2.34 (univariate) times higher for an ex-smoker best friend. The second important factor was the use of alcohol. The risk of being a daily smoker was about 7.98 (logistic regression), 12.66 (univariate) times higher for students who used to take more than once a week alcohol versus non users. For students with alcohol consumption once a month this risk decreased to 4.99 (logistic regression), 7.53 (univariate) times and for alcohol consumption of once in six months the risk decreased to 3.65 (logistic regression), 5.04 (univariate) times. The third factor was age. Fourth the agreement with that smoking is hazardous. 5th factor was sex. The next factor was where or with whom does the student during his/her education live. The 7'th factor was faculty. Than the older sister's smoking status. The 9th factor was the older brother's smoking status. The 10th factor was mothers smoking status. Then the money spent per month, and finally father's education. The order of these risk factors could be given only with the use of multivariate analysis because in univariate analysis the interaction of these risk factors could not be interpreted. The order of the risk factors was the order in which they enter the stepwise logistic regression equation. As shown multivariate and univariate results differ. This difference is due to modeling, where more than two variables can be analyzed together. And it is clear that the reliability of multivariate results are higher than univariate analysis results. In the comparison of dichotomous and polychotomous LR the number of variables and the variables found significant differ for these two multivariate analysis. For example fathers smoking status was found to be significant in polychotomous LR but not in one of the three models with dichotomous LR. The order and the variables themselves differ for the model in polychotomous logistic regression. The number of variables in the model for dichotomous LR was about 12 variables but only 10 for polychotomous LR. In the case of smoking status for students, the comparison of dichotomous and polychotomous LR showed that it is not possible to define whether a student smokes occasionally or is a never smoker and it is not possible to define whether a student was an ex-smoker or never smoker. But in the case of daily smoker and never smoker dichotomous LR seems to be better than polychotomous LR. Univariate data analysis was done with the chi-square method. Dichotomous and polychotomous logistic regression methods were used for multivariate analysis. Univariate analysis was done using SPSSPC+ and EPI6 and multivariate analysis was done with BMDP. Key words : Smoking, medical, logistic, multivariate, student. VI 80

ÖZ 6-12 Yaş Okul Çağı Çocuklarında Tüberkülin Taraması ve Tüberkülin Misk Faktörleri Amaç : Hızla göç alan bir bölge olan Adana Ceyhan ilçesinde tüberküloz infeksiyon prevelansı ve tüberkülin pozitifliğini etkileyen risk/aktörlerini saptamak. Yöntem ve Gereç: 1995-1996 yıllarında toplam 8275 ilkokul birinci ve beşinci sınıf öğrencisinde tüberkülin deri testi uygulandı ve BCG öyküsü, cinsiyet, doğum yeri, ailenin memleketi, göç yılı, oda sayısı, evde yaşayan kişi sayısı, tüberkülozlu hasta ile temas öyküsü, annenin eğitimi, babanın eğitimi, anne yaşı ve mesleği, baba yaşı ve mesleği, evde içilen sigara sayısı gibi risk faktörlerini araştıran bir anket ile 2836 çocukta bir cross-sectional çalışma düzenlendi. Bulgular : Aşılanma oranı %75 olarak saptanmıştır. Aşı sayısı artıkça tüberkülin pozitiflik oranmm arttığı saptanmıştır. Tüberkülin reaksiyonu, pozitiflik sınırı 10 mm olarak kabul edildiğinde aşısız çocuklarda %9.2, tek aşılı çocuklarda %17.4, birden fazla aşılılarda %46.4, pozitiflik sınırı 15 mm olarak kabul edildiğinde ise sırasıyla %4.7, %8.0 ve %24.5 olarak saptanmıştır. Anket uygulanan tüm çocuklar, üç ayrı logistik regresyon modeli ile incelenmiştir. Modellerde sırası ile 5 mm, 10 mm ve 15 mm reaksiyonları bağımlı değişken olarak kabul edilmiştir. Her üç modelde de tüberkülin pozitifliği ile istatistiksel olarak ilişkili ve anlamlı bulunan ortak risk faktörleri temas, çocukta tüberküloz öyküsü, BCG öyküsü, BM, yaş, doğum yeri kriterleridir (p^O.05). Tartışma : Tüberkülin taramalarının, risk gruplarını belirleyen kısa bir anket formu ile birlikte değerlendirilmesi ve ppd pozitiflik sınırlarının ve tübeküloz çalışmlarının risk gruplarına yönelik olarak yeniden değerlendirilmesinde yarar vardır. Anahtar Kelimeler: Tüberkülin taramaları, risk faktörleri, tüberküloz infeksiyon prevelansı, tüberkülin reaksiyonu. VI Abstract Tuberculin Survey in 6-12 years old schoolchildren and risk factor of tuberculin posiüvity. Objective: To estimate the tuberculosis infection prevalence in a special district which has a lot of migrants from East of Turkey and find out the risk factors that effect positive tuberculin reactivity. Methods: A cross-sectional survey of tuberculin skin test was conducted in Adana-Ceyhan in 1995 and 1996 and a family administred questionnaire was used to collect the details of BCG skar, sex, place of birth, family orijin, year of migration, number of rooms, family size, contact history of tuberculosis, family education, age of mother, age of father, job of family, cigarette smoking history in the family and income of family. Results: The rate of tuberculosis infection was found 9.2 % in unvaccinated children (k 10 mm), and 17.4% in children who had vaccinnated once, and 46.4% in children who had been vaccinated more than once (£ 10 mm). Three general logistic regression model with positive tuberculin skin test reactions £> 5 mm, <> 10 mm,,> 15 mm) as the dependent variable were fitted to the data to examine the joint effect of risk foctors. A number of statisticaly significant assosiation were found between positive tuberculin test (£ 5 mm, > 10 mm, £ 15 mm) and; history of contact with tuberculosis, age, BMI, history of tuberculosis disease, birth place, history of BCG vaccine (p^O.05). Conclusions; Results of tuberculin surveys must be interpreted with determined risk factors which has been obtain by a predesign questionnaire. Key words: Risk factors, tuberculin survey, tuberculosis infection prevalence, tuberculin reactivity. vn 111

ÖZ Bu çalışma, 0-6 yaş grubu çocuklardaki büyüme- gelişme ve bunları etkileyen faktörleri saptamada antropometrik ölçütlerin önemini belirlemek amacıyla yapılmıştır. Çalışmada Doğankent Eğitim Araştırma Bölgesindeki 6 sağlık ocağından 0-6 yaş grubu 1800 çocuğun antropometrik ölçütleri kullanılmıştır. Araştırmaya bazı hastalıklı çocuklar dahil edilmemiştir, Çocuklardaki malnütrisyonun belirlenmesinde antropometrik ölçütlerin değerini saptamak amacıyla 1720 çocuğun antropometrik ölçütleri, malnütrisyonu etkileyen faktörlerin analizi için 1052 çocuğun sosyo-demografik ve geçirdiği hastalıkları ile ilgili bilgileri kullanılmıştır. Çocuklardaki sağlık ve beslenme durumunun saptanmasında; z-skor, yüzde, persentil yöntemleri ve yaşa göre ağırlık, yaşa göre boy uzunluğu, boy uzunluğuna göre ağırlık ölçütleri kullanılmıştır. Analizlerde çocukların yaşları, yakın olan yaş ve tamamlanmış yaş olmak üzere farklı fonksiyonlar kullanılarak hesaplanmıştır, Çocuklardaki malnütrisyon oranı, çocukların yaşları yakın olan yaş yöntemine göre hesaplandığında daha yüksek bulunmuştur. Her iki yaş hesaplama yöntemine göre; yaşa göre ağırlık ve yaşa göre boy uzunluğu ölçütlerine göre malnütrisyonlu çocuğun en az olduğu grup 0-6 ay yaş grubu olup fizikzel muayene sonuçları ile uyumluluk göstermiştir. Fakat boy uzunluğuna göre ağırlık ölçütü kullanıldığında en az malnütrisyonlu çocuk 25-48 ay yaş viigrubundadır. Z-skor yöntemine- göre, kırsal bölgedeki malnütrisyonlu çocukların oranı kentsel bölgeye göre daha yüksek bulunmuştur. Annelerin eğitim düzeyi yükseldikçe malnütrisyonlu çocukların oranlarında azalma görülmüştür. Sosyo-ekonomik düzeyi yüksek olan ailelerde malnütrisyonlu çocukların sayısı azalırken, düşük olan ailelerde bu sayı artmıştır. 5 ve daha fazla çocuklu ailelerde malnütrisyon oranı diğer ailelere göre daha yüksek bulunmuştur. Günümüzdeki beslenme durumunu belirlerken, z-skor sonuçları fiziksel muayene sonuçları ile en yüksek uyumluluk gösteren metod oluşu nedeni ile izlemelerde bu ölçütün kullanılması önerilebilir. Çocuklardaki hafif malnütrisyon yalnızca fiziksel muayene takibi ile saptanamadığı için; 0-6 yaş grubu çocuklarda büyüme ve gelişme takip edilirken fiziksel muayenenin yanında antropometrik ölçütlerin kullanılması gereklidir. Anahtar Kelimeler: Malnütrisyon, Antropometrik, Beslenme, Z-skor viii ABSTRACT This study was carried out to determine the importance of anthropometric indicators in assessing factors that affect the growth and development of children aged 0-6 years. The anthropometric measurements of 1800 children aged 0-6 years from 6 health ceteras district in the Region of Dogankent have- been used. Some unhealthy children with growth retardation were excluded. Anthropometric measurements of 1720 children have been used in assessing the value of anthropometric indicators for malnutrition among children. The data on 1052 children were available for the analysis of factors that affects growth and developments in classifying children as malnutrition, normal and obese. CDC/NCHS reference population was used and for z-score, percent, percentiles methods, weight, height for age and weight for height measurements were calculated. In throughout the study, ages were calculated as both the nearest age and the completed age. The malnutrition rate is higher when nearest age method was used. Based on both type of age measurements malnutrition rates determined by both weight for age and height for age is minimum in the 0-6 month age group and similar to the findings from the physical examination. However according to weight, for height methods 25-48 month age groups showed the least malnutrition rate. According to z-score, malnutrition rate among children is higher in rural than in urban and is lower as the level of education of mother increases. In addition malnutrition rate is lower among the children with higher ixsocio-economic, statue and is higher among the families with five or more children. When determining the present nutritional status of children, the results of z-score should be followed since it is method showing significant, similarity with the outcome, of physical examination. In conclusion, the use of anthropometric measurements is necessary for monitoring growth and development of children aged 0-6 years in addition to physical examination. Because physical examinations misses the children with mild malnutrition. Keywords : Malnutrition, Anthropometric, Nutrition, Z-score x 67

PubMedID: 7854769 The intracellular glutathione (GSH) content was measured in 73 patients with leukemia and compared with controls. GSH content was between 1.16 and 5.55 µmol/g protein (mean 2.96 ± 0.86) in the study group and between 0.5 and 1.48 µmol/g protein (mean 1.31 ± 0.27) in the control group, statistically significant difference (p = 0.0000). There was no significant difference between acute and chronic leukemias, lymphoid and myeloid leukemias and, more importantly, newly diagnosed and relapsed patients. GSH content did not change significantly with clinical and hematologic parameters such as age, sex, and initial hematologic findings. In addition, variable changes were detected over 24 h in 9 patients. It can be concluded that GSH content in leukemic cells was higher than in controls and showed a wide range. The absence of a relationship between GSH content and clinical and laboratory parameters suggested that GSH is not the sole determinant of response to cytotoxic drugs. GSH variation over a 24-hour period may be important in the timing and success of chemotherapy for leukemias. © 1995 S. Karger AG, Basel.

PubMedID: 8176937 Paraffin-embedded primary tumor specimens from 48 patients with breast cancer were examined for DNA ploidy, S-phase fraction (SPF), and concanavalin A (Con A) reactivity. The results were correlated with clinicopathological prognostic factors, including patients' age and menopausal status, stage of disease, nuclear grade, and size of the primary tumor. There were no associations among ploidy, SPF, Con A reactivity, and menopausal status, stage of disease, or size of the primary tumor. However, among patients who were 50 years or older, 81 % had diploid tumors and 73% had good reactivity (3+ or better staining score) with Con A. In contrast, among patients who were younger than 50 years, 45% had diploid tumors (P < 0.05) and 21% had good Con A reactivity (P < 0.05). Seven of 19 (37%) poorly differentiated tumors and 7 of 9 (78%) moderately differentiated tumors had good reactivity with Con A (P < 0.05). Reactivity of tumor cells with Con A in primary breast cancer tissues deserves further evaluation as a potential biomarker of prognosis. © 1994 Wiley-Liss, Inc. Copyright © 1994 Wiley-Liss, Inc., A Wiley Company