ISSN: 1300-0292
İndekslendiği Dizinler: SCIENCE CITATION INDEX EXPANDED
CINAHL, Index Copernicus,
Chemical Abstracts (CA),
Excerpta Medica / EMBASE
Dil: Türkçe, İngilizce
İçerik: Orijinal Araştırma, Derleme, Editöre Mektup, Olgu Sunumu, Tıp Eğitimi, Tıbbi Kitap İncelemeleri
|
|
|
|
ORIGINAL RESEARCH ARTICLES |
The Estimatation Of Risk Factors In Leukemia With Competing Risks
Dr. Canan DEMİRÜSTÜ,a Dr. Kazım ÖZDAMARa
aBiyoistatistik AD, Osmangazi Üniversitesi Tıp Fakültesi, ESKİŞEHİR
Objective: If more than one reason to cause death is present in a given disease state, a competing risks approach is used to calculate risk probabilities as to which cause or causes may predominate. The objective of this study was to calculate risk factors of sepsis and other causes of morbidity and mortality in acute and chronic leukemia patients.
Material and Methods: The Cox regression method was used to determine the effect of prognostic factors on death in competing risks with a parametric approach. In the nonparametric approach, the number of deaths and the specific population under risk were used to determine risks.
Results: In the solutions attained upon utilization of the parametric approach, the age of death and the age at treatment were taken as covariates and sex was taken as factor. In acute and chronic leukemia patients, no important covariate was seen in the patients who died from sepsis, which was why none were included in the model. In acute leukemia patients, sepsis was found to be a determining risk factor. In chronic leukemia patients, deaths caused by other reasons was determined a risk factor. Deaths caused by other reasons in acute and chronic leukemia patients were assessed as determining risk factors by using the nonparametric approach.
Conclusion: Thus, in view of clinical studies, if more than one reason for death exists, it is appropriate to use a competing risk analysis to determine probable susceptibility.Keywords: Survival analysis, statistical models,proportional hazards models, leukemiaTurkiye Klinikleri J Med Sci 2005, 25:20-28
|
|
|
|
|
