محاسبه حجم نمونه در مطالعات بقا و كاربرد آن در ارزيابي بيماري لنفوم غير هوچكين

Calculation of Sample Size in Survival Studies and its Application in the Assessment of Non-Hodgkin Lymphoma

In non-random studies , calculation of sample size for failure time (survival) random variables is usually done on the basis of unadjusted log rank test . Supposing that a certain variable is independent from other variables of the patient , we developed the methods sugyested by lipstiz and Parzan to show that the power obtained from unadjusted log rank test for the event time random variable overestimated the real power when that variable correlated with other variables. Using exponential regression and the sample size obtained from the formula , we modelled the hazard continunm for both censored and uncensored data . We then used the suggested methods with the collected data for the prognosis of influential factors in non-hodgkin lymphoma .