quiz of the day You are reviewing a manuscript with the results of an observational study that looked at survival for several groups. The statistical methods section states stratified logrank tests were performed to compare overall survival between groups. The figures include several Kaplan-Meier curves and reports the p-values from the logrank tests. The manuscript does not mention any other statistical methods. What other statistical analysis method could you recommend be used before making conclusions about the study results? a. The answer is always the same regardless of the statistical analysis method. Kaplan Meier, any type of Cox regression, logistic regression, it does not matter which type of analysis is used to evaluate the time-to-event or survival data b. The Cox proportional hazards model is the best method to use for time-to-event analyses. You can always assume proportional hazards c. Cox models can be used to analyze the study data. Cox models can be used to make inference about continuous and categorical covariates such as age and gender in time-to-event models. For consistent unbiased estimates the Cox model requires independent censoring, either directly or given the covariates in the model. The proportional hazards assumption should be evaluated because it may not be true. There are Cox models that do not assume proportional hazards. Also, depending on the study design there are other statistical analysis methods that may be appropriate
Posted on: Mon, 05 May 2014 16:33:43 +0000
Trending Topics
Recently Viewed Topics
© 2015