Technique for Survival Plots Includes Individual Data
May complement Kaplan-Meier plots
WEDNESDAY, Jan. 9 (HealthDay News) -- A new technique to plot survival times for individuals in treatment groups shows more overlap than Kaplan-Meier plots and may complement them, according to a report in the Jan. 16 issue of the Journal of the National Cancer Institute.
Patrick Royston, from the Medical Research Council Clinical Trials Unit in London, United Kingdom, and colleagues developed a technique to produce scatter plots with survival data from individual patients rather than each treatment group, as is done for a Kaplan-Meier plot. They then applied the technique to a randomized trial of 347 patients with renal cancer, where 25 patients for whom data was available had been censored (i.e., the observation period ended before the event occurred).
The researchers found that the median survival was 3.0 months longer for patients receiving interferon compared with medroxyprogesterone acetate. While a Kaplan-Meier plot of the data showed clear separation between treatment groups and prognostic groups, their comparison of individual observed and inputted survival times between groups of patients overlapped.
The authors "have persuasively pointed out some important limitations of Kaplan-Meier curves and have proposed a method of presenting mortality data under limited censoring," Janet Wittes, Ph.D., from Statistics Collaborative in Washington, D.C., writes in an accompanying editorial.