Abstract:
|
The Kaplan-Meier (KM) curve is the most commonly reported summary of a survival analysis. Often, it is the only information available about the lifetime data underlying the curve, and actual individual patient data (IPD) is not readily available. For secondary analysis and meta-analysis, this is a major hindrance. This work is motivated by the need to provide a pooled estimate of survival rates (rather than hazard ratios) across several studies of end stage renal disease (ESRD) among patients with systematic lupus erythematosis (SLE) or lupus nephritis. We extend and simplify the methods proposed by Guyot, et al (2012), requiring only a Kaplan-Meier curve and the number of persons at risk at the start of the study to infer the numbers of events and censored patients, and simulate IPD conforming to the survival experience represented by the Kaplan-Meier curve. We find that KM curves computed from data simulated using our method conform very accurately to the source KM curve. Our method provides a way to leverage published KM curves to obtain reasonable simulated IPD for secondary and pooled analyses.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.