Recommendations for the Primary Analysis of Continuous Endpoints in Longitudinal Clinical Trials
- PDF / 17,641,463 Bytes
- 17 Pages / 612 x 792 pts (letter) Page_size
- 66 Downloads / 232 Views
303
Recommendations for the Primary Analysis of Continuous Endpoints in Longitudinal Clinical Trials Craig H. Mollinckrodt, PhD Research Advisor. Lilly Research Laboratories, Eli Lilly and Company, Indianapolis. Indiana
Peter W. lane, MA, CStat Director of Consultancy and Training. Research Statistics Unit. GlaxoSmithKline. Harlow, United Kingdom
Dan Scknell, PhD Section Head. Pharmaceutical Statistics. Procter 6 Gamble Pharmaceuticals. Mason. Ohio
Yahong Ping, PkD Senior Biometrician. Clinical Biostatistics. Merck Research Lab, Upper Gwynedd, Pennsylvania
James P. Mancuso, PkD Associate Director. Statistics. Pfizer Inc. Groton, Connecticut
This position paper summarizes relevant theory and current practice regarding the analysis of longitudinal clinical trials intended to support regulatory approval of medicinal products, and it reviews published research regarding methods for handling missing data. It is one strand of the PhRMA initiative to improve efficiency of late-stage clinical research and gives recommendationsfrom a cross-industry team. We concentrate specifically on continuous response measures analyzed using a linear model, when the goal is to estimate and test treatment differences at a given time point. Traditionally,the primary analysis of such trials handled missing data by simple imputation using the last, or baseline, observation cam'ed forward method (LOCF,B O G ) followed by analysis of (co)variance at the chosen time point. However, thegeneral statistical and scientific community has moved away from these
Key Words Missing data; Longitudinal data; Primary analysis; Clinical trials
Corrrspondence Address Craig Mallinckrodt, Eli Lilly and Company, Lilly Corporate Center. Indianapolis. IN 46285 (email: [email protected]).
INTRODUCTION In longitudinal trials, efficacy is often assessed in terms of treatment differences at a specific time point, usually the last time at which observations are planned while patients are under treatment. A major difficulty in analyses of such trials is missing data at the chosen time point, often due to patients withdrawing (or dropping out) from treatment. Inference from the results of a trial can be complicated by the method used to handle the missing data because the inference may depend on the method and its assumptions. Historically, the simple imputation method, called last observation carried forward (LOCF), has been used for the primary efficacy analysis of clinical trials supporting registration of new medicines (1).This approach is simple to carry out and is generally regarded as conservative in
simple methods in favor of joint analysis of data from all time points based on a multivariate model (eg. of a mixed-effects type). One such newer method, a likelihood-based mixedefiects model repeated measures (MMRM)approach, has received considerableattention in the clinical trials literature. We discuss specific concerns raised by regulatory agencies with regard to MMRM and review published evidence comparing LOCF and MMRM in terms of validi9, bias, powe
Data Loading...
