Journal of biopharmaceutical statistics
-
For a variety of reasons including poorly designed case report forms (CRFs), incomplete or invalid CRF data entries, and premature treatment or study discontinuations, missing data is a common phenomenon in controlled clinical trials. With the widely accepted use of the intent-to-treat (ITT) analysis dataset as the primary analysis dataset for the analysis of controlled clinical trial data, the presence of missing data could lead to complicated data analysis strategies and subsequently to controversy in the interpretation of trial results. In this article, we review the mechanisms of missing data and some common approaches to analyzing missing data with an emphasis on study dropouts. ⋯ Finally, we discuss the results of a comparative Monte Carlo investigation of the performance characteristics of commonly utilized statistical methods for the analysis of clinical trial data with dropouts. The methods investigated include the mixed effects model for repeated measurements (MMRM), weighted and unweighted generalized estimating equations (GEE) method for the available case data, multiple-imputation-based GEE (MI-GEE), complete case (CC) analysis of covariance (ANCOVA), and last observation carried forward (LOCF) ANCOVA. Simulation experiments for the repeated measures model with missing at random (MAR) dropout, under varying dropout rates and intrasubject correlation, show that the LOCF, ANCOVA, and weighted GEE methods perform poorly in terms of percent relative bias for estimating a difference in means effect, while the MI-GEE and weighted GEE methods both have less power for rejecting a zero difference in means hypothesis.
-
While randomized, well-controlled, clinical trials have been viewed as the gold standard in the evaluation of medical products, it is not uncommon for medical device clinical studies to depart from the paradigm of randomized trials, due to ethical or practical reasons. In nonrandomized studies, the advantages of well-designed and conducted randomized clinical trials are no longer available, and consequently the statistical inference obtained from such studies may carry a lower level of scientific assurance, compared to randomized trials. This paper provides a brief overview of nonrandomized medical device clinical studies in terms of design and statistical analysis as well as regulatory issues, including some challenges that frequently arise in those endeavors.
-
Cancer causes premature death and significant, often devastating, symptoms. While prolongation of survival is an obvious end point for new cancer drug approval, the US Food and Drug Administration (FDA) has also utilized end points that evaluate patient symptoms. In this article we discuss the end points, evidence, and analyses supporting cancer drug approvals based on evaluations of tumor-related signs and symptoms. ⋯ Drug sponsors are encouraged to include symptom assessments in cancer clinical trials and to perform further research to improve symptom-assessment methods. The FDA routinely meets with sponsors at End of Phase 2 Meetings to discuss drug development plans and the design of phase 3 trials. We encourage sponsors to request special protocol assessments (SPA) after meeting with the FDA to get written confirmation of the adequacy of plans for assessing cancer morbidity and quality of life, including protocols, end points, statistical analysis plans, and draft case report forms.