A Likelihood-Based, Counterfactual Approach to Accounting for Treatment Failures in Clinical Trials

• Published in: Journal of biopharmaceutical statistics Vol. 13; no. 3; pp. 481 - 494
• Main Authors: Shaffer, Michele L., Chinchilli, Vernon M.
• Format: Journal Article
• Language: English
• Published: England Taylor & Francis Group 01.08.2003
• Subjects: Bias; Clinical Trials as Topic - statistics & numerical data; Emergency medication; Expectation maximization (EM) algorithm; Intent-to-treat; Likelihood Functions; Non-compliance; Treatment Failure
• ISSN: 1054-3406; 1520-5711
• DOI: 10.1081/BIP-120022768

Consider a two-armed, placebo-controlled trial in which subjects may experience treatment failure. For ethical reasons, it is necessary to administer emergency or rescue medications for such subjects. However, the emergency medications may bias the set of response measurements. When analyzing the data from clinical trials, the standard approach is to perform an intent-to-treat (ITT) analysis, wherein the data are analyzed according to treatment assignment. Secondary statistical analyses that supplement the ITT analysis can be performed to account for the impact of treatment failures and emergency medications. A likelihood-based, counterfactual approach to supplemental analyses uses the expectation maximization (EM) algorithm for parameter estimation and a likelihood ratio test to test the equality of the placebo and experimental treatment means for subjects who would not fail under either treatment assignment. A simulation study is performed to assess the operating characteristics of the likelihood ratio test. An example from the Asthma Clinical Research Network (ACRN) is used to draw comparisons between the standard ITT procedure and the developed supplemental analysis.