A Meta-Analysis of Missing Data and Non-Compliance Data in Clinical Endpoint Bioequivalence Studies. (2nd July 2016)

• Record Type: Journal Article
• Authors: Sun, Wanjie; Zhou, Lingjie; Grosser, Stella; Kim, Carol
• Is Part Of: Statistics in biopharmaceutical research. Volume 8:Number 3(2016)
• Page Start: 334
• Page End: 344
• Publication Date: 2016-07-02
• Subjects: Bioequivalence -- Meta-analysis -- Missing data -- Non-compliance data; Pharmacy -- Statistical methods -- Periodicals; Pharmaceutical biotechnology -- Statistical methods -- Periodicals; Biopharmaceutics -- Periodicals; Biometry -- Periodicals; Pharmacy -- Statistical methods; Periodicals; 615.190727
• DOI: 10.1080/19466315.2016.1201000
• Languages: English
• ISSNs: 1946-6315

Abstract

ABSTRACT: Missing data and noncompliance data questions are especially important in evaluating locally acting generic drugs because primary equivalence analyses in clinical endpoint bioequivalence (BE) studies are based on the per-protocol (PP) population (generally, completers and compliers). A meta-analysis using six clinical endpoint BE studies for topical drugs reveals the following: (1) An average of 22% (95% CI: 15–29%) of randomized subjects are excluded from the PP population. (2) Of these excluded subjects, half (10.6%, 95% CI: 8.3–12.8%) dropped out. Most who dropped out (6.9%, 95% CI: 5.0–8.8%) did not specify reasons. (3) Noncompliance categories include out-of-window visits (7.7%, 95% CI: 5.5%–9.8%), dosing noncompliance (<75% or >125% of dose) (5%, 95% CI: 2.7%–7.4%), and restricted medication use (3.2%, 95% CI: 1.8%–4.7%). (4) Drop out and noncompliance are not completely at random: a better treatment effect is associated with less drop out and less noncompliance. (5) Drop out and noncompliance are correlated: noncompliers are more likely to drop out, and vice versa. These results will help regulators better understand the extent and pattern of drop out and noncompliance and shed light on designing appropriate analysis population, endpoints, estimands, and investigating primary and sensitivity methods for equivalence in clinical endpoint BE studies in presence of missing and noncompliance data.