Inference and sample size calculation for clinical trials with incomplete observations of paired binary outcomes

Song Zhang, Jing Cao, Chul Ahn

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

We investigate the estimation of intervention effect and sample size determination for experiments where subjects are supposed to contribute paired binary outcomes with some incomplete observations. We propose a hybrid estimator to appropriately account for the mixed nature of observed data: paired outcomes from those who contribute complete pairs of observations and unpaired outcomes from those who contribute either pre-intervention or post-intervention outcomes. We theoretically prove that if incomplete data are evenly distributed between the pre-intervention and post-intervention periods, the proposed estimator will always be more efficient than the traditional estimator. A numerical research shows that when the distribution of incomplete data is unbalanced, the proposed estimator will be superior when there is moderate-to-strong positive within-subject correlation. We further derive a closed-form sample size formula to help researchers determine how many subjects need to be enrolled in such studies. Simulation results suggest that the calculated sample size maintains the empirical power and type I error under various design configurations. We demonstrate the proposed method using a real application example.

Original languageEnglish (US)
Pages (from-to)581-591
Number of pages11
JournalStatistics in Medicine
Volume36
Issue number4
DOIs
StatePublished - Feb 20 2017

Fingerprint

Sample Size Calculation
Binary Outcomes
Clinical Trials
Sample Size
Estimator
Incomplete Data
Paired Data
Sample Size Determination
Research Personnel
Effect Size
Type I error
Research
Closed-form
Observation
Configuration
Demonstrate
Experiment
Simulation

Keywords

  • binary outcomes
  • incomplete
  • paire
  • sample size

ASJC Scopus subject areas

  • Epidemiology
  • Statistics and Probability

Cite this

Inference and sample size calculation for clinical trials with incomplete observations of paired binary outcomes. / Zhang, Song; Cao, Jing; Ahn, Chul.

In: Statistics in Medicine, Vol. 36, No. 4, 20.02.2017, p. 581-591.

Research output: Contribution to journalArticle

@article{e7d53ac9a6394c44b828a6a4890e6ab0,
title = "Inference and sample size calculation for clinical trials with incomplete observations of paired binary outcomes",
abstract = "We investigate the estimation of intervention effect and sample size determination for experiments where subjects are supposed to contribute paired binary outcomes with some incomplete observations. We propose a hybrid estimator to appropriately account for the mixed nature of observed data: paired outcomes from those who contribute complete pairs of observations and unpaired outcomes from those who contribute either pre-intervention or post-intervention outcomes. We theoretically prove that if incomplete data are evenly distributed between the pre-intervention and post-intervention periods, the proposed estimator will always be more efficient than the traditional estimator. A numerical research shows that when the distribution of incomplete data is unbalanced, the proposed estimator will be superior when there is moderate-to-strong positive within-subject correlation. We further derive a closed-form sample size formula to help researchers determine how many subjects need to be enrolled in such studies. Simulation results suggest that the calculated sample size maintains the empirical power and type I error under various design configurations. We demonstrate the proposed method using a real application example.",
keywords = "binary outcomes, incomplete, paire, sample size",
author = "Song Zhang and Jing Cao and Chul Ahn",
year = "2017",
month = "2",
day = "20",
doi = "10.1002/sim.7168",
language = "English (US)",
volume = "36",
pages = "581--591",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",
number = "4",

}

TY - JOUR

T1 - Inference and sample size calculation for clinical trials with incomplete observations of paired binary outcomes

AU - Zhang, Song

AU - Cao, Jing

AU - Ahn, Chul

PY - 2017/2/20

Y1 - 2017/2/20

N2 - We investigate the estimation of intervention effect and sample size determination for experiments where subjects are supposed to contribute paired binary outcomes with some incomplete observations. We propose a hybrid estimator to appropriately account for the mixed nature of observed data: paired outcomes from those who contribute complete pairs of observations and unpaired outcomes from those who contribute either pre-intervention or post-intervention outcomes. We theoretically prove that if incomplete data are evenly distributed between the pre-intervention and post-intervention periods, the proposed estimator will always be more efficient than the traditional estimator. A numerical research shows that when the distribution of incomplete data is unbalanced, the proposed estimator will be superior when there is moderate-to-strong positive within-subject correlation. We further derive a closed-form sample size formula to help researchers determine how many subjects need to be enrolled in such studies. Simulation results suggest that the calculated sample size maintains the empirical power and type I error under various design configurations. We demonstrate the proposed method using a real application example.

AB - We investigate the estimation of intervention effect and sample size determination for experiments where subjects are supposed to contribute paired binary outcomes with some incomplete observations. We propose a hybrid estimator to appropriately account for the mixed nature of observed data: paired outcomes from those who contribute complete pairs of observations and unpaired outcomes from those who contribute either pre-intervention or post-intervention outcomes. We theoretically prove that if incomplete data are evenly distributed between the pre-intervention and post-intervention periods, the proposed estimator will always be more efficient than the traditional estimator. A numerical research shows that when the distribution of incomplete data is unbalanced, the proposed estimator will be superior when there is moderate-to-strong positive within-subject correlation. We further derive a closed-form sample size formula to help researchers determine how many subjects need to be enrolled in such studies. Simulation results suggest that the calculated sample size maintains the empirical power and type I error under various design configurations. We demonstrate the proposed method using a real application example.

KW - binary outcomes

KW - incomplete

KW - paire

KW - sample size

UR - http://www.scopus.com/inward/record.url?scp=85009134626&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85009134626&partnerID=8YFLogxK

U2 - 10.1002/sim.7168

DO - 10.1002/sim.7168

M3 - Article

VL - 36

SP - 581

EP - 591

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 4

ER -