Accounting for twin births in sample size calculations for randomised trials

Lisa Yelland, Thomas Sullivan, Carmel Collins, David J. Price, Andrew McPhee, Katherine J. Lee

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Background: Including twins in randomised trials leads to non-independence or clustering in the data. Clustering has important implications for sample size calculations, yet few trials take this into account. Estimates of the intracluster correlation coefficient (ICC), or the correlation between outcomes of twins, are needed to assist with sample size planning. Our aims were to provide ICC estimates for infant outcomes, describe the information that must be specified in order to account for clustering due to twins in sample size calculations, and develop a simple tool for performing sample size calculations for trials including twins. Methods: ICCs were estimated for infant outcomes collected in four randomised trials that included twins. The information required to account for clustering due to twins in sample size calculations is described. A tool that calculates the sample size based on this information was developed in Microsoft Excel and in R as a Shiny web app. Results: ICC estimates ranged between -0.12, indicating a weak negative relationship, and 0.98, indicating a strong positive relationship between outcomes of twins. Example calculations illustrate how the ICC estimates and sample size calculator can be used to determine the target sample size for trials including twins. Conclusions: Clustering among outcomes measured on twins should be taken into account in sample size calculations to obtain the desired power. Our ICC estimates and sample size calculator will be useful for designing future trials that include twins. Publication of additional ICCs is needed to further assist with sample size planning for future trials.

LanguageEnglish
Pages380-387
Number of pages8
JournalPaediatric and Perinatal Epidemiology
Volume32
Issue number4
DOIs
Publication statusPublished - 1 Jul 2018

Keywords

  • generalised estimating equations
  • intracluster correlation
  • multiple birth
  • power
  • statistical methodology

ASJC Scopus subject areas

  • Epidemiology
  • Pediatrics, Perinatology, and Child Health

Cite this

@article{7ee2dda66f234685aed9967b814880cd,
title = "Accounting for twin births in sample size calculations for randomised trials",
abstract = "Background: Including twins in randomised trials leads to non-independence or clustering in the data. Clustering has important implications for sample size calculations, yet few trials take this into account. Estimates of the intracluster correlation coefficient (ICC), or the correlation between outcomes of twins, are needed to assist with sample size planning. Our aims were to provide ICC estimates for infant outcomes, describe the information that must be specified in order to account for clustering due to twins in sample size calculations, and develop a simple tool for performing sample size calculations for trials including twins. Methods: ICCs were estimated for infant outcomes collected in four randomised trials that included twins. The information required to account for clustering due to twins in sample size calculations is described. A tool that calculates the sample size based on this information was developed in Microsoft Excel and in R as a Shiny web app. Results: ICC estimates ranged between -0.12, indicating a weak negative relationship, and 0.98, indicating a strong positive relationship between outcomes of twins. Example calculations illustrate how the ICC estimates and sample size calculator can be used to determine the target sample size for trials including twins. Conclusions: Clustering among outcomes measured on twins should be taken into account in sample size calculations to obtain the desired power. Our ICC estimates and sample size calculator will be useful for designing future trials that include twins. Publication of additional ICCs is needed to further assist with sample size planning for future trials.",
keywords = "generalised estimating equations, intracluster correlation, multiple birth, power, statistical methodology",
author = "Lisa Yelland and Thomas Sullivan and Carmel Collins and Price, {David J.} and Andrew McPhee and Lee, {Katherine J.}",
year = "2018",
month = "7",
day = "1",
doi = "10.1111/ppe.12471",
language = "English",
volume = "32",
pages = "380--387",
journal = "Paediatric and Perinatal Epidemiology",
issn = "0269-5022",
publisher = "Wiley-Blackwell",
number = "4",

}

Accounting for twin births in sample size calculations for randomised trials. / Yelland, Lisa; Sullivan, Thomas; Collins, Carmel; Price, David J.; McPhee, Andrew; Lee, Katherine J.

In: Paediatric and Perinatal Epidemiology, Vol. 32, No. 4, 01.07.2018, p. 380-387.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Accounting for twin births in sample size calculations for randomised trials

AU - Yelland, Lisa

AU - Sullivan, Thomas

AU - Collins, Carmel

AU - Price, David J.

AU - McPhee, Andrew

AU - Lee, Katherine J.

PY - 2018/7/1

Y1 - 2018/7/1

N2 - Background: Including twins in randomised trials leads to non-independence or clustering in the data. Clustering has important implications for sample size calculations, yet few trials take this into account. Estimates of the intracluster correlation coefficient (ICC), or the correlation between outcomes of twins, are needed to assist with sample size planning. Our aims were to provide ICC estimates for infant outcomes, describe the information that must be specified in order to account for clustering due to twins in sample size calculations, and develop a simple tool for performing sample size calculations for trials including twins. Methods: ICCs were estimated for infant outcomes collected in four randomised trials that included twins. The information required to account for clustering due to twins in sample size calculations is described. A tool that calculates the sample size based on this information was developed in Microsoft Excel and in R as a Shiny web app. Results: ICC estimates ranged between -0.12, indicating a weak negative relationship, and 0.98, indicating a strong positive relationship between outcomes of twins. Example calculations illustrate how the ICC estimates and sample size calculator can be used to determine the target sample size for trials including twins. Conclusions: Clustering among outcomes measured on twins should be taken into account in sample size calculations to obtain the desired power. Our ICC estimates and sample size calculator will be useful for designing future trials that include twins. Publication of additional ICCs is needed to further assist with sample size planning for future trials.

AB - Background: Including twins in randomised trials leads to non-independence or clustering in the data. Clustering has important implications for sample size calculations, yet few trials take this into account. Estimates of the intracluster correlation coefficient (ICC), or the correlation between outcomes of twins, are needed to assist with sample size planning. Our aims were to provide ICC estimates for infant outcomes, describe the information that must be specified in order to account for clustering due to twins in sample size calculations, and develop a simple tool for performing sample size calculations for trials including twins. Methods: ICCs were estimated for infant outcomes collected in four randomised trials that included twins. The information required to account for clustering due to twins in sample size calculations is described. A tool that calculates the sample size based on this information was developed in Microsoft Excel and in R as a Shiny web app. Results: ICC estimates ranged between -0.12, indicating a weak negative relationship, and 0.98, indicating a strong positive relationship between outcomes of twins. Example calculations illustrate how the ICC estimates and sample size calculator can be used to determine the target sample size for trials including twins. Conclusions: Clustering among outcomes measured on twins should be taken into account in sample size calculations to obtain the desired power. Our ICC estimates and sample size calculator will be useful for designing future trials that include twins. Publication of additional ICCs is needed to further assist with sample size planning for future trials.

KW - generalised estimating equations

KW - intracluster correlation

KW - multiple birth

KW - power

KW - statistical methodology

UR - http://www.scopus.com/inward/record.url?scp=85046438984&partnerID=8YFLogxK

U2 - 10.1111/ppe.12471

DO - 10.1111/ppe.12471

M3 - Article

VL - 32

SP - 380

EP - 387

JO - Paediatric and Perinatal Epidemiology

T2 - Paediatric and Perinatal Epidemiology

JF - Paediatric and Perinatal Epidemiology

SN - 0269-5022

IS - 4

ER -