Accounting for a decaying correlation structure in cluster randomized trials with continuous recruitment

Kelsey L. Grantham, Jessica Kasza, Stephane Heritier, Karla Hemming, Andrew B. Forbes

Research output: Contribution to journalArticleResearchpeer-review

Abstract

A requirement for calculating sample sizes for cluster randomized trials (CRTs) conducted over multiple periods of time is the specification of a form for the correlation between outcomes of subjects within the same cluster, encoded via the within-cluster correlation structure. Previously proposed within-cluster correlation structures have made strong assumptions; for example, the usual assumption is that correlations between the outcomes of all pairs of subjects are identical (“uniform correlation”). More recently, structures that allow for a decay in correlation between pairs of outcomes measured in different periods have been suggested. However, these structures are overly simple in settings with continuous recruitment and measurement. We propose a more realistic “continuous-time correlation decay” structure whereby correlations between subjects' outcomes decay as the time between these subjects' measurement times increases. We investigate the use of this structure on trial planning in the context of a primary care diabetes trial, where there is evidence of decaying correlation between pairs of patients' outcomes over time. In particular, for a range of different trial designs, we derive the variance of the treatment effect estimator under continuous-time correlation decay and compare this to the variance obtained under uniform correlation. For stepped wedge and cluster randomized crossover designs, incorrectly assuming uniform correlation will underestimate the required sample size under most trial configurations likely to occur in practice. Planning of CRTs requires consideration of the most appropriate within-cluster correlation structure to obtain a suitable sample size.

Original languageEnglish
Pages (from-to)1918-1934
JournalStatistics in Medicine
Volume38
Issue number11
DOIs
Publication statusPublished - 20 May 2019

Keywords

  • clinical trial design
  • cluster randomized trial
  • crossover design
  • sample size
  • stepped wedge design
  • within-cluster correlation

Cite this

@article{e116323be50e4457b0327f8bf49593a3,
title = "Accounting for a decaying correlation structure in cluster randomized trials with continuous recruitment",
abstract = "A requirement for calculating sample sizes for cluster randomized trials (CRTs) conducted over multiple periods of time is the specification of a form for the correlation between outcomes of subjects within the same cluster, encoded via the within-cluster correlation structure. Previously proposed within-cluster correlation structures have made strong assumptions; for example, the usual assumption is that correlations between the outcomes of all pairs of subjects are identical (“uniform correlation”). More recently, structures that allow for a decay in correlation between pairs of outcomes measured in different periods have been suggested. However, these structures are overly simple in settings with continuous recruitment and measurement. We propose a more realistic “continuous-time correlation decay” structure whereby correlations between subjects' outcomes decay as the time between these subjects' measurement times increases. We investigate the use of this structure on trial planning in the context of a primary care diabetes trial, where there is evidence of decaying correlation between pairs of patients' outcomes over time. In particular, for a range of different trial designs, we derive the variance of the treatment effect estimator under continuous-time correlation decay and compare this to the variance obtained under uniform correlation. For stepped wedge and cluster randomized crossover designs, incorrectly assuming uniform correlation will underestimate the required sample size under most trial configurations likely to occur in practice. Planning of CRTs requires consideration of the most appropriate within-cluster correlation structure to obtain a suitable sample size.",
keywords = "clinical trial design, cluster randomized trial, crossover design, sample size, stepped wedge design, within-cluster correlation",
author = "Grantham, {Kelsey L.} and Jessica Kasza and Stephane Heritier and Karla Hemming and Forbes, {Andrew B.}",
year = "2019",
month = "5",
day = "20",
doi = "10.1002/sim.8089",
language = "English",
volume = "38",
pages = "1918--1934",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley & Sons",
number = "11",

}

Accounting for a decaying correlation structure in cluster randomized trials with continuous recruitment. / Grantham, Kelsey L.; Kasza, Jessica; Heritier, Stephane; Hemming, Karla; Forbes, Andrew B.

In: Statistics in Medicine, Vol. 38, No. 11, 20.05.2019, p. 1918-1934.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - Accounting for a decaying correlation structure in cluster randomized trials with continuous recruitment

AU - Grantham, Kelsey L.

AU - Kasza, Jessica

AU - Heritier, Stephane

AU - Hemming, Karla

AU - Forbes, Andrew B.

PY - 2019/5/20

Y1 - 2019/5/20

N2 - A requirement for calculating sample sizes for cluster randomized trials (CRTs) conducted over multiple periods of time is the specification of a form for the correlation between outcomes of subjects within the same cluster, encoded via the within-cluster correlation structure. Previously proposed within-cluster correlation structures have made strong assumptions; for example, the usual assumption is that correlations between the outcomes of all pairs of subjects are identical (“uniform correlation”). More recently, structures that allow for a decay in correlation between pairs of outcomes measured in different periods have been suggested. However, these structures are overly simple in settings with continuous recruitment and measurement. We propose a more realistic “continuous-time correlation decay” structure whereby correlations between subjects' outcomes decay as the time between these subjects' measurement times increases. We investigate the use of this structure on trial planning in the context of a primary care diabetes trial, where there is evidence of decaying correlation between pairs of patients' outcomes over time. In particular, for a range of different trial designs, we derive the variance of the treatment effect estimator under continuous-time correlation decay and compare this to the variance obtained under uniform correlation. For stepped wedge and cluster randomized crossover designs, incorrectly assuming uniform correlation will underestimate the required sample size under most trial configurations likely to occur in practice. Planning of CRTs requires consideration of the most appropriate within-cluster correlation structure to obtain a suitable sample size.

AB - A requirement for calculating sample sizes for cluster randomized trials (CRTs) conducted over multiple periods of time is the specification of a form for the correlation between outcomes of subjects within the same cluster, encoded via the within-cluster correlation structure. Previously proposed within-cluster correlation structures have made strong assumptions; for example, the usual assumption is that correlations between the outcomes of all pairs of subjects are identical (“uniform correlation”). More recently, structures that allow for a decay in correlation between pairs of outcomes measured in different periods have been suggested. However, these structures are overly simple in settings with continuous recruitment and measurement. We propose a more realistic “continuous-time correlation decay” structure whereby correlations between subjects' outcomes decay as the time between these subjects' measurement times increases. We investigate the use of this structure on trial planning in the context of a primary care diabetes trial, where there is evidence of decaying correlation between pairs of patients' outcomes over time. In particular, for a range of different trial designs, we derive the variance of the treatment effect estimator under continuous-time correlation decay and compare this to the variance obtained under uniform correlation. For stepped wedge and cluster randomized crossover designs, incorrectly assuming uniform correlation will underestimate the required sample size under most trial configurations likely to occur in practice. Planning of CRTs requires consideration of the most appropriate within-cluster correlation structure to obtain a suitable sample size.

KW - clinical trial design

KW - cluster randomized trial

KW - crossover design

KW - sample size

KW - stepped wedge design

KW - within-cluster correlation

UR - http://www.scopus.com/inward/record.url?scp=85060340033&partnerID=8YFLogxK

U2 - 10.1002/sim.8089

DO - 10.1002/sim.8089

M3 - Article

VL - 38

SP - 1918

EP - 1934

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 11

ER -