Person-time analysis of paired community intervention trials when the number of communities is small

Ron Brookmeyer, Ying Qing Chen

Research output: Contribution to journalArticle

Abstract

Community intervention trials involve randomization of communities to either an intervention or control arm. The objective of this paper is to evaluate person-time methods of analysis of paired community intervention trials when the number of community pairs is small. We consider several test procedures and evaluate their performance by simulation. Naive methods that ignore intracluster correlation, such as standard Mantel-Haenszel type statistics, can be misleading. The performance of the paired t-test depends on the distribution of the random community effects. Permutation tests perform well for the ranges of situations considered. However, there can be considerable loss of power with permutation methods compared to standard Mantel-Haenszel methods if in fact there is no intracluster correlation when the number of pairs is small. We consider methods to account for individual level covariates. Motivation for this work came from recent randomized community intervention trials in Africa to prevent transmission of the human immunodeficiency virus (HIV).

Original languageEnglish (US)
Pages (from-to)2121-2132
Number of pages12
JournalStatistics in Medicine
Volume17
Issue number18
DOIs
StatePublished - Sep 30 1998

Fingerprint

Person
Intracluster Correlation
Permutation Test
Evaluate
t-test
Random Allocation
Randomisation
Virus
Community
Covariates
Permutation
HIV
Statistics
Range of data
Simulation
Standards

ASJC Scopus subject areas

  • Epidemiology

Cite this

Person-time analysis of paired community intervention trials when the number of communities is small. / Brookmeyer, Ron; Chen, Ying Qing.

In: Statistics in Medicine, Vol. 17, No. 18, 30.09.1998, p. 2121-2132.

Research output: Contribution to journalArticle

Brookmeyer, Ron ; Chen, Ying Qing. / Person-time analysis of paired community intervention trials when the number of communities is small. In: Statistics in Medicine. 1998 ; Vol. 17, No. 18. pp. 2121-2132.
@article{d3d9f860d26c4fc487f10bf03145b7d8,
title = "Person-time analysis of paired community intervention trials when the number of communities is small",
abstract = "Community intervention trials involve randomization of communities to either an intervention or control arm. The objective of this paper is to evaluate person-time methods of analysis of paired community intervention trials when the number of community pairs is small. We consider several test procedures and evaluate their performance by simulation. Naive methods that ignore intracluster correlation, such as standard Mantel-Haenszel type statistics, can be misleading. The performance of the paired t-test depends on the distribution of the random community effects. Permutation tests perform well for the ranges of situations considered. However, there can be considerable loss of power with permutation methods compared to standard Mantel-Haenszel methods if in fact there is no intracluster correlation when the number of pairs is small. We consider methods to account for individual level covariates. Motivation for this work came from recent randomized community intervention trials in Africa to prevent transmission of the human immunodeficiency virus (HIV).",
author = "Ron Brookmeyer and Chen, {Ying Qing}",
year = "1998",
month = "9",
day = "30",
doi = "10.1002/(SICI)1097-0258(19980930)17:18<2121::AID-SIM907>3.0.CO;2-S",
language = "English (US)",
volume = "17",
pages = "2121--2132",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",
number = "18",

}

TY - JOUR

T1 - Person-time analysis of paired community intervention trials when the number of communities is small

AU - Brookmeyer, Ron

AU - Chen, Ying Qing

PY - 1998/9/30

Y1 - 1998/9/30

N2 - Community intervention trials involve randomization of communities to either an intervention or control arm. The objective of this paper is to evaluate person-time methods of analysis of paired community intervention trials when the number of community pairs is small. We consider several test procedures and evaluate their performance by simulation. Naive methods that ignore intracluster correlation, such as standard Mantel-Haenszel type statistics, can be misleading. The performance of the paired t-test depends on the distribution of the random community effects. Permutation tests perform well for the ranges of situations considered. However, there can be considerable loss of power with permutation methods compared to standard Mantel-Haenszel methods if in fact there is no intracluster correlation when the number of pairs is small. We consider methods to account for individual level covariates. Motivation for this work came from recent randomized community intervention trials in Africa to prevent transmission of the human immunodeficiency virus (HIV).

AB - Community intervention trials involve randomization of communities to either an intervention or control arm. The objective of this paper is to evaluate person-time methods of analysis of paired community intervention trials when the number of community pairs is small. We consider several test procedures and evaluate their performance by simulation. Naive methods that ignore intracluster correlation, such as standard Mantel-Haenszel type statistics, can be misleading. The performance of the paired t-test depends on the distribution of the random community effects. Permutation tests perform well for the ranges of situations considered. However, there can be considerable loss of power with permutation methods compared to standard Mantel-Haenszel methods if in fact there is no intracluster correlation when the number of pairs is small. We consider methods to account for individual level covariates. Motivation for this work came from recent randomized community intervention trials in Africa to prevent transmission of the human immunodeficiency virus (HIV).

UR - http://www.scopus.com/inward/record.url?scp=0032581930&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0032581930&partnerID=8YFLogxK

U2 - 10.1002/(SICI)1097-0258(19980930)17:18<2121::AID-SIM907>3.0.CO;2-S

DO - 10.1002/(SICI)1097-0258(19980930)17:18<2121::AID-SIM907>3.0.CO;2-S

M3 - Article

C2 - 9789918

AN - SCOPUS:0032581930

VL - 17

SP - 2121

EP - 2132

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 18

ER -