Analysis of repeated pregnancy outcomes

Germaine Buck Louis, Vanja Dukic, Patrick J. Heagerty, Thomas Louis, Courtney D. Lynch, Louise M. Ryan, Enrique F. Schisterman, Ann Trumble, Mark Klebanoff, Aiyi Liu, Kai Yu, James Collins, Geary Olsen

Research output: Contribution to journalArticle

Abstract

Women tend to repeat reproductive outcomes, with past history of an adverse outcome being associated with an approximate two-fold increase in subsequent risk. These observations support the need for statistical designs and analyses that address this clustering. Failure to do so may mask effects, result in inaccurate variance estimators, produce biased or inefficient estimates of exposure effects. We review and evaluate basic analytic approaches for analysing reproductive outcomes, including ignoring reproductive history, treating it as a covariate or avoiding the clustering problem by analysing only one pregnancy per woman, and contrast these to more modern approaches such as generalized estimating equations with robust standard errors and mixed models with various correlation structures. We illustrate the issues by analysing a sample from the Collaborative Perinatal Project dataset, demonstrating how the statistical model impacts summary statistics and inferences when assessing etiologic determinants of birth weight.

Original languageEnglish (US)
Pages (from-to)103-126
Number of pages24
JournalStatistical Methods in Medical Research
Volume15
Issue number2
DOIs
StatePublished - Apr 2006

Fingerprint

Pregnancy
Pregnancy Outcome
Cluster Analysis
Reproductive History
Statistical Models
Masks
Birth Weight
Clustering
Generalized Estimating Equations
Error Model
Variance Estimator
Correlation Structure
Mixed Model
Standard error
Inaccurate
Statistical Model
Mask
Biased
Standard Model
Covariates

ASJC Scopus subject areas

  • Epidemiology
  • Health Information Management
  • Nursing(all)

Cite this

Louis, G. B., Dukic, V., Heagerty, P. J., Louis, T., Lynch, C. D., Ryan, L. M., ... Olsen, G. (2006). Analysis of repeated pregnancy outcomes. Statistical Methods in Medical Research, 15(2), 103-126. https://doi.org/10.1191/0962280206sm434oa

Analysis of repeated pregnancy outcomes. / Louis, Germaine Buck; Dukic, Vanja; Heagerty, Patrick J.; Louis, Thomas; Lynch, Courtney D.; Ryan, Louise M.; Schisterman, Enrique F.; Trumble, Ann; Klebanoff, Mark; Liu, Aiyi; Yu, Kai; Collins, James; Olsen, Geary.

In: Statistical Methods in Medical Research, Vol. 15, No. 2, 04.2006, p. 103-126.

Research output: Contribution to journalArticle

Louis, GB, Dukic, V, Heagerty, PJ, Louis, T, Lynch, CD, Ryan, LM, Schisterman, EF, Trumble, A, Klebanoff, M, Liu, A, Yu, K, Collins, J & Olsen, G 2006, 'Analysis of repeated pregnancy outcomes', Statistical Methods in Medical Research, vol. 15, no. 2, pp. 103-126. https://doi.org/10.1191/0962280206sm434oa
Louis, Germaine Buck ; Dukic, Vanja ; Heagerty, Patrick J. ; Louis, Thomas ; Lynch, Courtney D. ; Ryan, Louise M. ; Schisterman, Enrique F. ; Trumble, Ann ; Klebanoff, Mark ; Liu, Aiyi ; Yu, Kai ; Collins, James ; Olsen, Geary. / Analysis of repeated pregnancy outcomes. In: Statistical Methods in Medical Research. 2006 ; Vol. 15, No. 2. pp. 103-126.
@article{42cebdcdb339414c93fc0de8bd4ca5dd,
title = "Analysis of repeated pregnancy outcomes",
abstract = "Women tend to repeat reproductive outcomes, with past history of an adverse outcome being associated with an approximate two-fold increase in subsequent risk. These observations support the need for statistical designs and analyses that address this clustering. Failure to do so may mask effects, result in inaccurate variance estimators, produce biased or inefficient estimates of exposure effects. We review and evaluate basic analytic approaches for analysing reproductive outcomes, including ignoring reproductive history, treating it as a covariate or avoiding the clustering problem by analysing only one pregnancy per woman, and contrast these to more modern approaches such as generalized estimating equations with robust standard errors and mixed models with various correlation structures. We illustrate the issues by analysing a sample from the Collaborative Perinatal Project dataset, demonstrating how the statistical model impacts summary statistics and inferences when assessing etiologic determinants of birth weight.",
author = "Louis, {Germaine Buck} and Vanja Dukic and Heagerty, {Patrick J.} and Thomas Louis and Lynch, {Courtney D.} and Ryan, {Louise M.} and Schisterman, {Enrique F.} and Ann Trumble and Mark Klebanoff and Aiyi Liu and Kai Yu and James Collins and Geary Olsen",
year = "2006",
month = "4",
doi = "10.1191/0962280206sm434oa",
language = "English (US)",
volume = "15",
pages = "103--126",
journal = "Statistical Methods in Medical Research",
issn = "0962-2802",
publisher = "SAGE Publications Ltd",
number = "2",

}

TY - JOUR

T1 - Analysis of repeated pregnancy outcomes

AU - Louis, Germaine Buck

AU - Dukic, Vanja

AU - Heagerty, Patrick J.

AU - Louis, Thomas

AU - Lynch, Courtney D.

AU - Ryan, Louise M.

AU - Schisterman, Enrique F.

AU - Trumble, Ann

AU - Klebanoff, Mark

AU - Liu, Aiyi

AU - Yu, Kai

AU - Collins, James

AU - Olsen, Geary

PY - 2006/4

Y1 - 2006/4

N2 - Women tend to repeat reproductive outcomes, with past history of an adverse outcome being associated with an approximate two-fold increase in subsequent risk. These observations support the need for statistical designs and analyses that address this clustering. Failure to do so may mask effects, result in inaccurate variance estimators, produce biased or inefficient estimates of exposure effects. We review and evaluate basic analytic approaches for analysing reproductive outcomes, including ignoring reproductive history, treating it as a covariate or avoiding the clustering problem by analysing only one pregnancy per woman, and contrast these to more modern approaches such as generalized estimating equations with robust standard errors and mixed models with various correlation structures. We illustrate the issues by analysing a sample from the Collaborative Perinatal Project dataset, demonstrating how the statistical model impacts summary statistics and inferences when assessing etiologic determinants of birth weight.

AB - Women tend to repeat reproductive outcomes, with past history of an adverse outcome being associated with an approximate two-fold increase in subsequent risk. These observations support the need for statistical designs and analyses that address this clustering. Failure to do so may mask effects, result in inaccurate variance estimators, produce biased or inefficient estimates of exposure effects. We review and evaluate basic analytic approaches for analysing reproductive outcomes, including ignoring reproductive history, treating it as a covariate or avoiding the clustering problem by analysing only one pregnancy per woman, and contrast these to more modern approaches such as generalized estimating equations with robust standard errors and mixed models with various correlation structures. We illustrate the issues by analysing a sample from the Collaborative Perinatal Project dataset, demonstrating how the statistical model impacts summary statistics and inferences when assessing etiologic determinants of birth weight.

UR - http://www.scopus.com/inward/record.url?scp=33645744332&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33645744332&partnerID=8YFLogxK

U2 - 10.1191/0962280206sm434oa

DO - 10.1191/0962280206sm434oa

M3 - Article

C2 - 16615652

AN - SCOPUS:33645744332

VL - 15

SP - 103

EP - 126

JO - Statistical Methods in Medical Research

JF - Statistical Methods in Medical Research

SN - 0962-2802

IS - 2

ER -