Who really gets strep sore throat? Confounding and effect modification of a time-varying exposure on recurrent events

Dean Follmann, Chiung Yu Huang, Erin Gabriel

Research output: Contribution to journalArticle

Abstract

Unmeasured confounding is the fundamental obstacle to drawing causal conclusions about the impact of an intervention from observational data. Typically, covariates are measured to eliminate or ameliorate confounding, but they may be insufficient or unavailable. In the special setting where a transient intervention or exposure varies over time within each individual and confounding is time constant, a different tack is possible. The key idea is to condition on either the overall outcome or the proportion of time in the intervention. These measures can eliminate the unmeasured confounding either by conditioning or by use of a proxy covariate. We evaluate existing methods and develop new models from which causal conclusions can be drawn from such observational data even if no baseline covariates are measured. Our motivation for this work was to determine the causal effect of Streptococcus bacteria in the throat on pharyngitis (sore throat) in Indian schoolchildren. Using our models, we show that existing methods can be badly biased and that sick children who are rarely colonized have a high probability that the Streptococcus bacteria are causing their disease. Published 2016. This article is a U.S. Government work and is in the public domain in the USA Published 2016. This article is a U.S. Government work and is in the public domain in the USA.

Original languageEnglish (US)
JournalStatistics in Medicine
DOIs
StateAccepted/In press - 2016

Fingerprint

Recurrent Events
Pharyngitis
Confounding
Time-varying
Public Sector
Streptococcus
Covariates
Bacteria
Eliminate
Proxy
Pharynx
Causal Effect
Time Constant
Conditioning
Biased
Baseline
Proportion
Vary
Evaluate
Model

Keywords

  • Attributable fraction
  • Case-crossover design
  • Pattern mixture models
  • Probability of causation
  • Self-controlled case series

ASJC Scopus subject areas

  • Epidemiology
  • Statistics and Probability

Cite this

Who really gets strep sore throat? Confounding and effect modification of a time-varying exposure on recurrent events. / Follmann, Dean; Huang, Chiung Yu; Gabriel, Erin.

In: Statistics in Medicine, 2016.

Research output: Contribution to journalArticle

@article{55d91b11f667485491e8a150471b36dd,
title = "Who really gets strep sore throat? Confounding and effect modification of a time-varying exposure on recurrent events",
abstract = "Unmeasured confounding is the fundamental obstacle to drawing causal conclusions about the impact of an intervention from observational data. Typically, covariates are measured to eliminate or ameliorate confounding, but they may be insufficient or unavailable. In the special setting where a transient intervention or exposure varies over time within each individual and confounding is time constant, a different tack is possible. The key idea is to condition on either the overall outcome or the proportion of time in the intervention. These measures can eliminate the unmeasured confounding either by conditioning or by use of a proxy covariate. We evaluate existing methods and develop new models from which causal conclusions can be drawn from such observational data even if no baseline covariates are measured. Our motivation for this work was to determine the causal effect of Streptococcus bacteria in the throat on pharyngitis (sore throat) in Indian schoolchildren. Using our models, we show that existing methods can be badly biased and that sick children who are rarely colonized have a high probability that the Streptococcus bacteria are causing their disease. Published 2016. This article is a U.S. Government work and is in the public domain in the USA Published 2016. This article is a U.S. Government work and is in the public domain in the USA.",
keywords = "Attributable fraction, Case-crossover design, Pattern mixture models, Probability of causation, Self-controlled case series",
author = "Dean Follmann and Huang, {Chiung Yu} and Erin Gabriel",
year = "2016",
doi = "10.1002/sim.7000",
language = "English (US)",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",

}

TY - JOUR

T1 - Who really gets strep sore throat? Confounding and effect modification of a time-varying exposure on recurrent events

AU - Follmann, Dean

AU - Huang, Chiung Yu

AU - Gabriel, Erin

PY - 2016

Y1 - 2016

N2 - Unmeasured confounding is the fundamental obstacle to drawing causal conclusions about the impact of an intervention from observational data. Typically, covariates are measured to eliminate or ameliorate confounding, but they may be insufficient or unavailable. In the special setting where a transient intervention or exposure varies over time within each individual and confounding is time constant, a different tack is possible. The key idea is to condition on either the overall outcome or the proportion of time in the intervention. These measures can eliminate the unmeasured confounding either by conditioning or by use of a proxy covariate. We evaluate existing methods and develop new models from which causal conclusions can be drawn from such observational data even if no baseline covariates are measured. Our motivation for this work was to determine the causal effect of Streptococcus bacteria in the throat on pharyngitis (sore throat) in Indian schoolchildren. Using our models, we show that existing methods can be badly biased and that sick children who are rarely colonized have a high probability that the Streptococcus bacteria are causing their disease. Published 2016. This article is a U.S. Government work and is in the public domain in the USA Published 2016. This article is a U.S. Government work and is in the public domain in the USA.

AB - Unmeasured confounding is the fundamental obstacle to drawing causal conclusions about the impact of an intervention from observational data. Typically, covariates are measured to eliminate or ameliorate confounding, but they may be insufficient or unavailable. In the special setting where a transient intervention or exposure varies over time within each individual and confounding is time constant, a different tack is possible. The key idea is to condition on either the overall outcome or the proportion of time in the intervention. These measures can eliminate the unmeasured confounding either by conditioning or by use of a proxy covariate. We evaluate existing methods and develop new models from which causal conclusions can be drawn from such observational data even if no baseline covariates are measured. Our motivation for this work was to determine the causal effect of Streptococcus bacteria in the throat on pharyngitis (sore throat) in Indian schoolchildren. Using our models, we show that existing methods can be badly biased and that sick children who are rarely colonized have a high probability that the Streptococcus bacteria are causing their disease. Published 2016. This article is a U.S. Government work and is in the public domain in the USA Published 2016. This article is a U.S. Government work and is in the public domain in the USA.

KW - Attributable fraction

KW - Case-crossover design

KW - Pattern mixture models

KW - Probability of causation

KW - Self-controlled case series

UR - http://www.scopus.com/inward/record.url?scp=84976900992&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84976900992&partnerID=8YFLogxK

U2 - 10.1002/sim.7000

DO - 10.1002/sim.7000

M3 - Article

C2 - 27313096

AN - SCOPUS:84976900992

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

ER -