Group testing regression models with dilution submodels

Md S. Warasi, Christopher S. McMahan, Joshua M. Tebbs, Christopher R Bilder

Research output: Contribution to journalArticle

Abstract

Group testing, where specimens are tested initially in pools, is widely used to screen individuals for sexually transmitted diseases. However, a common problem encountered in practice is that group testing can increase the number of false negative test results. This occurs primarily when positive individual specimens within a pool are diluted by negative ones, resulting in positive pools testing negatively. If the goal is to estimate a population-level regression model relating individual disease status to observed covariates, severe bias can result if an adjustment for dilution is not made. Recognizing this as a critical issue, recent binary regression approaches in group testing have utilized continuous biomarker information to acknowledge the effect of dilution. In this paper, we have the same overall goal but take a different approach. We augment existing group testing regression models (that assume no dilution) with a parametric dilution submodel for pool-level sensitivity and estimate all parameters using maximum likelihood. An advantage of our approach is that it does not rely on external biomarker test data, which may not be available in surveillance studies. Furthermore, unlike previous approaches, our framework allows one to formally test whether dilution is present based on the observed group testing data. We use simulation to illustrate the performance of our estimation and inference methods, and we apply these methods to 2 infectious disease data sets.

Original languageEnglish (US)
Pages (from-to)4860-4872
Number of pages13
JournalStatistics in Medicine
Volume36
Issue number30
DOIs
StatePublished - Dec 30 2017

Fingerprint

Group Testing
Regression Model
Biomarkers
Sexually Transmitted Diseases
Communicable Diseases
Binary Regression
Population
Infectious Diseases
Estimate
Surveillance
Maximum Likelihood
Covariates
Adjustment
Testing
Simulation
Datasets

Keywords

  • binary regression
  • dilution effect
  • likelihood ratio test
  • maximum likelihood
  • pooled testing
  • sensitivity

ASJC Scopus subject areas

  • Epidemiology
  • Statistics and Probability

Cite this

Group testing regression models with dilution submodels. / Warasi, Md S.; McMahan, Christopher S.; Tebbs, Joshua M.; Bilder, Christopher R.

In: Statistics in Medicine, Vol. 36, No. 30, 30.12.2017, p. 4860-4872.

Research output: Contribution to journalArticle

Warasi, Md S. ; McMahan, Christopher S. ; Tebbs, Joshua M. ; Bilder, Christopher R. / Group testing regression models with dilution submodels. In: Statistics in Medicine. 2017 ; Vol. 36, No. 30. pp. 4860-4872.
@article{45ca536df2e74a0f9876894158288578,
title = "Group testing regression models with dilution submodels",
abstract = "Group testing, where specimens are tested initially in pools, is widely used to screen individuals for sexually transmitted diseases. However, a common problem encountered in practice is that group testing can increase the number of false negative test results. This occurs primarily when positive individual specimens within a pool are diluted by negative ones, resulting in positive pools testing negatively. If the goal is to estimate a population-level regression model relating individual disease status to observed covariates, severe bias can result if an adjustment for dilution is not made. Recognizing this as a critical issue, recent binary regression approaches in group testing have utilized continuous biomarker information to acknowledge the effect of dilution. In this paper, we have the same overall goal but take a different approach. We augment existing group testing regression models (that assume no dilution) with a parametric dilution submodel for pool-level sensitivity and estimate all parameters using maximum likelihood. An advantage of our approach is that it does not rely on external biomarker test data, which may not be available in surveillance studies. Furthermore, unlike previous approaches, our framework allows one to formally test whether dilution is present based on the observed group testing data. We use simulation to illustrate the performance of our estimation and inference methods, and we apply these methods to 2 infectious disease data sets.",
keywords = "binary regression, dilution effect, likelihood ratio test, maximum likelihood, pooled testing, sensitivity",
author = "Warasi, {Md S.} and McMahan, {Christopher S.} and Tebbs, {Joshua M.} and Bilder, {Christopher R}",
year = "2017",
month = "12",
day = "30",
doi = "10.1002/sim.7455",
language = "English (US)",
volume = "36",
pages = "4860--4872",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",
number = "30",

}

TY - JOUR

T1 - Group testing regression models with dilution submodels

AU - Warasi, Md S.

AU - McMahan, Christopher S.

AU - Tebbs, Joshua M.

AU - Bilder, Christopher R

PY - 2017/12/30

Y1 - 2017/12/30

N2 - Group testing, where specimens are tested initially in pools, is widely used to screen individuals for sexually transmitted diseases. However, a common problem encountered in practice is that group testing can increase the number of false negative test results. This occurs primarily when positive individual specimens within a pool are diluted by negative ones, resulting in positive pools testing negatively. If the goal is to estimate a population-level regression model relating individual disease status to observed covariates, severe bias can result if an adjustment for dilution is not made. Recognizing this as a critical issue, recent binary regression approaches in group testing have utilized continuous biomarker information to acknowledge the effect of dilution. In this paper, we have the same overall goal but take a different approach. We augment existing group testing regression models (that assume no dilution) with a parametric dilution submodel for pool-level sensitivity and estimate all parameters using maximum likelihood. An advantage of our approach is that it does not rely on external biomarker test data, which may not be available in surveillance studies. Furthermore, unlike previous approaches, our framework allows one to formally test whether dilution is present based on the observed group testing data. We use simulation to illustrate the performance of our estimation and inference methods, and we apply these methods to 2 infectious disease data sets.

AB - Group testing, where specimens are tested initially in pools, is widely used to screen individuals for sexually transmitted diseases. However, a common problem encountered in practice is that group testing can increase the number of false negative test results. This occurs primarily when positive individual specimens within a pool are diluted by negative ones, resulting in positive pools testing negatively. If the goal is to estimate a population-level regression model relating individual disease status to observed covariates, severe bias can result if an adjustment for dilution is not made. Recognizing this as a critical issue, recent binary regression approaches in group testing have utilized continuous biomarker information to acknowledge the effect of dilution. In this paper, we have the same overall goal but take a different approach. We augment existing group testing regression models (that assume no dilution) with a parametric dilution submodel for pool-level sensitivity and estimate all parameters using maximum likelihood. An advantage of our approach is that it does not rely on external biomarker test data, which may not be available in surveillance studies. Furthermore, unlike previous approaches, our framework allows one to formally test whether dilution is present based on the observed group testing data. We use simulation to illustrate the performance of our estimation and inference methods, and we apply these methods to 2 infectious disease data sets.

KW - binary regression

KW - dilution effect

KW - likelihood ratio test

KW - maximum likelihood

KW - pooled testing

KW - sensitivity

UR - http://www.scopus.com/inward/record.url?scp=85037606187&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85037606187&partnerID=8YFLogxK

U2 - 10.1002/sim.7455

DO - 10.1002/sim.7455

M3 - Article

VL - 36

SP - 4860

EP - 4872

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 30

ER -