Extended matrix and inverse matrix methods utilizing internal validation data when both disease and exposure status are misclassified

Li Tang, Robert H. Lyles, Ye Ye, Yungtai Lo, Caroline C. King

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

The problem of misclassification is common in epidemiological and clinical research. In some cases, misclassification may be incurred when measuring both exposure and outcome variables. It is well known that validity of analytic results (e.g. point and confidence interval estimates for odds ratios of interest) can be forfeited when no correction effort is made. Therefore, valid and accessible methods with which to deal with these issues remain in high demand. Here, we elucidate extensions of well-studied methods in order to facilitate misclassification adjustment when a binary outcome and binary exposure variable are both subject to misclassification. By formulating generalizations of assumptions underlying well-studied "matrix" and "inverse matrix" methods into the framework of maximum likelihood, our approach allows the flexible modeling of a richer set of misclassification mechanisms when adequate internal validation data are available. The value of our extensions and a strong case for the internal validation design are demonstrated by means of simulations and analysis of bacterial vaginosis and trichomoniasis data from the HIV Epidemiology Research Study.

Original languageEnglish (US)
Pages (from-to)49-66
Number of pages18
JournalEpidemiologic Methods
Volume2
Issue number1
DOIs
StatePublished - Dec 1 2013

Fingerprint

Misclassification
Inverse Method
Inverse matrix
Matrix Method
Internal
Epidemiology
Maximum likelihood
Bacterial Vaginosis
Research
Reproducibility of Results
Binary Outcomes
Odds Ratio
HIV
Confidence Intervals
Maximum Likelihood
Confidence interval
Adjustment
Valid
Binary
Modeling

Keywords

  • Inverse matrix method
  • Likelihood
  • Matrix method
  • Misclassification

ASJC Scopus subject areas

  • Epidemiology
  • Applied Mathematics

Cite this

Extended matrix and inverse matrix methods utilizing internal validation data when both disease and exposure status are misclassified. / Tang, Li; Lyles, Robert H.; Ye, Ye; Lo, Yungtai; King, Caroline C.

In: Epidemiologic Methods, Vol. 2, No. 1, 01.12.2013, p. 49-66.

Research output: Contribution to journalArticle

@article{77aebd6e1b0244848b2ea7d3be5c3487,
title = "Extended matrix and inverse matrix methods utilizing internal validation data when both disease and exposure status are misclassified",
abstract = "The problem of misclassification is common in epidemiological and clinical research. In some cases, misclassification may be incurred when measuring both exposure and outcome variables. It is well known that validity of analytic results (e.g. point and confidence interval estimates for odds ratios of interest) can be forfeited when no correction effort is made. Therefore, valid and accessible methods with which to deal with these issues remain in high demand. Here, we elucidate extensions of well-studied methods in order to facilitate misclassification adjustment when a binary outcome and binary exposure variable are both subject to misclassification. By formulating generalizations of assumptions underlying well-studied {"}matrix{"} and {"}inverse matrix{"} methods into the framework of maximum likelihood, our approach allows the flexible modeling of a richer set of misclassification mechanisms when adequate internal validation data are available. The value of our extensions and a strong case for the internal validation design are demonstrated by means of simulations and analysis of bacterial vaginosis and trichomoniasis data from the HIV Epidemiology Research Study.",
keywords = "Inverse matrix method, Likelihood, Matrix method, Misclassification",
author = "Li Tang and Lyles, {Robert H.} and Ye Ye and Yungtai Lo and King, {Caroline C.}",
year = "2013",
month = "12",
day = "1",
doi = "10.1515/em-2013-0008",
language = "English (US)",
volume = "2",
pages = "49--66",
journal = "Epidemiologic Methods",
issn = "2194-9263",
publisher = "Walter De Gruyter",
number = "1",

}

TY - JOUR

T1 - Extended matrix and inverse matrix methods utilizing internal validation data when both disease and exposure status are misclassified

AU - Tang, Li

AU - Lyles, Robert H.

AU - Ye, Ye

AU - Lo, Yungtai

AU - King, Caroline C.

PY - 2013/12/1

Y1 - 2013/12/1

N2 - The problem of misclassification is common in epidemiological and clinical research. In some cases, misclassification may be incurred when measuring both exposure and outcome variables. It is well known that validity of analytic results (e.g. point and confidence interval estimates for odds ratios of interest) can be forfeited when no correction effort is made. Therefore, valid and accessible methods with which to deal with these issues remain in high demand. Here, we elucidate extensions of well-studied methods in order to facilitate misclassification adjustment when a binary outcome and binary exposure variable are both subject to misclassification. By formulating generalizations of assumptions underlying well-studied "matrix" and "inverse matrix" methods into the framework of maximum likelihood, our approach allows the flexible modeling of a richer set of misclassification mechanisms when adequate internal validation data are available. The value of our extensions and a strong case for the internal validation design are demonstrated by means of simulations and analysis of bacterial vaginosis and trichomoniasis data from the HIV Epidemiology Research Study.

AB - The problem of misclassification is common in epidemiological and clinical research. In some cases, misclassification may be incurred when measuring both exposure and outcome variables. It is well known that validity of analytic results (e.g. point and confidence interval estimates for odds ratios of interest) can be forfeited when no correction effort is made. Therefore, valid and accessible methods with which to deal with these issues remain in high demand. Here, we elucidate extensions of well-studied methods in order to facilitate misclassification adjustment when a binary outcome and binary exposure variable are both subject to misclassification. By formulating generalizations of assumptions underlying well-studied "matrix" and "inverse matrix" methods into the framework of maximum likelihood, our approach allows the flexible modeling of a richer set of misclassification mechanisms when adequate internal validation data are available. The value of our extensions and a strong case for the internal validation design are demonstrated by means of simulations and analysis of bacterial vaginosis and trichomoniasis data from the HIV Epidemiology Research Study.

KW - Inverse matrix method

KW - Likelihood

KW - Matrix method

KW - Misclassification

UR - http://www.scopus.com/inward/record.url?scp=84926407001&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84926407001&partnerID=8YFLogxK

U2 - 10.1515/em-2013-0008

DO - 10.1515/em-2013-0008

M3 - Article

AN - SCOPUS:84926407001

VL - 2

SP - 49

EP - 66

JO - Epidemiologic Methods

JF - Epidemiologic Methods

SN - 2194-9263

IS - 1

ER -