Hypothesis Tests for Neyman's Bias in Case–Control Studies

D. M. Swanson, C. D. Anderson, Rebecca Betensky

Research output: Contribution to journalArticle

Abstract

Survival bias is a long recognized problem in case–control studies, and many varieties of bias can come under this umbrella term. We focus on one of them, termed Neyman's bias or ‘prevalence–incidence bias’. It occurs in case–control studies when exposure affects both disease and disease-induced mortality, and we give a formula for the observed, biased odds ratio under such conditions. We compare our result with previous investigations into this phenomenon and consider models under which this bias may or may not be important. Finally, we propose three hypothesis tests to identify when Neyman's bias may be present in case–control studies. We apply these tests to three data sets, one of stroke mortality, another of brain tumors, and the last of atrial fibrillation, and find some evidence of Neyman's bias in the former two cases, but not the last case.

Original languageEnglish (US)
Pages (from-to)1956-1977
Number of pages22
JournalJournal of Applied Statistics
Volume45
Issue number11
DOIs
StatePublished - Aug 18 2018

Fingerprint

Case-control Study
Hypothesis Test
Mortality
Atrial Fibrillation
Brain Tumor
Odds Ratio
Hypothesis test
Stroke
Biased
Term

Keywords

  • Odds ratio
  • survival bias
  • truncation

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

Hypothesis Tests for Neyman's Bias in Case–Control Studies. / Swanson, D. M.; Anderson, C. D.; Betensky, Rebecca.

In: Journal of Applied Statistics, Vol. 45, No. 11, 18.08.2018, p. 1956-1977.

Research output: Contribution to journalArticle

Swanson, D. M. ; Anderson, C. D. ; Betensky, Rebecca. / Hypothesis Tests for Neyman's Bias in Case–Control Studies. In: Journal of Applied Statistics. 2018 ; Vol. 45, No. 11. pp. 1956-1977.
@article{a452a99ad09342d9a71029bc4b69c4eb,
title = "Hypothesis Tests for Neyman's Bias in Case–Control Studies",
abstract = "Survival bias is a long recognized problem in case–control studies, and many varieties of bias can come under this umbrella term. We focus on one of them, termed Neyman's bias or ‘prevalence–incidence bias’. It occurs in case–control studies when exposure affects both disease and disease-induced mortality, and we give a formula for the observed, biased odds ratio under such conditions. We compare our result with previous investigations into this phenomenon and consider models under which this bias may or may not be important. Finally, we propose three hypothesis tests to identify when Neyman's bias may be present in case–control studies. We apply these tests to three data sets, one of stroke mortality, another of brain tumors, and the last of atrial fibrillation, and find some evidence of Neyman's bias in the former two cases, but not the last case.",
keywords = "Odds ratio, survival bias, truncation",
author = "Swanson, {D. M.} and Anderson, {C. D.} and Rebecca Betensky",
year = "2018",
month = "8",
day = "18",
doi = "10.1080/02664763.2017.1401053",
language = "English (US)",
volume = "45",
pages = "1956--1977",
journal = "Journal of Applied Statistics",
issn = "0266-4763",
publisher = "Routledge",
number = "11",

}

TY - JOUR

T1 - Hypothesis Tests for Neyman's Bias in Case–Control Studies

AU - Swanson, D. M.

AU - Anderson, C. D.

AU - Betensky, Rebecca

PY - 2018/8/18

Y1 - 2018/8/18

N2 - Survival bias is a long recognized problem in case–control studies, and many varieties of bias can come under this umbrella term. We focus on one of them, termed Neyman's bias or ‘prevalence–incidence bias’. It occurs in case–control studies when exposure affects both disease and disease-induced mortality, and we give a formula for the observed, biased odds ratio under such conditions. We compare our result with previous investigations into this phenomenon and consider models under which this bias may or may not be important. Finally, we propose three hypothesis tests to identify when Neyman's bias may be present in case–control studies. We apply these tests to three data sets, one of stroke mortality, another of brain tumors, and the last of atrial fibrillation, and find some evidence of Neyman's bias in the former two cases, but not the last case.

AB - Survival bias is a long recognized problem in case–control studies, and many varieties of bias can come under this umbrella term. We focus on one of them, termed Neyman's bias or ‘prevalence–incidence bias’. It occurs in case–control studies when exposure affects both disease and disease-induced mortality, and we give a formula for the observed, biased odds ratio under such conditions. We compare our result with previous investigations into this phenomenon and consider models under which this bias may or may not be important. Finally, we propose three hypothesis tests to identify when Neyman's bias may be present in case–control studies. We apply these tests to three data sets, one of stroke mortality, another of brain tumors, and the last of atrial fibrillation, and find some evidence of Neyman's bias in the former two cases, but not the last case.

KW - Odds ratio

KW - survival bias

KW - truncation

UR - http://www.scopus.com/inward/record.url?scp=85034236342&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85034236342&partnerID=8YFLogxK

U2 - 10.1080/02664763.2017.1401053

DO - 10.1080/02664763.2017.1401053

M3 - Article

VL - 45

SP - 1956

EP - 1977

JO - Journal of Applied Statistics

JF - Journal of Applied Statistics

SN - 0266-4763

IS - 11

ER -