Confidence limits for prevalence of disease adjusted for estimated sensitivity and specificity

Zsolt Lang, J. Reiczigel

Research output: Contribution to journalArticle

11 Citations (Scopus)

Abstract

Prevalence of a disease is usually assessed by diagnostic tests that may produce false results. Rogan and Gladen (1978) described a method to estimate the true prevalence correcting for sensitivity and specificity of the diagnostic procedure, and Reiczigel et al. (2010) provided exact confidence intervals for the true prevalence assuming sensitivity and specificity were known. In this paper we propose a new method to construct approximate confidence intervals for the true prevalence when sensitivity and specificity are estimated from independent samples. To improve coverage we applied an adjustment similar to that described in Agresti and Coull (1998). According to an extensive simulation study the new confidence intervals maintain the nominal level fairly well even for sample sizes as small as 30; minimum coverage is above 88%, 93%, and 98% at nominal 90%, 95%, and 99%, respectively. We illustrate the advantages of the proposed method with real-life applications.

Original languageEnglish
Pages (from-to)13-22
Number of pages10
JournalPreventive Veterinary Medicine
Volume113
Issue number1
DOIs
Publication statusPublished - Jan 1 2014

Fingerprint

disease prevalence
confidence interval
Sensitivity and Specificity
Confidence Intervals
diagnostic specificity
diagnostic sensitivity
Agrostis
diagnostic techniques
methodology
Routine Diagnostic Tests
Sample Size
sampling

Keywords

  • Binomial confidence interval
  • Coverage probability
  • Exact and approximate inference
  • False negatives
  • False positives
  • Imperfect diagnostic test

ASJC Scopus subject areas

  • Animal Science and Zoology
  • Food Animals

Cite this

Confidence limits for prevalence of disease adjusted for estimated sensitivity and specificity. / Lang, Zsolt; Reiczigel, J.

In: Preventive Veterinary Medicine, Vol. 113, No. 1, 01.01.2014, p. 13-22.

Research output: Contribution to journalArticle

@article{5231a24524eb4378aee4ca6cf26d419f,
title = "Confidence limits for prevalence of disease adjusted for estimated sensitivity and specificity",
abstract = "Prevalence of a disease is usually assessed by diagnostic tests that may produce false results. Rogan and Gladen (1978) described a method to estimate the true prevalence correcting for sensitivity and specificity of the diagnostic procedure, and Reiczigel et al. (2010) provided exact confidence intervals for the true prevalence assuming sensitivity and specificity were known. In this paper we propose a new method to construct approximate confidence intervals for the true prevalence when sensitivity and specificity are estimated from independent samples. To improve coverage we applied an adjustment similar to that described in Agresti and Coull (1998). According to an extensive simulation study the new confidence intervals maintain the nominal level fairly well even for sample sizes as small as 30; minimum coverage is above 88{\%}, 93{\%}, and 98{\%} at nominal 90{\%}, 95{\%}, and 99{\%}, respectively. We illustrate the advantages of the proposed method with real-life applications.",
keywords = "Binomial confidence interval, Coverage probability, Exact and approximate inference, False negatives, False positives, Imperfect diagnostic test",
author = "Zsolt Lang and J. Reiczigel",
year = "2014",
month = "1",
day = "1",
doi = "10.1016/j.prevetmed.2013.09.015",
language = "English",
volume = "113",
pages = "13--22",
journal = "Preventive Veterinary Medicine",
issn = "0167-5877",
publisher = "Elsevier",
number = "1",

}

TY - JOUR

T1 - Confidence limits for prevalence of disease adjusted for estimated sensitivity and specificity

AU - Lang, Zsolt

AU - Reiczigel, J.

PY - 2014/1/1

Y1 - 2014/1/1

N2 - Prevalence of a disease is usually assessed by diagnostic tests that may produce false results. Rogan and Gladen (1978) described a method to estimate the true prevalence correcting for sensitivity and specificity of the diagnostic procedure, and Reiczigel et al. (2010) provided exact confidence intervals for the true prevalence assuming sensitivity and specificity were known. In this paper we propose a new method to construct approximate confidence intervals for the true prevalence when sensitivity and specificity are estimated from independent samples. To improve coverage we applied an adjustment similar to that described in Agresti and Coull (1998). According to an extensive simulation study the new confidence intervals maintain the nominal level fairly well even for sample sizes as small as 30; minimum coverage is above 88%, 93%, and 98% at nominal 90%, 95%, and 99%, respectively. We illustrate the advantages of the proposed method with real-life applications.

AB - Prevalence of a disease is usually assessed by diagnostic tests that may produce false results. Rogan and Gladen (1978) described a method to estimate the true prevalence correcting for sensitivity and specificity of the diagnostic procedure, and Reiczigel et al. (2010) provided exact confidence intervals for the true prevalence assuming sensitivity and specificity were known. In this paper we propose a new method to construct approximate confidence intervals for the true prevalence when sensitivity and specificity are estimated from independent samples. To improve coverage we applied an adjustment similar to that described in Agresti and Coull (1998). According to an extensive simulation study the new confidence intervals maintain the nominal level fairly well even for sample sizes as small as 30; minimum coverage is above 88%, 93%, and 98% at nominal 90%, 95%, and 99%, respectively. We illustrate the advantages of the proposed method with real-life applications.

KW - Binomial confidence interval

KW - Coverage probability

KW - Exact and approximate inference

KW - False negatives

KW - False positives

KW - Imperfect diagnostic test

UR - http://www.scopus.com/inward/record.url?scp=84888203241&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84888203241&partnerID=8YFLogxK

U2 - 10.1016/j.prevetmed.2013.09.015

DO - 10.1016/j.prevetmed.2013.09.015

M3 - Article

C2 - 24416798

AN - SCOPUS:84888203241

VL - 113

SP - 13

EP - 22

JO - Preventive Veterinary Medicine

JF - Preventive Veterinary Medicine

SN - 0167-5877

IS - 1

ER -