schliessen

Filtern

 

Bibliotheken

Prevalence odds ratio or prevalence ratio in the analysis of cross sectional data: what is to be done?

OBJECTIVES: To review the appropriateness of the prevalence odds ratio (POR) and the prevalence ratio (PR) as effect measures in the analysis of cross sectional data and to evaluate different models for the multivariate estimation of the PR. METHODS: A system of linear differential equations corresp... Full description

Journal Title: Occupational and environmental medicine (London England), 1998-04, Vol.55 (4), p.272-277
Main Author: Thompson, M L
Other Authors: Myers, J E , Kriebel, D
Format: Electronic Article Electronic Article
Language: English
Subjects:
Quelle: Alma/SFX Local Collection
Publisher: London: BMJ Publishing Group Ltd
ID: ISSN: 1351-0711
Zum Text:
SendSend as email Add to Book BagAdd to Book Bag
Staff View
recordid: cdi_pubmedcentral_primary_oai_pubmedcentral_nih_gov_1757577
title: Prevalence odds ratio or prevalence ratio in the analysis of cross sectional data: what is to be done?
format: Article
creator:
  • Thompson, M L
  • Myers, J E
  • Kriebel, D
subjects:
  • Biological and medical sciences
  • Chronic Disease - epidemiology
  • Chronic diseases
  • Cross sectional studies
  • Disease models
  • Disease prevalence rates
  • Epidemiology
  • Estimation methods
  • General aspects
  • Humans
  • Incidence
  • Interval estimators
  • Linear models
  • Logistic regression
  • Medical sciences
  • Methodology
  • Models, Statistical
  • Multivariate Analysis
  • Occupational Diseases - epidemiology
  • Odds Ratio
  • Papers
  • Prevalence
  • Public health. Hygiene
  • Public health. Hygiene-occupational medicine
  • Regression Analysis
  • Statistical variance
ispartof: Occupational and environmental medicine (London, England), 1998-04, Vol.55 (4), p.272-277
description: OBJECTIVES: To review the appropriateness of the prevalence odds ratio (POR) and the prevalence ratio (PR) as effect measures in the analysis of cross sectional data and to evaluate different models for the multivariate estimation of the PR. METHODS: A system of linear differential equations corresponding to a dynamic model of a cohort with a chronic disease was developed. At any point in time, a cross sectional analysis of the people then in the cohort provided a prevalence based measure of the effect of exposure on disease. This formed the basis for exploring the relations between the POR, the PR, and the incidence rate ratio (IRR). Examples illustrate relations for various IRRs, prevalences, and differential exodus rates. Multivariate point and interval estimation of the PR by logistic regression is illustrated and compared with the results from proportional hazards regression (PH) and generalised linear modelling (GLM). RESULTS: The POR is difficult to interpret without making restrictive assumptions and the POR and PR may lead to different conclusions with regard to confounding and effect modification. The PR is always conservative relative to the IRR and, if PR > 1, the POR is always > PR. In a fixed cohort and with an adverse exposure, the POR is always > or = IRR, but in a dynamic cohort with sufficient underlying follow up the POR may overestimate or underestimate the IRR, depending on the duration of follow up. Logistic regression models provide point and interval estimates of the PR (and POR) but may be intractable in the presence of many covariates. Proportional hazards and generalised linear models provide statistical methods directed specifically at the PR, but the interval estimation in the case of PH is conservative and the GLM procedure may require constrained estimation. CONCLUSIONS: The PR is conservative, consistent, and interpretable relative to the IRR and should be used in preference to the POR. Multivariate estimation of the PR should be executed by means of generalised linear models or, conservatively, by proportional hazards regression.
language: eng
source: Alma/SFX Local Collection
identifier: ISSN: 1351-0711
fulltext: fulltext
issn:
  • 1351-0711
  • 1470-7926
url: Link


@attributes
NO1
SEARCH_ENGINEprimo_central_multiple_fe
SEARCH_ENGINE_TYPEPrimo Central Search Engine
RANK2.3226063
LOCALfalse
PrimoNMBib
record
control
sourceidjstor_pubme
recordidTN_cdi_pubmedcentral_primary_oai_pubmedcentral_nih_gov_1757577
sourceformatXML
sourcesystemPC
jstor_id27730920
sourcerecordid27730920
originalsourceidFETCH-LOGICAL-1635t-da10d889718e518b98f6aa76c5034475416a850d3da84625bec1aa4ad4eae76b3
addsrcrecordideNqFks1vEzEQxVcIVErhxhXJEkhcSLC9_toeQChQilQVkABxs2btWeqwWQd7U-h_j0OihPYA8sGW38_P4-epqoeMThmr1fOIi6mUUzHlmt-qDpnQdKIbrm6XdS3ZhGrG7lb3cp5Tympd84PqoFFccMMPq-5DwkvocXBIoveZJBhDJDGR5V7Y7IWBjBdIYID-KodMYkdcijmTjK7oZZt4GOGY_LyAkRRgjKRF4uOAL-9XdzroMz7YzkfV55M3n2ank7P3b9_NXp1NmKrlOPHAqDem0cygZKZtTKcAtHKS1kJoKZgCI6mvPRihuGzRMQABXiCgVm19VL3Y-C5X7QK9w2FM0NtlCgtIVzZCsNeVIVzYb_HSMi3L0MXgdGMQlzhASHjtrB9wtNFbrrQ1SjtD61KX6YxgotYdaM-BI3UoJCtWT7e1pPhjhXm0i5Ad9j0MGFfZ6nKEN4I3hXx8g5zHVSqB5lKXZoKr8vJCPdtQf1JP2O1KY9SuO8GWTrBSWmFLJxT80d9R7ODt1xf9yVaH7KDvEgwu5B3GOW1MLQvGb9zqwrjuh3V-of_P3fM8xrT3LAHThtOiTzZ6yCP-2umQvlulay3t-ZeZ_XjylQp6ru3rfZTtYv7vV_8GLqzwZQ
sourcetypeOpen Access Repository
isCDItrue
recordtypearticle
pqid1771426541
display
typearticle
titlePrevalence odds ratio or prevalence ratio in the analysis of cross sectional data: what is to be done?
sourceAlma/SFX Local Collection
creatorThompson, M L ; Myers, J E ; Kriebel, D
creatorcontribThompson, M L ; Myers, J E ; Kriebel, D
descriptionOBJECTIVES: To review the appropriateness of the prevalence odds ratio (POR) and the prevalence ratio (PR) as effect measures in the analysis of cross sectional data and to evaluate different models for the multivariate estimation of the PR. METHODS: A system of linear differential equations corresponding to a dynamic model of a cohort with a chronic disease was developed. At any point in time, a cross sectional analysis of the people then in the cohort provided a prevalence based measure of the effect of exposure on disease. This formed the basis for exploring the relations between the POR, the PR, and the incidence rate ratio (IRR). Examples illustrate relations for various IRRs, prevalences, and differential exodus rates. Multivariate point and interval estimation of the PR by logistic regression is illustrated and compared with the results from proportional hazards regression (PH) and generalised linear modelling (GLM). RESULTS: The POR is difficult to interpret without making restrictive assumptions and the POR and PR may lead to different conclusions with regard to confounding and effect modification. The PR is always conservative relative to the IRR and, if PR > 1, the POR is always > PR. In a fixed cohort and with an adverse exposure, the POR is always > or = IRR, but in a dynamic cohort with sufficient underlying follow up the POR may overestimate or underestimate the IRR, depending on the duration of follow up. Logistic regression models provide point and interval estimates of the PR (and POR) but may be intractable in the presence of many covariates. Proportional hazards and generalised linear models provide statistical methods directed specifically at the PR, but the interval estimation in the case of PH is conservative and the GLM procedure may require constrained estimation. CONCLUSIONS: The PR is conservative, consistent, and interpretable relative to the IRR and should be used in preference to the POR. Multivariate estimation of the PR should be executed by means of generalised linear models or, conservatively, by proportional hazards regression.
identifier
0ISSN: 1351-0711
1EISSN: 1470-7926
2DOI: 10.1136/oem.55.4.272
3PMID: 9624282
languageeng
publisherLondon: BMJ Publishing Group Ltd
subjectBiological and medical sciences ; Chronic Disease - epidemiology ; Chronic diseases ; Cross sectional studies ; Disease models ; Disease prevalence rates ; Epidemiology ; Estimation methods ; General aspects ; Humans ; Incidence ; Interval estimators ; Linear models ; Logistic regression ; Medical sciences ; Methodology ; Models, Statistical ; Multivariate Analysis ; Occupational Diseases - epidemiology ; Odds Ratio ; Papers ; Prevalence ; Public health. Hygiene ; Public health. Hygiene-occupational medicine ; Regression Analysis ; Statistical variance
ispartofOccupational and environmental medicine (London, England), 1998-04, Vol.55 (4), p.272-277
rights
01998 INIST-CNRS
1Copyright BMJ Publishing Group LTD Apr 1998
lds50peer_reviewed
oafree_for_read
citedbyFETCH-LOGICAL-1635t-da10d889718e518b98f6aa76c5034475416a850d3da84625bec1aa4ad4eae76b3
citesFETCH-LOGICAL-1635t-da10d889718e518b98f6aa76c5034475416a850d3da84625bec1aa4ad4eae76b3
links
openurl$$Topenurl_article
openurlfulltext$$Topenurlfull_article
thumbnail$$Usyndetics_thumb_exl
backlink
0$$Uhttp://pascal-francis.inist.fr/vibad/index.php?action=getRecordDetail&idt=2209835$$DView record in Pascal Francis
1$$Uhttps://www.ncbi.nlm.nih.gov/pubmed/9624282$$D View this record in MEDLINE/PubMed
search
creatorcontrib
0Thompson, M L
1Myers, J E
2Kriebel, D
title
0Prevalence odds ratio or prevalence ratio in the analysis of cross sectional data: what is to be done?
1Occupational and environmental medicine (London, England)
addtitleOccup Environ Med
descriptionOBJECTIVES: To review the appropriateness of the prevalence odds ratio (POR) and the prevalence ratio (PR) as effect measures in the analysis of cross sectional data and to evaluate different models for the multivariate estimation of the PR. METHODS: A system of linear differential equations corresponding to a dynamic model of a cohort with a chronic disease was developed. At any point in time, a cross sectional analysis of the people then in the cohort provided a prevalence based measure of the effect of exposure on disease. This formed the basis for exploring the relations between the POR, the PR, and the incidence rate ratio (IRR). Examples illustrate relations for various IRRs, prevalences, and differential exodus rates. Multivariate point and interval estimation of the PR by logistic regression is illustrated and compared with the results from proportional hazards regression (PH) and generalised linear modelling (GLM). RESULTS: The POR is difficult to interpret without making restrictive assumptions and the POR and PR may lead to different conclusions with regard to confounding and effect modification. The PR is always conservative relative to the IRR and, if PR > 1, the POR is always > PR. In a fixed cohort and with an adverse exposure, the POR is always > or = IRR, but in a dynamic cohort with sufficient underlying follow up the POR may overestimate or underestimate the IRR, depending on the duration of follow up. Logistic regression models provide point and interval estimates of the PR (and POR) but may be intractable in the presence of many covariates. Proportional hazards and generalised linear models provide statistical methods directed specifically at the PR, but the interval estimation in the case of PH is conservative and the GLM procedure may require constrained estimation. CONCLUSIONS: The PR is conservative, consistent, and interpretable relative to the IRR and should be used in preference to the POR. Multivariate estimation of the PR should be executed by means of generalised linear models or, conservatively, by proportional hazards regression.
subject
0Biological and medical sciences
1Chronic Disease - epidemiology
2Chronic diseases
3Cross sectional studies
4Disease models
5Disease prevalence rates
6Epidemiology
7Estimation methods
8General aspects
9Humans
10Incidence
11Interval estimators
12Linear models
13Logistic regression
14Medical sciences
15Methodology
16Models, Statistical
17Multivariate Analysis
18Occupational Diseases - epidemiology
19Odds Ratio
20Papers
21Prevalence
22Public health. Hygiene
23Public health. Hygiene-occupational medicine
24Regression Analysis
25Statistical variance
issn
01351-0711
11470-7926
fulltexttrue
rsrctypearticle
creationdate1998
recordtypearticle
recordideNqFks1vEzEQxVcIVErhxhXJEkhcSLC9_toeQChQilQVkABxs2btWeqwWQd7U-h_j0OihPYA8sGW38_P4-epqoeMThmr1fOIi6mUUzHlmt-qDpnQdKIbrm6XdS3ZhGrG7lb3cp5Tympd84PqoFFccMMPq-5DwkvocXBIoveZJBhDJDGR5V7Y7IWBjBdIYID-KodMYkdcijmTjK7oZZt4GOGY_LyAkRRgjKRF4uOAL-9XdzroMz7YzkfV55M3n2ank7P3b9_NXp1NmKrlOPHAqDem0cygZKZtTKcAtHKS1kJoKZgCI6mvPRihuGzRMQABXiCgVm19VL3Y-C5X7QK9w2FM0NtlCgtIVzZCsNeVIVzYb_HSMi3L0MXgdGMQlzhASHjtrB9wtNFbrrQ1SjtD61KX6YxgotYdaM-BI3UoJCtWT7e1pPhjhXm0i5Ad9j0MGFfZ6nKEN4I3hXx8g5zHVSqB5lKXZoKr8vJCPdtQf1JP2O1KY9SuO8GWTrBSWmFLJxT80d9R7ODt1xf9yVaH7KDvEgwu5B3GOW1MLQvGb9zqwrjuh3V-of_P3fM8xrT3LAHThtOiTzZ6yCP-2umQvlulay3t-ZeZ_XjylQp6ru3rfZTtYv7vV_8GLqzwZQ
startdate199804
enddate199804
creator
0Thompson, M L
1Myers, J E
2Kriebel, D
general
0BMJ Publishing Group Ltd
1BMJ Publishing Group
2BMJ
3BMJ Publishing Group LTD
4BMJ Group
scope
0BSCLL
1IQODW
2CGR
3CUY
4CVF
5ECM
6EIF
7NPM
8AAYXX
9CITATION
103V.
117RV
127X7
137XB
1488E
158C1
168FE
178FG
188FI
198FJ
208FK
21ABJCF
22ABUWG
23ATCPS
24AZQEC
25BENPR
26BGLVJ
27BHPHI
28BTHHO
29DWQXO
30FYUFA
31GHDGH
32GNUQQ
33HCIFZ
34K9.
35KB0
36L6V
37M0S
38M1P
39M7S
40NAPCQ
41PATMY
42PQEST
43PQQKQ
44PQUKI
45PRINS
46PTHSS
47PYCSY
487QQ
497TB
508FD
51FR3
52JG9
53KR7
54BOBZL
55CLFQK
565PM
sort
creationdate199804
titlePrevalence odds ratio or prevalence ratio in the analysis of cross sectional data: what is to be done?
authorThompson, M L ; Myers, J E ; Kriebel, D
facets
frbrtype5
frbrgroupidcdi_FETCH-LOGICAL-1635t-da10d889718e518b98f6aa76c5034475416a850d3da84625bec1aa4ad4eae76b3
rsrctypearticles
prefilterarticles
languageeng
creationdate1998
topic
0Biological and medical sciences
1Chronic Disease - epidemiology
2Chronic diseases
3Cross sectional studies
4Disease models
5Disease prevalence rates
6Epidemiology
7Estimation methods
8General aspects
9Humans
10Incidence
11Interval estimators
12Linear models
13Logistic regression
14Medical sciences
15Methodology
16Models, Statistical
17Multivariate Analysis
18Occupational Diseases - epidemiology
19Odds Ratio
20Papers
21Prevalence
22Public health. Hygiene
23Public health. Hygiene-occupational medicine
24Regression Analysis
25Statistical variance
toplevel
0peer_reviewed
1online_resources
creatorcontrib
0Thompson, M L
1Myers, J E
2Kriebel, D
collection
0Istex
1Pascal-Francis
2Medline
3MEDLINE
4MEDLINE (Ovid)
5MEDLINE
6MEDLINE
7PubMed
8CrossRef
9ProQuest Central (Corporate)
10Nursing & Allied Health Database
11Health & Medical Collection
12ProQuest Central (purchase pre-March 2016)
13Medical Database (Alumni Edition)
14Public Health Database
15ProQuest SciTech Collection
16ProQuest Technology Collection
17Hospital Premium Collection
18Hospital Premium Collection (Alumni Edition)
19ProQuest Central (Alumni) (purchase pre-March 2016)
20Materials Science & Engineering Collection
21ProQuest Central (Alumni Edition)
22Agricultural & Environmental Science Collection
23ProQuest Central Essentials
24ProQuest Central
25Technology Collection
26Natural Science Collection
27BMJ Journals
28ProQuest Central Korea
29Health Research Premium Collection
30Health Research Premium Collection (Alumni)
31ProQuest Central Student
32SciTech Premium Collection
33ProQuest Health & Medical Complete (Alumni)
34Nursing & Allied Health Database (Alumni Edition)
35ProQuest Engineering Collection
36Health & Medical Collection (Alumni Edition)
37Medical Database
38Engineering Database
39Nursing & Allied Health Premium
40Environmental Science Database
41ProQuest One Academic Eastern Edition
42ProQuest One Academic
43ProQuest One Academic UKI Edition
44ProQuest Central China
45Engineering Collection
46Environmental Science Collection
47Ceramic Abstracts
48Mechanical & Transportation Engineering Abstracts
49Technology Research Database
50Engineering Research Database
51Materials Research Database
52Civil Engineering Abstracts
53OpenAIRE (Open Access)
54OpenAIRE
55PubMed Central (Full Participant titles)
jtitleOccupational and environmental medicine (London, England)
delivery
delcategoryRemote Search Resource
fulltextfulltext
addata
au
0Thompson, M L
1Myers, J E
2Kriebel, D
formatjournal
genrearticle
ristypeJOUR
atitlePrevalence odds ratio or prevalence ratio in the analysis of cross sectional data: what is to be done?
jtitleOccupational and environmental medicine (London, England)
addtitleOccup Environ Med
date1998-04
risdate1998
volume55
issue4
spage272
epage277
pages272-277
issn1351-0711
eissn1470-7926
abstractOBJECTIVES: To review the appropriateness of the prevalence odds ratio (POR) and the prevalence ratio (PR) as effect measures in the analysis of cross sectional data and to evaluate different models for the multivariate estimation of the PR. METHODS: A system of linear differential equations corresponding to a dynamic model of a cohort with a chronic disease was developed. At any point in time, a cross sectional analysis of the people then in the cohort provided a prevalence based measure of the effect of exposure on disease. This formed the basis for exploring the relations between the POR, the PR, and the incidence rate ratio (IRR). Examples illustrate relations for various IRRs, prevalences, and differential exodus rates. Multivariate point and interval estimation of the PR by logistic regression is illustrated and compared with the results from proportional hazards regression (PH) and generalised linear modelling (GLM). RESULTS: The POR is difficult to interpret without making restrictive assumptions and the POR and PR may lead to different conclusions with regard to confounding and effect modification. The PR is always conservative relative to the IRR and, if PR > 1, the POR is always > PR. In a fixed cohort and with an adverse exposure, the POR is always > or = IRR, but in a dynamic cohort with sufficient underlying follow up the POR may overestimate or underestimate the IRR, depending on the duration of follow up. Logistic regression models provide point and interval estimates of the PR (and POR) but may be intractable in the presence of many covariates. Proportional hazards and generalised linear models provide statistical methods directed specifically at the PR, but the interval estimation in the case of PH is conservative and the GLM procedure may require constrained estimation. CONCLUSIONS: The PR is conservative, consistent, and interpretable relative to the IRR and should be used in preference to the POR. Multivariate estimation of the PR should be executed by means of generalised linear models or, conservatively, by proportional hazards regression.
copLondon
pubBMJ Publishing Group Ltd
pmid9624282
doi10.1136/oem.55.4.272
oafree_for_read