Imbalance p values for baseline covariates in randomized controlled trials: a last resort for the use of p values? A pro and contra debate
Andreas Stang,1,2 Christopher Baethge3,4 1Center of Clinical Epidemiology, Institute of Medical Informatics, Biometry and Epidemiology, Medical Faculty, University Hospital of Essen, Hufelandstr, Essen, Germany; 2Department of Epidemiology, School of Public Health, Boston University, Boston, MA, USA...
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
Dove Medical Press
2018-05-01
|
Series: | Clinical Epidemiology |
Subjects: | |
Online Access: | https://www.dovepress.com/imbalance-p-values-for-baseline-covariates-in-randomized-controlled-tr-peer-reviewed-article-CLEP |
id |
doaj-b1452fd961974b099afa5e17bb8709fc |
---|---|
record_format |
Article |
spelling |
doaj-b1452fd961974b099afa5e17bb8709fc2020-11-25T00:24:11ZengDove Medical PressClinical Epidemiology1179-13492018-05-01Volume 1053153538186Imbalance p values for baseline covariates in randomized controlled trials: a last resort for the use of p values? A pro and contra debateStang ABaethge CAndreas Stang,1,2 Christopher Baethge3,4 1Center of Clinical Epidemiology, Institute of Medical Informatics, Biometry and Epidemiology, Medical Faculty, University Hospital of Essen, Hufelandstr, Essen, Germany; 2Department of Epidemiology, School of Public Health, Boston University, Boston, MA, USA; 3Department of Psychiatry and Psychotherapy, University of Cologne Medical School, Cologne, Germany; 4Editorial Offices, Deutsches Ärzteblatt and Deutsches Ärzteblatt International, Deutscher Ärzte-Verlag, Cologne, Germany Background: Results of randomized controlled trials (RCTs) are usually accompanied by a table that compares covariates between the study groups at baseline. Sometimes, the investigators report p values for imbalanced covariates. The aim of this debate is to illustrate the pro and contra of the use of these p values in RCTs.Pro: Low p values can be a sign of biased or fraudulent randomization and can be used as a warning sign. They can be considered as a screening tool with low positive-predictive value. Low p values should prompt us to ask for the reasons and for potential consequences, especially in combination with hints of methodological problems.Contra: A fair randomization produces the expectation that the distribution of p values follows a flat distribution. It does not produce an expectation related to a single p value. The distribution of p values in RCTs can be influenced by the correlation among covariates, differential misclassification or differential mismeasurement of baseline covariates. Given only a small number of reported p values in the reports of RCTs, judging whether the realized p value distribution is, indeed, a flat distribution becomes difficult. If p values ≤0.005 or ≥0.995 were used as a sign of alarm, the false-positive rate would be 5.0% if randomization was done correctly, and five p values per RCT were reported.Conclusion: Use of a low p value as a warning sign that randomization is potentially biased can be considered a vague heuristic. The authors of this debate are obviously more or less enthusiastic with this heuristic and differ in the consequences they propose. Keywords: randomized controlled trial, distribution, statistical, random allocationhttps://www.dovepress.com/imbalance-p-values-for-baseline-covariates-in-randomized-controlled-tr-peer-reviewed-article-CLEPRandomized controlled trialdistributionstatisticalrandom allocation; |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Stang A Baethge C |
spellingShingle |
Stang A Baethge C Imbalance p values for baseline covariates in randomized controlled trials: a last resort for the use of p values? A pro and contra debate Clinical Epidemiology Randomized controlled trial distribution statistical random allocation; |
author_facet |
Stang A Baethge C |
author_sort |
Stang A |
title |
Imbalance p values for baseline covariates in randomized controlled trials: a last resort for the use of p values? A pro and contra debate |
title_short |
Imbalance p values for baseline covariates in randomized controlled trials: a last resort for the use of p values? A pro and contra debate |
title_full |
Imbalance p values for baseline covariates in randomized controlled trials: a last resort for the use of p values? A pro and contra debate |
title_fullStr |
Imbalance p values for baseline covariates in randomized controlled trials: a last resort for the use of p values? A pro and contra debate |
title_full_unstemmed |
Imbalance p values for baseline covariates in randomized controlled trials: a last resort for the use of p values? A pro and contra debate |
title_sort |
imbalance p values for baseline covariates in randomized controlled trials: a last resort for the use of p values? a pro and contra debate |
publisher |
Dove Medical Press |
series |
Clinical Epidemiology |
issn |
1179-1349 |
publishDate |
2018-05-01 |
description |
Andreas Stang,1,2 Christopher Baethge3,4 1Center of Clinical Epidemiology, Institute of Medical Informatics, Biometry and Epidemiology, Medical Faculty, University Hospital of Essen, Hufelandstr, Essen, Germany; 2Department of Epidemiology, School of Public Health, Boston University, Boston, MA, USA; 3Department of Psychiatry and Psychotherapy, University of Cologne Medical School, Cologne, Germany; 4Editorial Offices, Deutsches Ärzteblatt and Deutsches Ärzteblatt International, Deutscher Ärzte-Verlag, Cologne, Germany Background: Results of randomized controlled trials (RCTs) are usually accompanied by a table that compares covariates between the study groups at baseline. Sometimes, the investigators report p values for imbalanced covariates. The aim of this debate is to illustrate the pro and contra of the use of these p values in RCTs.Pro: Low p values can be a sign of biased or fraudulent randomization and can be used as a warning sign. They can be considered as a screening tool with low positive-predictive value. Low p values should prompt us to ask for the reasons and for potential consequences, especially in combination with hints of methodological problems.Contra: A fair randomization produces the expectation that the distribution of p values follows a flat distribution. It does not produce an expectation related to a single p value. The distribution of p values in RCTs can be influenced by the correlation among covariates, differential misclassification or differential mismeasurement of baseline covariates. Given only a small number of reported p values in the reports of RCTs, judging whether the realized p value distribution is, indeed, a flat distribution becomes difficult. If p values ≤0.005 or ≥0.995 were used as a sign of alarm, the false-positive rate would be 5.0% if randomization was done correctly, and five p values per RCT were reported.Conclusion: Use of a low p value as a warning sign that randomization is potentially biased can be considered a vague heuristic. The authors of this debate are obviously more or less enthusiastic with this heuristic and differ in the consequences they propose. Keywords: randomized controlled trial, distribution, statistical, random allocation |
topic |
Randomized controlled trial distribution statistical random allocation; |
url |
https://www.dovepress.com/imbalance-p-values-for-baseline-covariates-in-randomized-controlled-tr-peer-reviewed-article-CLEP |
work_keys_str_mv |
AT stanga imbalancepvaluesforbaselinecovariatesinrandomizedcontrolledtrialsalastresortfortheuseofpvaluesaproandcontradebate AT baethgec imbalancepvaluesforbaselinecovariatesinrandomizedcontrolledtrialsalastresortfortheuseofpvaluesaproandcontradebate |
_version_ |
1725353431072243712 |