Relaxing the Rule of Ten Events per Variable in Logistic and Cox Regression

Dados Bibliográficos

ID23364674
AutoresEric Vittinghoff (0000-0001-8535-0920, University of California, San Francisco), Charles E Mcculloch (0000-0002-1279-6179, University of California, San Francisco)
Ano2007
Volume165
Fascículo6
Páginas710-718
Data de publicação2007-01-12
Peer ReviewedSim
Open AccessNão
TipoARTICLE
PeriódicoAmerican Journal of Epidemiology (JOURNAL)
Identificadores do periódicoISSN: 0002-9262 • E-ISSN: 1476-6256
EditoraOxford University Press (OUP) (PUBLISHER)
DOI10.1093/aje/kwk052
PMID17182981
OpenAlexW2130373985
IdiomaEN
Citações recebidas219
Referências citadas4

The rule of thumb that logistic and Cox models should be used with a minimum of 10 outcome events per predictor variable (EPV), based on two simulation studies, may be too conservative. The authors conducted a large simulation study of other influences on confidence interval coverage, type I error, relative bias, and other model performance measures. They found a range of circumstances in which coverage and bias were within acceptable levels despite less than 10 EPV, as well as other factors that were as influential as or more influential than EPV. They conclude that this rule can be relaxed, in particular for sensitivity analyses undertaken to demonstrate adequate control of confounding.

Confidence interval · Confounding · Econometrics · Logistic regression · Range (aeronautics) · Regression · Regression analysis · Rule of thumb · Statistics · Type I and type II errors · Variable (mathematics) · Variables · Advanced Causal Inference Techniques · Engineering · Mathematics · Medicine · Statistical Methods and Bayesian Inference · Statistical Methods and Inference