Open Access Open Access  Restricted Access Subscription Access

Comparison of Regression Models for Binary Outcome Variables in Clinical Trials


Affiliations
1 Department of Statistics, ICMR-National Institute for Research in Tuberculosis, No. 1, Mayor Sathiyamoorthy Road, Chetpet, Chennai 600 031, India
 

The widely used logistic regression may not be suitable to model binary outcomes in clinical trials. The present study compares and assesses various binary regression models such as logistic, log-binomial, Poisson and Cox proportional models for clinical trials. A dataset obtained from a clinical trial conducted on tuberculosis patients is used to illustrate the models. The estimated odds ratios from logistic regression severely overestimated the relative risks, thereby overestimating the overall relationship. Log-binomial and Poisson with sandwich covariance estimator were found to be suitable for estimating adjusted relative risks in clinical trials.

Keywords

Binary Outcomes, Clinical Trials, Regression Models, Relative Risk.
User
Notifications
Font Size

Abstract Views: 7

PDF Views: 0




  • Comparison of Regression Models for Binary Outcome Variables in Clinical Trials

Abstract Views: 7  |  PDF Views: 0

Authors

Adhin Bhaskar
Department of Statistics, ICMR-National Institute for Research in Tuberculosis, No. 1, Mayor Sathiyamoorthy Road, Chetpet, Chennai 600 031, India
Chinnaiyan Ponnuraja
Department of Statistics, ICMR-National Institute for Research in Tuberculosis, No. 1, Mayor Sathiyamoorthy Road, Chetpet, Chennai 600 031, India

Abstract


The widely used logistic regression may not be suitable to model binary outcomes in clinical trials. The present study compares and assesses various binary regression models such as logistic, log-binomial, Poisson and Cox proportional models for clinical trials. A dataset obtained from a clinical trial conducted on tuberculosis patients is used to illustrate the models. The estimated odds ratios from logistic regression severely overestimated the relative risks, thereby overestimating the overall relationship. Log-binomial and Poisson with sandwich covariance estimator were found to be suitable for estimating adjusted relative risks in clinical trials.

Keywords


Binary Outcomes, Clinical Trials, Regression Models, Relative Risk.



DOI: https://doi.org/10.18520/cs%2Fv119%2Fi12%2F2010-2013