Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Reporting Bayesian analyses of clinical trials

M D Hughes1

  • 1Department of Biostatistics, Harvard School of Public Health, Boston, MA 02115.

Statistics in Medicine
|September 30, 1993
PubMed
Summary

Bayesian analysis offers a valid alternative for interpreting clinical trial results, moving beyond common misinterpretations of p-values and confidence intervals. This approach updates pre-trial beliefs to post-trial probabilities, enhancing clinical understanding of treatment effects.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

HIV testing uptake among the household contacts of multidrug-resistant tuberculosis index cases in eight countries.

The international journal of tuberculosis and lung disease : the official journal of the International Union against Tuberculosis and Lung Disease·2019
Same author

Resource utilization for multidrug-resistant tuberculosis household contact investigations (A5300/I2003).

The international journal of tuberculosis and lung disease : the official journal of the International Union against Tuberculosis and Lung Disease·2018
Same author

Antiretroviral regimens in pregnancy and breast-feeding in Botswana.

The New England journal of medicine·2010
Same author

Predictors of growth and body composition in HIV-infected children beginning or changing antiretroviral therapy.

HIV medicine·2010
Same author

Safety and efficacy of nevirapine- and efavirenz-based antiretroviral treatment in adults treated for TB-HIV co-infection in Botswana.

The international journal of tuberculosis and lung disease : the official journal of the International Union against Tuberculosis and Lung Disease·2009
Same author

Magnetic resonance imaging as a surrogate outcome for multiple sclerosis relapses.

Multiple sclerosis (Houndmills, Basingstoke, England)·2008

Area of Science:

  • Biostatistics
  • Clinical Trial Methodology
  • Medical Informatics

Background:

  • Clinicians frequently misinterpret frequentist statistical measures like p-values and confidence intervals.
  • Common misinterpretations include viewing p-values as the probability of adverse treatment effects or confidence intervals as probability intervals.
  • Bayesian analysis provides a framework for validly drawing such inferences by updating prior beliefs with trial data.

Purpose of the Study:

  • To propose Bayesian methods for reporting clinical trial results that are more intuitively understood by clinicians.
  • To address the need for consensus between statisticians and clinicians on choosing appropriate prior distributions for trial reporting.
  • To demonstrate how Bayesian analyses can provide post-trial probability distributions of treatment effects.

Related Experiment Videos

Main Methods:

  • Utilized Bayesian inference to update prior beliefs about treatment effects using clinical trial data.
  • Explored two types of prior distributions: non-informative priors and priors with a spike at the point of no effect.
  • Emphasized the use of graphical displays to facilitate clinician exploration of posterior probability distributions.

Main Results:

  • Non-informative priors yield standardized likelihood distributions as post-trial probability distributions, potentially appealing to clinicians.
  • Priors with a spike at the point of no treatment effect allow for the illustration of sensitivity to prior skepticism.
  • Graphical displays enhance the interpretability of Bayesian trial results for clinical readers.

Conclusions:

  • Bayesian analysis offers a more interpretable approach to clinical trial reporting compared to traditional frequentist methods.
  • The proposed methods, particularly the use of non-informative priors and informative priors with skepticism, can improve clinician understanding.
  • Wider adoption requires consensus on prior distribution selection, but graphical tools can aid in exploring results and facilitating acceptance.