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

Statistical issues in studies of individual response.

D J Spiegelhalter1

  • 1MRC Biostatistics Unit, Cambridge.

Scandinavian Journal of Gastroenterology. Supplement
|January 1, 1988
PubMed
Summary

This study differentiates between single-patient (N=1) pragmatic trials and group (N>1) explanatory trials for assessing therapy response. It highlights the importance of individual patient data analysis and permutation tests for variable 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

Statistics. The future lies in uncertainty.

Science (New York, N.Y.)·2014
Same author

The power of the MicroMort.

BJOG : an international journal of obstetrics and gynaecology·2014
Same author

Using routine intelligence to target inspection of healthcare providers in England.

Quality & safety in health care·2009
Same author

How to interpret your dot: decoding the message of clinical performance indicators.

Journal of perinatology : official journal of the California Perinatal Association·2008
Same author

Flexible random-effects models using Bayesian semi-parametric models: applications to institutional comparisons.

Statistics in medicine·2006
Same author

Intracortically distributed neurovascular coupling relationships within and between human somatosensory cortices.

Cerebral cortex (New York, N.Y. : 1991)·2006

Area of Science:

  • Biostatistics
  • Clinical Trial Design
  • Pharmacology

Background:

  • Individual patient responses to therapies can vary significantly.
  • Aggregate measures of treatment effect are often inappropriate for highly variable responses.
  • Distinguishing between study types is crucial for accurate interpretation.

Purpose of the Study:

  • To differentiate between strict 'N = 1' and 'N much greater than 1' study designs.
  • To discuss design, measurement, and statistical significance in intensive individual response studies.
  • To emphasize the utility of permutation tests for analyzing variable treatment effects.

Main Methods:

  • Controlled experimental design.
  • Focus on intensive studies of individual patient response.
  • Application of permutation tests for statistical significance.

Main Results:

  • 'N = 1' studies are pragmatic, focusing on the individual patient.
  • 'N much greater than 1' studies are explanatory, aiming for generalizable statements.
  • Permutation tests are emphasized for analyzing highly variable responses.

Conclusions:

  • Careful distinction between study types (N=1 vs. N>1) is essential for interpreting therapy response.
  • Individualized analysis is key when treatment effects are highly variable.
  • Permutation tests offer a robust statistical approach for such studies.

Related Experiment Videos