Interpreting trial results: time for confidence and magnitude and not P values please

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

The problems with the P value as the single metric to summarize the results of a study are being recognized. It captures a single domain—random error—but it is relatively uninformative about more critical domains for deciding whether the results should be applied to clinical care and policy. Alternatives include the components of the outcomes reported (relevance, magnitude, frailty, and net benefit) and confidence (risk of bias and directness).

Original languageEnglish
Pages (from-to)28-30
Number of pages3
JournalKidney International
Volume95
Issue number1
DOIs
Publication statusPublished - 1 Jan 2019

Keywords

  • clinical trials
  • P-values
  • test statistics

Fingerprint Dive into the research topics of 'Interpreting trial results: time for confidence and magnitude and not P values please'. Together they form a unique fingerprint.

  • Cite this