Biostatistics journal club: p-Values and the New England Journal of Medicine (NEJM)

David Harrington, PhD, Harvard T.H. Chan School of Public Health, New England Journal of Medicine (NEJM), will review the NEJM 2019 statistical reporting guidelines and accompanying editorial with a focus on elimination of all p-values in manuscripts except those for a pre-specified primary analysis and multiplicity-adjusted analyses of secondary outcomes. Discussion will be based around the changing view of p-values as reflected in the ASA statement on the use of p-values, and the implications for both undergraduate and graduate training in statistics.

Registration required.

David Harrington, PhD
Professor of Biostatistics, Emeritus, Harvard T.H. Chan School of Public Health
Professor of Statistics, Emeritus, Harvard University Faculty of Arts & Sciences
Lead Statistical Editor, New England Journal of Medicine

p-Values and the New England Journal of Medicine (NEJM)

Abstract
This talk will discuss the NEJM 2019 statistical reporting guidelines and the accompanying editorial. The 2019 guidelines discussed several issues, including the handling of missing data, the importance of the fidelity of an analysis to study design, and the reporting of subgroup effects, but the decision given the most attention was the elimination of all p-values in manuscripts except those for a pre-specified primary analysis and multiplicity-adjusted analyses of secondary outcomes. More generally, the talk will discuss the changing view of p-values as reflected in the ASA statement on the use of p-values, and the implications for both undergraduate and graduate training in statistics. The talk will also cover the issues likely to be covered in the upcoming revision of the guidelines.

Useful References

• New Guidelines for Statistical Reporting in NEJM https://www.nejm.org/doi/full/10.1056/NEJMe1906559
• ASA Statement on p-values