Cited by 102Statistical Methods in Clinical TrialsStatistical Methods in EpidemiologyMeta-analysis and systematic reviews

Underpowering in randomized trials reporting a sample size calculation

Vickers, Andrew J·2003·Journal of Clinical Epidemiology

Abstract

Objective The objective of this study was to determine whether standard deviations (SDs) used in sample size calculations are smaller than those found in the resulting study sample, thereby leading to underpowered studies.Method The predicted SD used in the sample size calculation and the actual SD of the study sample were recorded for randomized trials recently published in one of four major journals.Results Sample SD was greater than predicted SD for 80% of endpoints. About one quarter of trials required five times as many patients as specified in the sample size calculation.Conclusion Trials reporting sample size calculations for continuous endpoints published in the most reputable medical journals are often underpowered. There seems to be insufficient understanding that the SD of a sample of patients is a random variable, associated with imprecision, that cannot easily be extrapolated from one population to another.

Integrity signals

Registry

Crossref · 10.1016/s0895-4356(03)00141-0

Cites

16 of 16 references verified0 citation errors · 0 not checkable

Statistics

Recomputing p-values needs the paper’s full text.

Claims the paper makes

This paper’s full text isn’t openly available, so its claims and self-check weren’t run. The registry and citation checks still apply.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • The values of certain prespecified variables tend to be inappropriately inflated (eg, clinically important target difference) or underestimated (eg, standard deviation for continuous outcomes), leading to trials ultimately having less power than originally intended.

    “The values of certain prespecified variables tend to be inappropriately inflated (eg, clinically important target difference) or underestimated (eg, standard deviation for continuous outcomes),295 leading to trials ultimately having less power than originally intended.”

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Authors
Vickers, A. J.
Affiliations
Memorial Sloan Kettering Cancer Center (US)
Corresponding author
Vickers, Andrew J
Venue
Journal of Clinical Epidemiology
Publisher
Elsevier BV
Published
2003-08
Volume
56
Issue
8
Pages
717-720
ISSN
0895-4356 (ISSN-L)
OpenAlex
W1999562241
Cited by
102
References
16
Citations by year
2026: 3 · 2025: 5 · 2024: 6 · 2023: 1 · 2022: 3 · 2021: 4 · 2020: 4 · 2019: 1 · 2018: 6 · 2017: 4 · 2016: 7 · 2015: 6 · 2014: 7 · 2013: 10 · 2012: 3
FWCI
5.16
Type
journal article
Language
en
Open access
closed
Locations
  • Journal of Clinical Epidemiology (publishedVersion)
  • PubMed (publishedVersion)
Topics
  • Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods in Clinical Trials
  • Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods in Epidemiology
  • Social Sciences › Decision Sciences › Statistics, Probability and Uncertainty › Meta-analysis and systematic reviews
Keywords
underpowered trials, sample size determination, standard deviation, statistical power, continuous outcomes, trial reporting
MeSH terms
Data Interpretation, Statistical, Humans, Randomized Controlled Trials as Topic, Sample Size

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review