Two-thirds of clinical studies in cardiology spin nonsignificant findings

  • Khan MS & al.
  • JAMA Netw Open
  • 3 May 2019

  • International Clinical Digest
Access to the full content of this site is available only to registered healthcare professionals. Access to the full content of this site is available only to registered healthcare professionals.

Takeaway

  • Authors writing up results of cardiovascular randomized controlled trials (RCTs) often spin statistically nonsignificant primary findings in ways that do not match what the data say. 
  • These authors say 67% of the main text of published articles contain some kind of spin to gloss findings.

Why this matters

  • Authors say clinicians should be aware that peer review does not always catch spin.  
  • Editor-in-chief of JAMA Network Open, which published these findings, commented that “we live in an age of spin,” and that the scientific literature is not immune.
  • Accompanying editorial says such spin can lead to “exuberant adoption of marginally effective, useless, or even harmful clinical interventions” and create patient anxiety.

Key results

  • 57% (95% CI, 47%-67%) of abstracts contained spin.
  • 67% (95% CI, 57%-75%) of main text articles contained spin.
  • 11% (95% CI, 6%-19%) had spin in their titles.
  • 54% (95% CI, 44%-64%) had spin in the conclusions.
  • Spin included methods misreporting, shifting objectives/hypothesis to match findings, muddying prior vs post hoc analyses, selective report, misleading interpretation, P value misinterpretation, unfounded extrapolation.

Study design

  • Analysis of 93 “high-impact” RCTs (January 2015-December 2017) with nonsignificant findings for a primary outcome.
  • Funding: None disclosed.

Limitations

  • Short publication time frame, a few select journals represented.