Preclinical Statin Studies Meta-Analysis on Sponsorship Bias

63 preclinical animal studies on statins and atherosclerosis outcomes were meta-analyzed to assess sponsorship bias. The dataset includes extracted study design criteria, efficacy outcomes, sponsorship source, and investigator financial ties, covering literature from 1966 to 2012. It was created by David Krauth to examine differences in standardized mean difference (SMD) estimates between industry and non-industry sponsored research.

Use Cases

• Analyze the association between sponsorship source and the standardized mean difference (SMD) of statin efficacy outcomes.
• Model the relationship between study design criteria (e.g., blinding, randomization) and reported effect sizes.
• Investigate correlations between the disclosure of investigator financial ties and the risk of bias metrics.
• Compare pooled average SMD estimates across subgroups defined by outcome measured and optimal time window for assessment.

Strengths

• Includes 63 distinct preclinical studies, providing a substantive sample for meta-analysis.
• Data spans a 46-year literature search period (1966-2012), offering historical perspective.
• Key variables like sponsorship source and standardized mean difference (SMD) are explicitly coded for analysis.

Limitations

• Sample size is limited to 63 studies, which may constrain statistical power for some subgroup analyses.
• Data is temporally stale, with the literature search ending in 2012 and no updates indicated since 2020.
• Potential for publication bias and selective reporting in the underlying studies, as noted by the authors.

Provenance

Source

Dryad digital repository.

Collection Method

Meta-analysis of 63 studies identified via Medline search, with independent coding by two reviewers.

Time Range

Literature covered from January 1966 to April 2012.

Freshness

Last updated 2020-06-24; no update frequency specified.