Scientific fraud in 20 falsified anesthesia papers : detection using financial auditing methods

Area of Science:

• Medical research integrity
• Data analysis and statistics

Background:

• Benford's Law describes a predictable pattern in the leading digits of naturally occurring numerical data.
• This law is widely used in financial and tax auditing to detect fraudulent activities.
• Anesthesiology research is vulnerable to issues like plagiarism, ghostwriting, and data counterfeiting.

Purpose of the Study:

• To investigate whether Benford's Law can identify falsified data in anesthesiology publications.
• To analyze 20 known retracted anesthesiology studies for deviations from Benford's Law.

Main Methods:

• Application of the chi-squared (χ(2)) test and the Z-test to analyze the distribution of first and second digits in 20 retracted anesthesiology publications.
• Comparison of observed digit distributions against the expected distribution patterns defined by Benford's Law.
• A control meta-analysis of non-falsified data was used for comparison.

Main Results:

• The 20 retracted publications exhibited significant deviations from Benford's Law for both first and second digits (p<0.01).
• 17 out of 20 studies showed significant first-digit deviations, and 18 out of 20 showed significant second-digit deviations.
• A control meta-analysis confirmed a distribution consistent with Benford's Law.

Conclusions:

• The analysis suggests that Benford's Law is a sensitive tool for detecting potential fraud in anesthesiology research.
• Deviations from Benford's Law in published data may indicate falsification.
• Further research is needed to confirm specificity and explore conformity in non-falsified data to prevent publication of fraudulent studies.