Quantifying Undetected Severe Misconduct in Medical Research

Scientific misconduct in the medical and health sciences remains a major concern for practitioners, patients, scientists, and funders. Even as retractions of fraudulent articles have increased sharply in recent years, concerns remain that detected cases represent only the tip of the iceberg and that actual prevalence of misconduct in the scientific system may be much higher. In this study, we estimate the size of the hidden population of scientists associated with articles that have been retracted for severe misconduct by applying capturerecapture methods. Distinguishing first authors and last authors associated with retracted articles, we find a capture rate of approximately 23% and a population of around 4,250 first authors associated with articles that have been retracted due to severe misconduct in the medical and health sciences. Among last authors, we find a capture rate of 34% and a population of around 4,125 scientists. Our results imply that most fraud in the medical and health sciences remains undetected, yet suggest that only a small fraction of the scientific workforce is associated with articles affected by severe scientific misconduct.

Hussinger, Katrin and Maikel Pellens (2026), Quantifying Undetected Severe Misconduct in Medical Research, ZEW Discussion Paper No. 26-048, Mannheim.