[Probabilities cannot be calculated retrospectively--not even in the courtroom]

Area of Science:

• Medical Statistics
• Legal Medicine
• Probability Theory

Background:

• Disease clusters and chance events frequently trigger suspicions of hidden causes.
• Legal cases have involved statistical 'predictions' to assign accountability for unexplained health occurrences.
• The misinterpretation of statistical probabilities has led to wrongful convictions.

Purpose of the Study:

• To examine the role of statistical arguments in legal and medical contexts.
• To differentiate between coincidence and causation in disease patterns.
• To emphasize the need for evidence-based reasoning in interpreting statistical data.

Main Methods:

• Analysis of legal case precedents involving statistical evidence.
• Review of expert witness testimony and its impact on judicial decisions.
• Discussion of the limitations of probability in establishing causality.

Main Results:

• Statistical arguments alone were rejected in Dutch legal cases without supporting evidence.
• A UK case highlighted the dangers of relying on retrospective probabilities, leading to a wrongful conviction.
• The absence of causal evidence renders statistical predictions unreliable in legal and medical settings.

Conclusions:

• Interpreting coincidences requires identifying plausible causal factors supported by new evidence, not just statistical probabilities.
• Physicians, judges, and politicians must exercise scrupulous reasoning when evaluating probabilities.
• Mathematical solutions alone cannot resolve the issue of chance events in disease patterns.