In theory 100% audit with source data verification should lead to a 100% detection of research fraud. In practice 100% audit is impossible within the bounds of reasonable costs. Most companies therefore have decided to introduce sampling methods, which by definition guarantee less than 100% verification.
company and specifies what action the company will take if it suspects fraud has been perpetrated against it
• A declaration that the company is willing to prosecute all cases of fraud
• The training of clinical research associates, trial monitors, data managers, and medical advisers to be alert to the telltale signs of fraud
• The building in of routine check statistics to the analysis of data
• The introduction of a standard operating procedure, signed by all levels of management, both commercial and clinical, which outlines unambiguously the steps to be followed in cases of suspected clinical fraud.
The next major step forward must be the generation and sharing of details of cases of suspected fraud among companies, the trade associations, and the boards of health and ethics committees. This, sadly, remains something for the future but how good it would be if it came to pass.
Was this article helpful?