The effect of scientific misconduct on the results of clinical trials: a Delphi survey.

OBJECTIVES: To discover what types of scientific misconduct are most likely to influence the results of a clinical trial. DESIGN: Delphi survey of expert opinion with three rounds of consultation. SETTING: Non-industry clinical trial "community". PARTICIPANTS: Experts identified from invitees to a previous MRC consultation on clinical trials. 32 out of the 40 experts approached agreed to participate. RESULTS: We identified thirteen forms of scientific misconduct for which there was majority agreement (>50%) that they would be likely or very likely to distort the results and majority agreement (>50%) that they would be likely or very likely to occur. Of these, the over-interpretation of 'significant' findings in small trials, selective reporting and inappropriate subgroup analyses were the main themes. CONCLUSIONS: According to this expert group, the most important forms of scientific misconduct in clinical trials are selective reporting and the opportunistic use of the play of chance. Data fabrication and falsification were not rated highly because it was considered that these were unlikely to occur. Registration and publication of detailed clinical trial protocols could make an important contribution to preventing scientific misconduct.