Quality control in clinical trials can comprise several different domains (Box). In many trials, this may primarily include ensuring adherence to study protocols and data integrity. Procedural trials introduce an additional difficulty, ie, the variability inherent in procedures performed by different health care professionals in heterogeneous environments for slightly variable clinical conditions.
Data integrity is fundamental to the performance of a quality clinical trial. Intentional falsifications are much less common than unintentional errors and oversights. Good data begin with accurate data entry and are greatly facilitated by electronic medical record systems and automated data transfer. Participants can directly enter their pain scores or questionnaire answers into electronic systems, thereby obviating the chance for transcriptional errors. A surveillance strategy to prevent missing data should be created, as well as a plan for the statistical handling of missing data. Data integrity should be maintained by periodic audits verifying the recorded study data against original source documents. The timing and frequency of audits should be determined before study commencement, with a frequency and rigor appropriate to the risks and complexity of the study.