Evidence-based medical practice: as viewed by a clinical researcher.

In this era of information technology, clinicians need to understand how to conduct ethical and effective clinical research. The principles of 'good clinical practice' have been established to prevent mistakes and malpractice, and to protect research subjects. A researcher must carefully design the study and perform a thorough review of the literature. A detailed protocol that has been approved by the institution's Ethics Committee must be constructed before commencement of the study and followed accordingly. Comparable control group(s), recruited in a double-blind fashion, ensure that the effects observed are specific to the intervention itself. While a retrospective data review is sometimes useful, only prospective studies provide solid evidence. Outcome parameters must be defined that are clinically appropriate and measurable. Subjects should only be recruited with consent and in a randomised fashion. Elimination of inter- and intra-observer errors can be achieved by appointing an assessor (or team) and by training personnel. Proper documentation of all the events and outcome parameters is of paramount importance and is a requirement for good clinical practice. Data analysis should be done by a qualified and experienced biostatistician and the personnel involved in the study must be fully aware of their moral, legal, and scientific responsibilities and be able to work with each other.