The importance of large, high quality clinical databases in nephrology.

While clinical trials constitute an undisputed standard, they are not perfect because they are focused on a well defined segment of patients with a given disease. Generalizability is a problem with these studies because the experimental setting often differs substantially from everyday clinical practice. Clinical files are traditionally the main source of information for clinical medicine. The quality of recorded information in clinical files is often unsatisfactory. Yet clinical files are data-rich and ought to be considered as a valuable resource, particularly when clinical information is stored in electronic format. Clinicians have already started to construct very large, high quality clinical databases. Outcome research based on high quality data retrieved from well organized and exhaustive clinical files is bound to become increasingly important. Standardization is fundamental for these projects to be successful. The task may be daunting but worth while; due to complexity and cost the randomised clinical trial is unlikely to remain the only source of information on which to base clinical decisions about treatments.