[Is the randomized controlled trial overvalued as a basis for clinical decision-making? A review with comments]

The randomized controlled trial (RCT) may have considerable limitations in clinical research. Lacking the possibility of blinding impairs the internal validity of the trials. The external validity is often impaired, as results of RCTs obtained in an ideal situation, may be difficult to generalize to a clinical routine situation. Pragmatic randomized trials move from ideal situations towards routine situations, and by modifying the design it is possible to reduce selection bias due to patient and physician preferences. Quasi-experimental studies have varying degrees of problems with internal validity but are necessary contributions to our knowledge of the effect of treatment in clinical routine situations. Limitations of the usefulness of RCTs as well as pragmatic and quasi-experimental studies in clinical research make it necessary to recognise that different methods complement one another. Research in development of RCTs and new methods in clinical research should be encouraged.