In the accepted hierarchy of evidence generation, the results of a randomized, controlled clinicaltrial usually take precedence over other forms of evidence such as meta-analysis and observational studies.
The typical clinicaltrial model may not be best suited for generating evidence that a drug is safe and effective against diseases where patient enrollment can be difficult and diseases progress slowly and unevenly over a long period of time, according to Steven Grossman, who worked on the original Orphan Drug Act in 1983 and is now a policy and regulatory consultant.