Clinical trials bias problems

At one end of the spectrum, the event may be a simple dosage problem which could be an error on the part of the prescriber or an unanticipated hypersensitivity for that particular patient. At the other end of the spectrum, is an uncommon, serious adverse reaction not revealed in premarketing clinical trials. Somewhere between those two extremes are more or less serious adverse events which are not entirely unexpected but appear to be more common than is accepted for comparable products in the same therapeutic category. This maybe a real increase in frequency or may be due to patient selection bias. The later has arisen with new products which claim a lower incidence of certain adverse reactions which encourages doctors to precribe them preferentially for patients who have suffered such reactions with older products. 

---