RESUMEN
A prospective randomized double blind study was carried out on 60 patients undergoing elective gastrointestinal endoscopy. Patients in group A were sedated with diazepam (0.15 mg/kg) and those in group B received midazolam (0.10 mg/kg). There were no differences between the groups in age, sex, previous drug intake, level of anxiety prior to the procedure and from the endoscopist point of view the easiness to perform the studies. The group of patients who received midazolam (B) showed a significantly greater sedation, had less pain in the site of injection, their retrograde amnesia was more effective and tolerated much better the overall procedure (p < 0.05). There were no complications in either group. The cost per dosage was 2:1 greater with midazolam. It is concluded that midazolam has some advantages over diazepam used for sedation of patients undergoing gastrointestinal endoscopy. The time for recovery and cost are its disadvantages.