Midazolam and pethidine versus propofol and fentanyl patient controlled sedation/analgesia for upper gastrointestinal tract ultrasound endoscopy: a prospective randomized controlled trial

Abstract

Background and objectives: The aim of this prospective, randomized study was to compare the standard regimen of midazolam and pethidine administered by the gastroenterologist versus patient controlled sedation with a propofol-fentanyl mixture during upper gastrointestinal tract endoscopic ultrasonography. Our primary end-points were patient satisfaction and patient cooperation assessed by endoscopist.

Methods: Fifty-four consecutive patients, undergoing endoscopic ultrasonography, received sedation with midazolam and pethidine (Group M: n=27) or propofol and fentanyl (Group P: n=27). Group M: pethidine 0.7mg/kg midazolam 0.04mg/kg before examination; boluses of same drugs if the sedation was insufficient plus a sham patient controlled sedation analgesia; Group P: propofol 17mg plus fentanyl 15microg before examination and a patient controlled sedation analgesia pump containing 170mg propofol plus 150microg fentanyl injecting 0.5ml every time the patient pressed the button (no "lock out"). Boluses of 1ml of the same mixture if the sedation was insufficient.

Results: Group M: mean dosage of pethidine and midazolam 88.6 and 5mg, respectively. Group P: mean dosage of propofol and fentanyl 119.7mg and 106microg, respectively. Both groups were similar for duration and difficulty of the procedure, the grade of sedation (Observer's Assessment of Alertness/Sedation Score) and judgement by endoscopist and patient about cooperation and satisfaction. The only difference between groups was about the extra boluses administered during the procedure.

Conclusion: This study demonstrated that a patient controlled sedation analgesia with propofol and fentanyl is an effective and safe technique for upper gastrointestinal tract endoscopic ultrasonography procedures and results in a high level of satisfaction both for patients and operator.