Erschienen in:
01.08.2014 | Original Article
Significant and Safe Shortening of the Recovery Time After Flumazenil-Reversed Midazolam Sedation
verfasst von:
Elisabeth M. H. Mathus-Vliegen, Linda de Jong, Hedwig A. Kos-Foekema
Erschienen in:
Digestive Diseases and Sciences
|
Ausgabe 8/2014
Einloggen, um Zugang zu erhalten
Abstract
Background
Endoscopy under midazolam sedation requires a 2-h recovery facility.
Aim
To study the potential of shortening patients’ stay without jeopardizing patients’ safety by the use of the benzodiazepine-antagonist flumazenil in the everyday practice and to investigate the feasibility of a study comparing midazolam with recovery with midazolam–flumazenil and immediate discharge.
Methods
Consecutive ambulatory patients referred for endoscopy under midazolam sedation with ASA I or II, escorted by a person, were eligible. Flumazenil was given on arrival in the recovery room. Patients were discharged when adequate Aldrete scores and physical mobility were present. The next day, they were contacted by telephone.
Results
A total of 1,506 patients participated. They received 5 mg midazolam, while 887 patients also received 50 mcg fentanyl. The median dose of flumazenil was 0.2 mg. Oxygen desaturation (sO2 <92 %) occurred in 15 % during the procedure without an effect on recovery and discharge times. Patients left the department 65 min after the last midazolam administration. The majority (82.7 %) were fully alert during their journey home. At home, 2.7 % went to bed, 45.2 % took a nap, and 40 % undertook activities. Almost every patient (98.8 %) liked the shortened recovery time. Three patients had an incident (fainting, fall, and near-fall) without consequences. Based on this low incidence, a non-inferiority comparison of midazolam–flumazenil with midazolam-recovery would require a total of 32,650 patients.
Conclusions
Administration of flumazenil resulted in a safe shortening of the recovery period and offers the possibility for substantial savings in time, space, and nurse resources. A non-inferiority comparison will not be practicable.