Clinical evaluation of performance of target controlled infusion system of midazolam for sedation during operation / 中华麻醉学杂志

ABSTRACT

ve To evaluate the performance of the target controlled infusion (TCI) system of midazolam for sedation during operation in terms of safety and ease of use in the elderly and the young. Methods Forty-four unpremedicated ASA I - II patients scheduled for elective surgery under epidural or combined spinal-epidural (CSE) anesthesia were included in this study. The patients were divided into two age groups the elderly group aged 61-82 yr ( n = 22) and the young group aged 19-43 yr (n = 22) . Patients with neurological or severe cardiopulmonary diseases or abnormal liver or kidney function were excluded. The block height was maintained below T4. The TCI system consisted of Intel Pentium III 450 MHz computer and Graseby 3500 computerized infusion pump. The software we used was Stelpump Version 1.05 written by Pina and Coetzee and the pharmacokinetic model and parameter were described by Arram and Buhrer. The effect site concentration of midazolam was targeted and started from 50ng?ml-1, then increased in increment of 50 ng?ml-1 (in the young group) or 25ng?ml-1 (in the elderly group) until the patient did not respond to light shaking (defined as unconsciousness) . Each incremental increase was maintained for 15 min. Arterial blood samples were taken in ten patients randomly selected from each group at 10 and 15min after each incremental increase of midazolam concentration for determination of plasma midazolam concentration by high performance liquid chromatography ( HPLC) .Results The bias (median performance error) of the TCI system of midazolam was 7.9% in the elderly group and 11.6% in young groups. The precision (median absolute performance error) was 20.0% in both groups. They were clinically acceptable. Conclusions The TCI system of midzolam can provided sedation for patients undergoing surgery under spinal and/or epidural anesthesia safely and efficiently in both the old and young.