Design of the multicenter Australian study of epidural anesthesia and analgesia in major surgery: The MASTER Trial

Citation
Jra. Rigg et al., Design of the multicenter Australian study of epidural anesthesia and analgesia in major surgery: The MASTER Trial, CONTR CL TR, 21(3), 2000, pp. 244-256
Citations number
26
Categorie Soggetti
Pharmacology,"Medical Research General Topics
Journal title
CONTROLLED CLINICAL TRIALS
ISSN journal
01972456 → ACNP
Volume
21
Issue
3
Year of publication
2000
Pages
244 - 256
Database
ISI
SICI code
0197-2456(200006)21:3<244:DOTMAS>2.0.ZU;2-C
Abstract
The Multicenter Australian Study of Epidural Anesthesia and Analgesia in Ma jor Surgery (The MASTER Trial) was designed to evaluate the possible benefi t of epidural block in improving outcome in high-risk patients. The trial b egan in 1995 and is scheduled to reach the planned sample size of 900 durin g 2001. This paper describes the trial design and presents data comparing 4 55 patients randomized in 21 institutions in Australia, Hong Kong, and Mala ysia, with 237 patients from the same hospitals who were eligible but not r andomized. Nine categories of high-risk patients were defined as entry crit eria for the trial. Protocols for ethical review, informed consent, randomi zation, clinical anesthesia and analgesia, and perioperative management wer e determined following extensive consultation with anesthesiologists throug hout Australia. Clinical and research information was collected in particip ating hospitals by research staff who may not have been blind to allocation . Decisions about the presence or absence of endpoints were made primarily by a computer algorithm, supplemented by blinded clinical experts. Without unblinding the trial, comparison of eligibility criteria and incidence of e ndpoints between randomized and nonrandomized patients showed only small di fferences. We conclude that there is no strong evidence of important demogr aphic or clinical differences between randomized and nonrandomized patients eligible for the MASTER Trial. Thus, the trial results are likely to be br oadly generalizable. Control Clin Trials 2000;21:244-256 (C) Elsevier Scien ce Inc. 2000.