Format

Send to

Choose Destination
J Surg Res. 2011 Jan;165(1):118-27. doi: 10.1016/j.jss.2009.09.054. Epub 2009 Oct 30.

Design of surgical randomized controlled trials involving multiple interventions.

Author information

1
University Department of Surgery, Royal Free Campus, UCL Medical School, London, United Kingdom. kurinchi2k@hotmail.com

Abstract

BACKGROUND:

Surgical trials evaluating multiple interventions can be carried out by using parallel or factorial trial design. The optimal trial method has not been established.

METHODS:

Trials involving interventions to facilitate day-case laparoscopic cholecystectomy have been chosen to illustrate the strengths and weaknesses of these two trials designs.

RESULTS AND CONCLUSIONS:

The parallel trial design is simple to design and interpret, but requires multiple randomized controlled trials to assess multiple interventions. It does not detect interactions between the interventions. Factorial trial design allows reduction in overall sample size and allows for exploration of interactions. However, it is not suitable when interventions are mutually exclusive, and it is associated with a higher risk of type I and type II statistical errors. Increased sample size is required in trials in which there are group interactions, which are present in a small proportion of the interventions. Both parallel trial design and factorial trial design have their advantages and disadvantages in the assessment of multiple interventions that can be applied in conjunction. Modifications to the factorial trial design appear to overcome many of the shortcomings of factorial trial design but require further increase in sample size to decrease type I and type II errors. In practice, the choice of trial design for multiple surgical interventions depends on the potential costs of type I and type II errors and the risk of missing important antagonistic and negative interactions.

PMID:
20097373
DOI:
10.1016/j.jss.2009.09.054
[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Elsevier Science
Loading ...
Support Center