Blinding in randomised controlled trials of surgical procedures in children: who, when and how

Abstract text
Background: Blinding is the process by which people are kept unaware of trial intervention allocations after randomisation. It is designed to prevent performance bias and detection bias. Conducting trials in surgery and in paediatrics is acknowledged as particularly challenging. By studying how trials of surgical procedures in children have evolved to overcome the difficulties of implementing blinding, all specialties can potentially learn valuable lessons.

Objectives: To describe, who is blinded, as well as when and how blinding occurs in randomised controlled trials of surgical procedures in children.

Methods: We searched the Cochrane Central Register of ControlLed Trials (CENTRAL) on the 10/02/2012 and selected randomised controlled trials only involving children (<18 years) undergoing surgical procedures published from the years 2000 to 2001 and 2010 to 2011. We assessed the epidemiological and methodological of these trials, we analysed risk of bias using the Cohrane handbook and conducted qualitative analyses of the blinding process.

Results: We included 112 randomised controlled trials. The median sample size per trial was 54 (IQR: 38 to 123) and the mean patient age was 5.7 years (95%CI: 3.4 to 8.0). Blinding was implemented in some form in at least 48/112 trials (42.9%), through a variety of means. Many innovative blinding techniques were noted, see Table 1. The people most often blinded were the data collectors and analysts (36.7%). No trial described testing the adequacy of blinding. There was no significant improvement in the blinding of participants and personnel or outcome assessors over time (P=0.06 & P=0.13 respectively), see Table 2.

Conclusions: Blinding is an important component of randomised controlled trials. Despite the difficulties faced, this study demonstrates that blinding in trials of surgical procedures in children is feasabile. Our study also provides many practical examples of how blinding can be achieved, which is useful for those conducting trials.
Authors
McGee R1, Hua M1, Craig J2, Webster A2
1 Sydney School of Public Health, University of Sydney
2 Cochrane Renal Group, Sydney
Presenting author and contact person
Presenting author: 
Richard McGee
Contact person Affiliation Country
Richard McGee (Contact this person) Sydney School of Public Health Australia
Attachments
Attachments: 
Date and Location
Session: 
Oral session A4O2
Date: 
Monday 1 October 2012 - 11:20 - 11:40
Location: