Abstract
Background: Blinding is recommended widely as a strategy in randomized controlled trials (RCTs) to reduce biases that can result from awareness of who is receiving the intervention being tested. The absence of blinding, especially when the primary outcomes are subjective, has been found to be associated with inflated estimates of treatment effects, yet little is known about the use of blinding in nursing RCTs.
Objectives: The purposes of this study were (a) to describe the extent to which nurse researchers state that they used blinding as a bias-reduction strategy or explain why it was not used, (b) to identify the groups that are blinded when blinding is used, (c) to assess whether the term blinding or masking is more prevalent, and (d) to explore factors that might affect the use or acknowledgement of blinding in nursing trials.
Methods: Data regarding blinding were extracted systematically and coded from a consecutive sample of 199 RCTs published in 16 nursing journals in 2007 to 2009.
Results: Blinding, the term used predominantly in nursing, was reported to have been used in 33% of the studies. Rates of blinding ranged from 2.5% for data analysts to 28% for data collectors. The absence of blinding was discussed as a limitation in only 13% of the studies in which blinding did not occur. The use or mention of blinding was related strongly to whether the journal in which the article was published had endorsed the Consolidated Standards for Reporting Trials guidelines (CONSORT).
Conclusions: Nurse researchers conducting RCTs should be more diligent in following the CONSORT guidelines regarding the use and description of blinding. Although it is often impossible to blind study participants and intervention agents, the blinding of data collectors, data analysts, and nonresearch healthcare staff should be pursued.