Blinding may be unnecessary, but please divestBMJ 2020; 368 doi: https://doi.org/10.1136/bmj.m255 (Published 23 January 2020) Cite this as: BMJ 2020;368:m255
- Fiona Godlee, editor in chief
Follow Fiona on Twitter @fgodlee
In the hierarchy of evidence, randomised trials are near the top, trumped only by meta-analysis of such trials, with blinding of patients and clinicians firmly established as being key to their validity. But new research published in The BMJ this week casts doubt on the benefits of blinding. Helene Moustgaard and colleagues have meta-analysed 142 Cochrane meta-analyses, incorporating 1153 randomised trials. They conclude that there’s no evidence that a lack of blinding leads to exaggerated estimates of treatment effects (doi:10.1136/bmj.l6802).
This apparent lack of benefit should be seen in the context of other problems with blinding that Rohan Anand and colleagues highlight (doi:10.1136/bmj.l6228). Blinding can be expensive and may hamper recruitment and retention of participants, they say. It can also compromise patients’ safety and render the evidence less applicable to real life care.
As their preferred method for trials, Anand and colleagues champion PROBE: open label but with adequate randomisation, allocation concealment, and blinded objective outcome assessment. Accepting that blinding may not be necessary could bring more existing trials back into the evidence fold, say editorialists Aaron Drucker and An-Wen Chan (doi:10.1136/bmj.m229), and could mean that more randomised trials are undertaken and successfully completed, especially in areas of healthcare where trials have been deemed hard to do. But they also caution that the effect of blinding in any particular trial is influenced by contextual factors that we don’t yet fully understand. So “until we have further evidence, use of blinding should remain the default standard for protection against performance and detection biases.”
A re-evaluation of the role of blinding may be especially welcome in surgery, where it presents a particular challenge. Two recent articles look at the evidence from trials of surgical interventions for knee arthritis. Both deliver useful clinical messages. John Orchard looks at the uncertainty around the benefits of intra-articular corticosteroid injections (doi:10.1136/bmj.l6923). He finds only limited short term benefit and the potential for long term harm. Exercise and weight loss remain the mainstays of management. And the latest in our series of NIHR Signals concludes that partial knee replacement offers similar clinical outcomes to total knee replacement but with better cost effectiveness because of reduced stays in hospital (doi:10.1136/bmj.l5994).
Also this week we launched a campaign for divestment from fossil fuels (doi:10.1136/bmj.m167). The BMJ joins some major medical colleges and the BMA in committing to divest, and we encourage health professionals and organisations to follow suit. Guidance on how to divest is provided courtesy of the UK Health Alliance on Climate Change (of which The BMJ is a member). We hope you will sign our online declaration of intent to divest and give us your comments on our three criteria for potential divestment from other health harming industries.