Of course there are some cases when blinding would be the wrong thing to do: for example, if one wishes to compare two inhalers for ease of use. This can and should be done in an open fashion. Eksewhere a technical problem is that one cannot give unblinded studies a standard lower weight because how much weight they should be given depends on how much higher quality information is available. This is very esay to show using mean square error.
Some discussion of issues affecting blinding is given in the references below.
Stephen
Senn, S. J. (1994). "Fisher's game with the devil." Statistics in Medicine 13(3): 217-230.
The publication of Fisher's correspondence on statistics has shed new light on his views on randomization. Quotations from this correspondence and from other works of Fisher are used to illustrate the role of randomization in clinical trials. It is concluded that Fisher's views not only are coherent but, despite having been developed over 60 years ago and with particular reference to agricultural experiments, are still relevant to the planning and analysis of clinical trials today.
Senn, S. J. (1995). "A personal view of some controversies in allocating treatment to patients in clinical trials [see comments]." Statistics in Medicine 14(24): 2661-2674.
A non-exhaustive but nevertheless wide-ranging if biased and personal view of various matters affecting the allocation of treatments to patients in controlled clinical trials will be undertaken from the standpoint of a statistician working in drug development. The following topics are considered: the use of placebos' in the run-in phase; the use of crossover trials; selection of patients; various matters concerning centres; two extreme alternatives to randomization, and various matters linking blinding and allocation.
Senn, S. J. (2004). "Turning a blind eye: Authors have blinkered view of blinding." BMJ 328(7448): 1135-b-1136.
Senn, S. J. (2007). Statistical Issues in Drug Development. Hoboken, Wiley.
Stephen Senn
Professor of Statistics
School of Mathematics and Statistics
Direct line: +44 (0)141 330 5141
Fax: +44 (0)141 330 4814
Private Webpage: http://www.senns.demon.co.uk/home.html
University of Glasgow
15 University Gardens
Glasgow G12 8QW
The University of Glasgow, charity number SC004401
________________________________________
From: Evidence based health (EBH) [[log in to unmask]] On Behalf Of Michael Brown [[log in to unmask]]
Sent: 10 February 2011 22:19
To: [log in to unmask]
Subject: Re: How much should study results be downgraded if blinding/masking is incomplete?
Mark: You may find the recently published GRADE series useful: http://www.gradeworkinggroup.org/publications/index.htm
In this series, the authors describe an explicit approach to rating the quality of the evidence for each outcome. As per Paul's response, the evidence from RCTs with inadequate masking may be rated down for risk of bias for a subjective outcome, but would not be downgraded if the outcome was objective (eg, mortality). The risk of bias assessment would be reported under Limitations in the evidence profile.
Michael Brown MD, MSc
Director of the Division of Emergency Medicine
Michigan State University College of Human Medicine
15 Michigan Street NE, Grand Rapids, MI 49503
PH 616.234.2732
CELL 616.490.0920
[log in to unmask]<mailto:[log in to unmask]>
-----Original Message-----
From: Evidence based health (EBH) [mailto:[log in to unmask]] On Behalf Of Paul Glasziou
Sent: Thursday, February 10, 2011 4:51 PM
To: [log in to unmask]
Subject: Re: How much should study results be downgraded if blinding/masking is incomplete?
Dear Mark
A good question. The blinding/masking is critical for only *subjective* outcomes, and appears of similar importance to adequate randomization.
So we teach students the3-item "RAMbo" appraisal method:
Randomized?
Attrition (>20% or differential)?
Measurements - blinded OR objective
(the OR is crucial and seems to need repeating so sink in).
The best evidence I know for this is from the 2008 BMJ study (below) which suggested:
"In trials with subjective outcomes effect estimates were exaggerated when there was inadequate or unclear allocation concealment (ratio of odds ratios 0.69 (95% CI 0.59 to 0.82)) or lack of blinding (0.75 (0.61 to 0.93)). In contrast, there was little evidence of bias in trials with objective outcomes: ratios of odds ratios 0.91 (0.80 to 1.03) for inadequate or unclear allocation concealment and 1.01 (0.92 to 1.10) for lack of blinding."
Best wishes
Paul GLasziou
Wood L, Egger M, Gluud LL, Schulz KF, Jüni P, Altman DG, Gluud C, Martin RM, Wood AJ, Sterne JA. Empirical evidence of bias in treatment effect estimates in controlled trials with different interventions and outcomes: meta-epidemiological study. BMJ. 2008 Mar 15;336(7644):601-5. Epub 2008 Mar 3.
<http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2267990/?tool=pubmed>
________________________________________
From: Evidence based health (EBH) [[log in to unmask]] On Behalf Of Mark V. Johnston [[log in to unmask]]
Sent: 10 February 2011 21:12
To: [log in to unmask]
Subject: How much should study results be downgraded if blinding/masking is incomplete?
We have been debating this question in the Clinical Practice Committee of the American Congress of Rehabilitation.
In past reviews, we had to downgrade studies because of absence of masking or blinding at least as often as for failure to randomize.
Does anyone know of resources on the topic? Do any of the resources provide data to support their opinions?
Explanation:
In the past, we have employed the evidence grading scheme of the American Academy of Neurology (AAN) or the PEDRO system.
- The AAN system downgrades a study if the outcome is not blinded or masked. An unblinded study cannot provide greater than Class 4 or suggestive (level U) evidence, unless the outcome is entirely objective (e.g. mortality, a serum level done by a professional clinical laboratory which follows uniform procedures across patients).
- The PEDRO system is a checklist of study characteristics. It asks whether outcome assessors, clinicians, or patients are blinded, and also total allocation concealment. There are then 4 points for blinding. By comparison, just one point is given for randomization or not.
My opinion: Masking is not 4 times as dangerous as failure to randomize. Blinding is relatively easy in drug studies. (Who knows what's in the pill?) But it is difficult and sometimes impossible to blind therapists and patients so that they do not know what therapeutic activities they are engaged in. Nonetheless, some behavioral or activity therapies are more effective than others, and we need to know which.
We have read many opinions, but we are still looking for resources and data to support our decisions. Hope some of your are interested.
--
Mark V. Johnston, Ph.D.
Professor, Occupational Science and Technology,
College of Health Sciences
University of Wisconsin – Milwaukee
(414) 229-3616
|