the issue of patient preferences is a vexing one and thus some suggest the patient trial trial...those who seek a particular treatment get their treatment, and the rest are randomized as usual e.g. simple randomisation etc..I find a complex approach but it is an option now but the experts can weigh in. thus it may be in such a preference trial, you will get 4 possible groups...those randomized to treatment 1, prefer treatment 1; randomized to treatment 2, and those who prefer treatment 2....no doubt raises lots of questions as to how best to analyze as this is not now a classic RCT...you cannot compare a randomzied group to a non randomized group as the groups are different just be the fact that the randomzied arm has spread out the known and unknown
prognostic factors near evenly across groups and the preference arms has not...quite biased with confounders. one option may be to analyze the randomzied arm via traditional stats such as RR etc and the other preference arms as cohort observational designs which they are, via analysis that adjusts statistically for the potential alternative explanations (confounders)..some suggest a disadvantage of the preference portion of your study is likely to escalate costs...some say conduct it whereby you get patients to list their preferences and take it into explanatory account in interpreting the results (factoring in subject preferences) yet randomize all fully as a full RCT...
2.) the informed
consent issue is complicated by the fact that you have to ensure the patient is fully informed just as if in person...maybe you can ensure that even if they consent with a click button, it is prefaced on the fact that they then must see you or someone in person or via phone so as to allow for the full trial issues to be clarified, questions they may have , confidentiality issues and concerns, their rights of withdrawal, and how you will manage their electronic information and digital signature. also, this helps you ensure that the person enrolling electronically is the real deal and not a fraud...can ensure baseline data is accurate etc...
Best, Paul E. Alexander "Batallamos hasta la victoria, siempre"
--- On Sat, 3/16/13, Amy Price <[log in to unmask]> wrote:
From: Amy Price <[log in to unmask]>
Subject: Comprehensive Cohort Design/patient preference trial as alternative to the randomized controlled trial and online informed consent
To: [log in to unmask]
Received: Saturday, March 16, 2013, 9:41 PM
Dear all,
1) We are using Public led trials as a form of interactive patient education citizen science so these will be interventions that will be safe for the public to do for themselves. We are considering using a Comprehensive Cohort Design/patient preference trial as alternative to the randomized controlled trial.
The advantages are possible better compliance, (a concern because attrition in an online platform is about 65-85%) choose randomization offers more opportunities for teaching, less concern about cross over as the public has chosen the randomization and the ones that have no preference will be randomized so it is essentially a 4 arm trial. Some of the papers say there is less opportunity for bias, others say there is more. To resolve this we can analyze the groups individually as well as show the total results. One difficulty a paper mentioned was that should the results be accepted for publication and be biased this would hurt science and medicine. I think this is overstated because we are using PROMS it would be more hypothesis generating for a lab controlled study than the final word so to
speak. If you have any thoughts, see more advantages/disadvantages or have come across great papers on this please share, I am looking at this theoretically and am open to how this turns out in practice or what you see with the eyes of experience.
2) My second problem to solve is that of online informed consent. Most people will just click the agree button for an online consent, me included I could have sold my organs and given my husband and children into slavery and I would never know. I would define the button as 'consent' but not informed and ethical practice requires both. They will be coming in too fast to hold only Webinairs with Q and A although we will make this available and use
other medium. Does anyone have a good idea for solving this. We are considering an interactive process, unfortunately I can't find good literature on this except for a couple of patents and phase 2 FDA trials on this. I also need to know how far people read down before clicking I agree.
I will be certain to give full credit for any help. This group is such a wonderful resource and often I learn much from answers arguments and ideas communicated within the group. Thanks for being here….
Best
regards, Amy Price
|
