GETSET - Graded Exercise Therapy guided Self-hElp Treatment (GETSET)
- Thread starter Bob
- Start date
- Messages
- 149
- Likes
- 203
First, you should read the study's abstract. The study was partly blinded, as described here:
Second, you should read about what an ethics committee does. The purpose of an ethics committee is to protect the patients in the study. The committee's job is to make sure the study is safe to participate in. As long a filling out a extra questionnaire is safe for the study participants, they should approve it. They are not a "quality control" committee. That's what peer review is for.
Primary outcomes were rated by participants, who were necessarily unmasked to assignment; the statistician was masked to treatment arm assignment for the analysis.
So the patients were not blinded, but the data analysis was. This is what would be expected for a trial testing a physical treatment that can't be blinded.Second, you should read about what an ethics committee does. The purpose of an ethics committee is to protect the patients in the study. The committee's job is to make sure the study is safe to participate in. As long a filling out a extra questionnaire is safe for the study participants, they should approve it. They are not a "quality control" committee. That's what peer review is for.
As you say, it was open label. With post-hoc methodology including primary outcomes.
I think user meant a steering committee. Peer review isn't a quality control process. It's an academic process.
I think user meant a steering committee. Peer review isn't a quality control process. It's an academic process.
The point is that those carrying out the study are aware of what is going on - the people doing treatments know so unless there is a very strong Chinese wall between those carrying out therapies and those running the trial they will have a good idea about its success. Also statisticians will have an idea of the significance of differences between two groups as well as the effect sizes even if they do not know which group has changed (there are only two) so they will know if there is not a significant gap although could get a shock about the direction. Think through who who knows what and when.
As I understand it ethics committees are responsible for the overall ethics. A long time ago I read through some of the EMA regulations that talked about the research ethics committees needing to review protocol changes. With PACE the MRC said they were happy with outcome switching because they were reviewed including by the MREC.
http://forums.phoenixrising.me/inde...e-from-the-mrc-to-hooper-pace-complaint.8867/
The point is they should be checking the ethics of the whole trial which includes all the conduct of the trial and how it is reported. Now maybe they were presented with exceptional reasons for changes to be made to the protocols after the end of the trial. But the PACE group failed to give good reasons and in some cases reasons they gave were simply wrong. This is some of the same group. The role is not about quality control but about having strong governance in place and they seem to be failing in this by not asking questions. If we want to wipe out arbitary outcome switching then we need ethics committees that will enforce rules and basic ethics of good reporting. If they are fail they should go and be replaced by people who will.
As I understand it ethics committees are responsible for the overall ethics. A long time ago I read through some of the EMA regulations that talked about the research ethics committees needing to review protocol changes. With PACE the MRC said they were happy with outcome switching because they were reviewed including by the MREC.
http://forums.phoenixrising.me/inde...e-from-the-mrc-to-hooper-pace-complaint.8867/
The point is they should be checking the ethics of the whole trial which includes all the conduct of the trial and how it is reported. Now maybe they were presented with exceptional reasons for changes to be made to the protocols after the end of the trial. But the PACE group failed to give good reasons and in some cases reasons they gave were simply wrong. This is some of the same group. The role is not about quality control but about having strong governance in place and they seem to be failing in this by not asking questions. If we want to wipe out arbitary outcome switching then we need ethics committees that will enforce rules and basic ethics of good reporting. If they are fail they should go and be replaced by people who will.
No I did mean the "London Bridge Research Ethics Committee" who approved the trial and should be approving or denying approval to major protocol changes. It is unethical to do outcome switching and it leads to misreporting of a trial. There can be technical reasons as to why it is necessary but they should be properly outlined. In an area where they have run many trials this should not be the case. As far as I can see the ethics committees are the basic governance structure for medical trials and yet they just seem to rubber stamp changes.
Why do these people not want us to get better...like back as functional productive members of society?
It seems as though in some sick and twisted way they want us to stay sick and live within the torture that is ME.
Some days I am ashamed to be be part of the same human race as people like this....
If I could I would.......I would not choose this life nor would I choose it for anyone else....
It seems as though in some sick and twisted way they want us to stay sick and live within the torture that is ME.
Some days I am ashamed to be be part of the same human race as people like this....
If I could I would.......I would not choose this life nor would I choose it for anyone else....
- Messages
- 149
- Likes
- 203
At the very worse they added a new primary outcome. "Outcome switching" would be either removing an old outcome or not reporting on it, and reporting on a new outcome in it's place. That would be the switch. And they clearly did not do that. The worst thing that can be said is that they added a new outcome and reported on it.
Most scientists would call what they did "reporting more data than they planned to report".
It doesn't work that way in the US, and I doubt it does in Europe or the UK, either. Their are two limitations on changing a study: one is patient safety (as enforced by the ethics committee and the US FDA) and the other is fear that it will fail peer-review or that it will not be published. That's it. Once you have your data, how you analyse it is not a safety issue, so is covered by the "fear of peer review failure or fear of publication rejection". In the US, there is no pre-approval of any of that. Not by an ethics committee or any one else.
Most scientists would call what they did "reporting more data than they planned to report".
It doesn't work that way in the US, and I doubt it does in Europe or the UK, either. Their are two limitations on changing a study: one is patient safety (as enforced by the ethics committee and the US FDA) and the other is fear that it will fail peer-review or that it will not be published. That's it. Once you have your data, how you analyse it is not a safety issue, so is covered by the "fear of peer review failure or fear of publication rejection". In the US, there is no pre-approval of any of that. Not by an ethics committee or any one else.
- Messages
- 149
- Likes
- 203
That would make even less sense, since most studies don't have steering committees at all. Usually there is just a PI (the principal or primary investigator) and he or she just makes decisions, often by consulting with the lead statistician or the other researchers involved, but not with a formal committee of any kind. It is only the largest, and most complex studies that have a formal steering committee. I've covered dozens of studies (not in ME/CFS), and only a handful had steering committees, and all of those that did had more than 1000 patients involved. PACE certainly had a steering committee, but that was unusual (especially for a 400+ person study). I supposed it's possible that it's a UK thing, but I've covered some trials in the UK, and they did not have steering committees, either.
Lack of blinding generates misleading results. The rest of medicine has long accepted this.
It would be possible to minimize this problem with a study design that doesn't rely entirely on subjectively rated health.
Cardiopulmonary exercise testing could be used as measure of the actual fitness capacity. With this test it's possible to tell how much effort a person is making.
Actigraphy would probably generate more reliable data than self rated health as well.
Of course we can already guess why White et. al. have abandoned objective measures in GETSET. Because they show no or marginal improvement, as seen in PACE or other studies.
It would be possible to minimize this problem with a study design that doesn't rely entirely on subjectively rated health.
Cardiopulmonary exercise testing could be used as measure of the actual fitness capacity. With this test it's possible to tell how much effort a person is making.
Actigraphy would probably generate more reliable data than self rated health as well.
Of course we can already guess why White et. al. have abandoned objective measures in GETSET. Because they show no or marginal improvement, as seen in PACE or other studies.
I take your point that they have not removed the original primary outcome. But this is not about reporting more data than planned it is about how they describe the success of the trial. They choose to only use the SF36 scale as a primary outcome and then added an additional primary outcome and they will be talking about the overall success of the trial in terms of meeting this new set of primary outcomes. It changes the way the trial is reported and results are seen.
I've never liked the idea of pre-guessing the primary outcome because in more mature trials it allows those who pick the outcomes to use their experience to pick what the ones they consider will give the best results. They should give detailed reasons as to their choices but I've rarely seen this. I would prefer for a range of different end-points to be collected and for success to be declared as they are all changing or where their is good reason to understand why certain measures are not changing. But they preferred methodology for clinical trials is to design the protocol with primary end points at the start and use these to judge the trial. To then change the set of primary outcomes after the end of the trial just says 'dodgy' especially when it is an open label trial. I have not seen the full paper (I don't think it is published yet) so they may give some very good reasons for this but in the past they have not so I am judging them on past behaviour.
So after doing a quick search for changing primary end points I found a European Medical Agency document that talks about getting approval for trials
http://ec.europa.eu/health/files/pharmacos/docs/doc2005/10_05/ca_14-2005_en.pdf
To quote a paragraph
The point they make is that the protocols are there for multiple things including the scientific value as well as safety.
I'm somewhat alarmed that "fear of peer review failure or fear of publication rejection" could be used as a reason for changing primary end-points in a trial. This places a primary motivation of getting published over providing good information. Of course if there is a deviation from the predefined primary end points for a trial without being very clear and with very good reasons then peer reviewers should reject the paper. However, most will never read the protocol and most reviews don't spend that much time reviewing papers.
I've never liked the idea of pre-guessing the primary outcome because in more mature trials it allows those who pick the outcomes to use their experience to pick what the ones they consider will give the best results. They should give detailed reasons as to their choices but I've rarely seen this. I would prefer for a range of different end-points to be collected and for success to be declared as they are all changing or where their is good reason to understand why certain measures are not changing. But they preferred methodology for clinical trials is to design the protocol with primary end points at the start and use these to judge the trial. To then change the set of primary outcomes after the end of the trial just says 'dodgy' especially when it is an open label trial. I have not seen the full paper (I don't think it is published yet) so they may give some very good reasons for this but in the past they have not so I am judging them on past behaviour.
So after doing a quick search for changing primary end points I found a European Medical Agency document that talks about getting approval for trials
http://ec.europa.eu/health/files/pharmacos/docs/doc2005/10_05/ca_14-2005_en.pdf
To quote a paragraph
The point they make is that the protocols are there for multiple things including the scientific value as well as safety.
I'm somewhat alarmed that "fear of peer review failure or fear of publication rejection" could be used as a reason for changing primary end-points in a trial. This places a primary motivation of getting published over providing good information. Of course if there is a deviation from the predefined primary end points for a trial without being very clear and with very good reasons then peer reviewers should reject the paper. However, most will never read the protocol and most reviews don't spend that much time reviewing papers.
That's a helpful blog...
This is the research paper associated with it, which could be a handy reference to keep a note of...
Ernst E, Lee MS. A trial design that generates only "positive" results. J Postgrad Med. 2008;54:214-6.
http://www.jpgmonline.com/text.asp?2008/54/3/214/41806
http://www.ncbi.nlm.nih.gov/pubmed/18626172
It also mentions objective outcome measures...
More discussion...
This is the research paper associated with it, which could be a handy reference to keep a note of...
Ernst E, Lee MS. A trial design that generates only "positive" results. J Postgrad Med. 2008;54:214-6.
http://www.jpgmonline.com/text.asp?2008/54/3/214/41806
http://www.ncbi.nlm.nih.gov/pubmed/18626172
It also mentions objective outcome measures...
More discussion...
Last edited:
For anyone like me, who hadn't worked this out, it seems that the full GETSET paper hasn't been published yet, but just the abstract as part of a list of conference abstracts: The 4th annual scientific conference of the European Association for Psychosomatic Medicine (EAPM) in Luleå, Sweden, 16-18 June 2016. The conference is titled: "Transforming health through evidence and empathy." (If you believe it!) http://eapm2016.com/
Last edited:
- Messages
- 2,051
- Likes
- 10,097
"Transforming health through irony and bullshit"?
Not a therapy recommendation.
For anyone interested, the chronic fatigue Service at the NHS East London Foundation Trust (ELFT), based at St Leonard’s Hospital in Hackney, has a "GETSET" (i.e. self-guided GET) leaflet from Barts available for patients on its website:
https://www.elft.nhs.uk/uploads/files/1/Services/Chronic Fatique/GETSET Leaflet.pdf
It is to be found at the bottom of this webpage:
https://www.elft.nhs.uk/service/308/Chronic-Fatique-Service/
Maybe they're getting ready to roll out GETSET as a proven treatment strategy, but the leaflet acknowledges (page 2) that it's not yet a proven treatment.
Note that it's an NHS leaflet, and the ELFT is an NHS service.
For anyone interested, the chronic fatigue Service at the NHS East London Foundation Trust (ELFT), based at St Leonard’s Hospital in Hackney, has a "GETSET" (i.e. self-guided GET) leaflet from Barts available for patients on its website:
https://www.elft.nhs.uk/uploads/files/1/Services/Chronic Fatique/GETSET Leaflet.pdf
It is to be found at the bottom of this webpage:
https://www.elft.nhs.uk/service/308/Chronic-Fatique-Service/
Maybe they're getting ready to roll out GETSET as a proven treatment strategy, but the leaflet acknowledges (page 2) that it's not yet a proven treatment.
Note that it's an NHS leaflet, and the ELFT is an NHS service.
Last edited:
