This paper will soon appear in Journal of Thyroid Research, It's extremely important as it asks the question as to why basic physiological discoveries about how the thyroid and body interact are not mirrored equivalently by clinical trial results, that do not consistently support combo therapy over mono therapy with T4. Put simply, ALL randomized clinical trials attempting to answer this question are faultily based, and by their basic error in statistical analysis and inadequate power of numbers of patients, cannot address the questions they are asking. We go into great detail as to how this should be so and suggest remedies for better trial design in the future. The irony is that NICE are at present drawing up guidelines based on the very completely flawed "evidence" from trials. Obviously it follows their conclusions are fatally flawed and no weight can be given to the outcome if it is based on such work. It should not take too long for this paper in J Thyroid Research (an open journal) to be published and when this happens, it should be addressed to the NICE deliberators to give them pause for thought. The title is:
"Lessons from Randomised Clinical Trials for Triiodothyronine Treatment of Hypothyroidism: Have They Achieved Their Objectives?"
It will quite a blow for the thyroid establishment to find they have wasted their time and come to wrong conclusions as a result.
Written by
diogenes
Remembering
To view profiles and participate in discussions please or .
This time they will have to copy Trumpism, saying "t'aint so" with their brains, eyes and ears tight shut. But the killer is that no longer can they confidently state "what is so" with an conviction.
There will never be a "mea culpa" moment by the medical profession. That isn't how it works. When the evidence is overwhelming, they quietly change track without reference to the bad ideas held before. It's like a pinprick in a balloon. At first very gradually, the balloon imperceptibly loses its size. Then the hole gets bigger, the balloon shrinks faster and at last bang, its gone. No wrong idea is for ever - it just takes time for it to die.
I’m looking forward to reading this paper.. it can’t be published soon enough.
NICE have handed the Job of compiling the Thyroid Guideline to the National Guideline Centre, who in turn asked (?) the Royal College of Physicians. I’m not sure how all this is interconnected or who asked who. You would have to make further enquiries to make sure this information is correct.
How fabulous! Is this paper going to be given to Lord Hunt at the meeting? Would be a great ace up the sleeve to produce when the BTA and BTF representatives may be being a little unhelpful.
A lot of papers I've read on thyroid disease are woefully inadequate. As you say, too small sample size, etc., oh and statistical anomalies. Many raise too many questions than give answers. A lot of their "research" appears flawed by too many variables being present in their analyses to be able to get any sort of conclusion. Unless one seeks a "conclusion" with bias. Very amateur work compared to other fields of science and engineering. Looking forward to reading your paper. If I hadn't suffered Hashimoto's disease I would never have found out just how bad some "academics" are in the thyroid world!!
Possibly because a lot of medical research is done by medics, who are practitioners rather than scientists, so they have no research training. I'm a social scientist, and years ago I used to give a single lecture on statistics to doctors planning their first clinical trial, and that was all they got. The same kind of material our own students had as first years.
I worked in education for years and there's the same problem, the field is full of teachers rather than researchers, so research practice isn't valued.
Yeah I think you've hit the nail on the head there. I can imagine the maths and statistics taught is rather meek and mild, and not too in depth. Assuming they stayed for the full half hour! Joking aside, medical students/practitioners have a lot of other things to cover. Shame really though, as a lot of medical research papers tend to exhibit some form of statistical analysis and many are lacking in quality. So are you saying your own social science students were much more equipped to analyse data using statistical methods than the medical students?
Yes, I'd say the social science students were more equipped, they'd have studied statistical methods for 10-15% of their time all the way through, including a lot on research design and questions. I think one of the biggest issues with tacked on statistics is not having an understanding of the kind of question that can be addressed.
And to clarify, I am very very far away from a statistical methods specialist, in fact I am primarily a social theorist, which is more or less philosophy! This was the absolute basics, and very new ideas for most of them
From the paper I deposited with Louise Roberts the other day, a little excerpt:
That is, even in the best-case scenario, we
should not think of a correlation in group data as an estimate that generalizes to any given individual in the population. Stated
bluntly, this implies that the temptation to use aggregate estimates
to draw inferences at the basic unit of social and psychological
organization—the person—is far less accurate or valid
than it may appear in the literature. Indeed, even the best-case scenario
is quite alarming: Only 68% of all individual correlational values
fall within a range that would be predicted by group data to cover
99.7% of all possible correlations—a discrepancy of nearly 32%.
The worst-case scenario is clearly dire: It is plausible that inattention
to nonergodicity and a lack of group-to-individual generalizability threaten the veracity of countless studies, conclusions,
and best-practice recommendations.
A strong statement indeed!
NB Ergodicity is a term that states that variability within an individual is equivalent to that within a group. Noergodicity is when this doesn't happen, as in thyroid clinical trials.
I see, so the social science students cover statistics a lot more. Does a lot of the data to be analysed come from questionnaires? Quite often with tacked on modules and courses, the surface of knowledge is barely scratched, resulting in a low depth of knowledge that isn't really worth anything in the real world. Whether that real world is industry or academia.
Blimey social theorist! Sounds impressive. Most of my statistical analysis relates to experimental and computational data. So fairly mathematical. Physical data usually measured and computational data derived from mathematical modelling of thermodynamics and fluid mechanics systems and processes. Much greater fun now my brain fog has lifted!
I smell a scientist! It's true: there is an enormous gap between the essential rigour of legitimate biochemistry/physiology findings and the sloppy realm of medical-based analysis. No meeting of minds, but one we are trying to bridge.
It is great that we are lucky enough to have researchers who can come to the conclusion you and your team have as so many people aren't able to argue their case of why they feel awful whilst being told you're 'normal' and given a prescription for the symptom.
Content on HealthUnlocked does not replace the relationship between you and doctors or other healthcare professionals nor the advice you receive from them.
Never delay seeking advice or dialling emergency services because of something that you have read on HealthUnlocked.