Stop the teaching of pseudoscience - Comments by John Harvey
Posted: Fri Sep 02, 2011 12:29 pm
Dear Colleagues - Please see this superb discourse by John
From: UBM Medica Liferay Administrator [mailto:do-not-reply@ubmmedica.com]
Sent: 01 September 2011 17:25
To: Soroush Ebrahimi
Subject: -----SPAM----- New Blogs Comments by John Harvey
Dear finrod@finrod.co.uk,
John Harvey commented on your Blog.
----------
Les/Simon's seven points summed up his reasoning -- and, as it happens, the reasoning of many others -- in contending
(a) that homoeopathy cannot possibly achieve clinical results;
(b) that it does not achieve them;
(c) that seriously researching the question of its clinical effectiveness is not worth doing; and
(d) that such serious research has never been done.
Leaving aside that not all four of these contentions are mutually compatible, let's examine the evidence for each of them.
(a) That homoeopathy cannot possibly achieve clinical results. This contention is based on three commonly propagated misconceptions.
(1) The first misconception is a negative rather than a positive one. It is the systematic overlooking of what homoeopathy is. It is not treatment by ultramolecular doses (which I'll come to next). It is in fact treatment by the best available mimic of the patient's condition. In critical discussion of homoeopathy's success or otherwise, it is intellectually negligent to overlook what it is that we're discussing. The necessity of relevance is too obvious a point to labour. There is something else, though, significant to distinguishing homoeopathy from random prescription of ultramolecular medication, and it is this: that in no criticisms of homoeopathy has it been seriously suggested that this observation -- that a patient is peculiarly susceptible to a substance capable of inflicting the symptoms he or she already suffers -- is illusory. In other words, in attacking its credibility as capable of achieving clinical results, homoeopathy's critics consistently fail to attack the credibility of homoeopathy itself, instead choosing to attack the credibility of ultramolecular doses' capability of deranging health. There is in fact a long tradition of medical oddities and experimental results to suggest very strongly that a medicinal mimic of a patient's condition is a substance to which a patient is peculiarly susceptible. And it is this peculiar susceptibility to the substances most homoeopathic to a person's natural (i.e. not medicinal) state of health that offers the most readily available predictors of individual hypersensitivities. But that is another story.
(2) It is that susceptibility to a substance with the potential to aggravates one's entire extant collection of symptoms that points to the significance of the second misconception, which is that homoeopathy (necessarily) uses ultramolecular doses. It does not. The fact that traditional allopathic medicine cannot be successfully practised using such doses is testament to the medicinal "force" necessary to alter a state of health by use of a medicine whose pathogenesis (i.e. the medicinal illness it is capable of causing in the healthy) is fundamentally dissimilar from that state. Homoeopathy's common use of such doses -- and generally few of them -- is not a necessity as those ignorant of its definition (see above) commonly assume, but merely a handy possibility. That possibility is due to the similarity -- in a homoeopathically competent prescription -- between the illness the medicine is capable of causing and the condition of the patient. Homoeopathy may be practised entirely without use of ultramolecular doses, as it was in its first decade or so, and with no less success for it. The use of such doses is merely a refinement commonly adhered to because it makes very uncommon the patient reactions that otherwise commonly arise to a medicine whose primary effect is to cause the very symptoms from which the patient is already suffering.
(3) The third misconception common in offhand dismissal of homoeopathy is that all ultramolecular medicines are biologically equivalent to the solvent from which they have been eluted. So extremely obvious is this tenet that it has become an article of faith by those with an immovable belief that homoeopathy cannot have effects and with utter contempt for any profession of evidence for such effects. Yet the presumption, so obviously correct, has been demonstrated to be catastrophically wrong. A large body of materials science by O'Reilly, Rey, and the late Rustum Roy, amongst others, demonstrates that the properties of such a simple substance as water are not confined to its chemistry alone but reside also in its structure -- structure that a host of calibrated tests show to be susceptible to such events as recent dilution, elution, and succussion. (b) The contention that homoeopathy has been shown never to achieve clinical results. This contention, which I notice is never made by the more cautious (i.e. scientifically tentative) commentators, arises from a misunderstanding that failure of a single meta-analysis, no matter how poorly executed, or even of a single experiment, to demonstrate results constitutes disproof of all capacity for valid positive results to appear in another experiment.
Many high-quality studies have shown positive results. Not all of these have been based on the random-controlled trial (RCT); but then the misconception that the RCT is the only form of scientifically valid evidence is itself a howler, whose common repetition among those with this particular axe to grind merely overlooks the other forms of experimental study accepted, with good reason, as valuable in assessing the results of medical practice. (The use of consecutive cases is just one common example.)
Even staying within the bounds of RCTs, however, it is possible to discern that a number of studies strongly suggest that something is going on. It's not possible to prove absolutely, ever, that any medicine of any kind results in the effects it obviously does. But if reasonable rigour and statistical power combine to suggest that any medicine is effective, then they're sufficient basis too for assessing homoeopathy. There is one caveat that must be made with respect to analysis of any multicause–multieffect relationship to be examined through an RCT: that the RCT's design must, if it is to be at all suitable, take into account the particulars of the complexity of the relationship it is seeking to test. The application of such caution in research design to homoeopathy as well as to any other similarly complex relationship is a necessity, not a luxury, in doing good science. Those who beg off the work of appropriate design do themselves and their field no service even if they obfuscate their negligence by simplistically asserting that a design appropriate to assess single cause – single effect relationships must suffice for assessment of a multicause–multieffect one.
(c) The contention that clinical research on homoeopathy is not worth doing.
This contention always seems to be based in the first one: that homoeopathy cannot possibly achieve results. When one removes that premise, one is left with a contention that appears to be completely baseless.
(d) The contention that no serious research on clinical results of homoeopathic treatment has been done (except, possibly, a single meta-analysis or a single clinical trial claiming negative results).
For a bookful of such studies, see Bellavite and Signorini's Homeopathy: A Frontier in Medical Science. A more accessible summary of meta-analyses and reviews appears at the Faculty of Homeopathy's site, at , which concludes that 10 of 24 systematic reviews focusing on RCTs of homoeopathy "in specific clinical areas" yielded conclusions of positive effect, nine were inconclusive, and six found no evidence of effectiveness. For those interested, I'm happy to cite individual peer-reviewed papers.
One meta-analysis I will particularly mention is the 2005 Shang et al. meta-analysis (The Lancet), based on just eight clinical trials of homoeopathy that were selected from 21 high-quality trials. The meta-analysis purported to show that those trials demonstrated no effect. Subsequent analyses (e.g. Lüdtke & Rutten, "The conclusions on the effectiveness of homeopathy highly depend on the set of analysed trials", J. Clin. Epidem. 2008) of the authors' data showed that the arbitrariness of their selections had played an undue role in the analysis, and others (e.g. Rutten & Stolper, "The 2005 meta-analysis of homeopathy: the importance of post-publication data", Homeopathy 2008, http://www.sciencedirect.com/science/ar ... 1608000891) showed that the quality of all the homoeopathy trials was higher than that of those of traditional medicine and that Shang et al. had mismatched the homoeopathy trials to the traditional ones.
Another critique (Fisher, "Homeopathy and The Lancet", Evid Based Complement Alternat Med. 2006 March; 3(1): 145–147, www.ncbi.nlm.nih.gov/pmc/articles/PMC1375230) showed that the authors had predefined some outcomes for certain failure. Even the programme that had commissioned the study criticised it heavily for its methodological failures (http://www.bag.admin.ch/kv/forschung/f/ ... ht_PEK.pdf), and its methodology has come under heavy fire by many respected researchers, some of whom have shown that corrected meta-analysis of Shang et al.'s eight trials showed that their results had in fact supported an effect significantly beyond placebo (e.g. Fisher, Berman, et al., "Are the clinical effects of homeopathy placebo effects?", The Lancet 366(9503): 2082–83, http://www.thelancet.com/journals/lance ... 8/fulltext; Rutten & Stolper, "'Proof' against homeopathy in fact supports homeopathy", Homeopathy Jan 2006; 95(1) 57–61, http://www.homeopathy.org/research/edit ... Rutten.pdf).
----------
http://www.pulsetoday.co.uk/comment-blo ... udoscience
Sincerely,
UBM Medica Liferay Administrator
do-not-reply@ubmmedica.com
From: UBM Medica Liferay Administrator [mailto:do-not-reply@ubmmedica.com]
Sent: 01 September 2011 17:25
To: Soroush Ebrahimi
Subject: -----SPAM----- New Blogs Comments by John Harvey
Dear finrod@finrod.co.uk,
John Harvey commented on your Blog.
----------
Les/Simon's seven points summed up his reasoning -- and, as it happens, the reasoning of many others -- in contending
(a) that homoeopathy cannot possibly achieve clinical results;
(b) that it does not achieve them;
(c) that seriously researching the question of its clinical effectiveness is not worth doing; and
(d) that such serious research has never been done.
Leaving aside that not all four of these contentions are mutually compatible, let's examine the evidence for each of them.
(a) That homoeopathy cannot possibly achieve clinical results. This contention is based on three commonly propagated misconceptions.
(1) The first misconception is a negative rather than a positive one. It is the systematic overlooking of what homoeopathy is. It is not treatment by ultramolecular doses (which I'll come to next). It is in fact treatment by the best available mimic of the patient's condition. In critical discussion of homoeopathy's success or otherwise, it is intellectually negligent to overlook what it is that we're discussing. The necessity of relevance is too obvious a point to labour. There is something else, though, significant to distinguishing homoeopathy from random prescription of ultramolecular medication, and it is this: that in no criticisms of homoeopathy has it been seriously suggested that this observation -- that a patient is peculiarly susceptible to a substance capable of inflicting the symptoms he or she already suffers -- is illusory. In other words, in attacking its credibility as capable of achieving clinical results, homoeopathy's critics consistently fail to attack the credibility of homoeopathy itself, instead choosing to attack the credibility of ultramolecular doses' capability of deranging health. There is in fact a long tradition of medical oddities and experimental results to suggest very strongly that a medicinal mimic of a patient's condition is a substance to which a patient is peculiarly susceptible. And it is this peculiar susceptibility to the substances most homoeopathic to a person's natural (i.e. not medicinal) state of health that offers the most readily available predictors of individual hypersensitivities. But that is another story.
(2) It is that susceptibility to a substance with the potential to aggravates one's entire extant collection of symptoms that points to the significance of the second misconception, which is that homoeopathy (necessarily) uses ultramolecular doses. It does not. The fact that traditional allopathic medicine cannot be successfully practised using such doses is testament to the medicinal "force" necessary to alter a state of health by use of a medicine whose pathogenesis (i.e. the medicinal illness it is capable of causing in the healthy) is fundamentally dissimilar from that state. Homoeopathy's common use of such doses -- and generally few of them -- is not a necessity as those ignorant of its definition (see above) commonly assume, but merely a handy possibility. That possibility is due to the similarity -- in a homoeopathically competent prescription -- between the illness the medicine is capable of causing and the condition of the patient. Homoeopathy may be practised entirely without use of ultramolecular doses, as it was in its first decade or so, and with no less success for it. The use of such doses is merely a refinement commonly adhered to because it makes very uncommon the patient reactions that otherwise commonly arise to a medicine whose primary effect is to cause the very symptoms from which the patient is already suffering.
(3) The third misconception common in offhand dismissal of homoeopathy is that all ultramolecular medicines are biologically equivalent to the solvent from which they have been eluted. So extremely obvious is this tenet that it has become an article of faith by those with an immovable belief that homoeopathy cannot have effects and with utter contempt for any profession of evidence for such effects. Yet the presumption, so obviously correct, has been demonstrated to be catastrophically wrong. A large body of materials science by O'Reilly, Rey, and the late Rustum Roy, amongst others, demonstrates that the properties of such a simple substance as water are not confined to its chemistry alone but reside also in its structure -- structure that a host of calibrated tests show to be susceptible to such events as recent dilution, elution, and succussion. (b) The contention that homoeopathy has been shown never to achieve clinical results. This contention, which I notice is never made by the more cautious (i.e. scientifically tentative) commentators, arises from a misunderstanding that failure of a single meta-analysis, no matter how poorly executed, or even of a single experiment, to demonstrate results constitutes disproof of all capacity for valid positive results to appear in another experiment.
Many high-quality studies have shown positive results. Not all of these have been based on the random-controlled trial (RCT); but then the misconception that the RCT is the only form of scientifically valid evidence is itself a howler, whose common repetition among those with this particular axe to grind merely overlooks the other forms of experimental study accepted, with good reason, as valuable in assessing the results of medical practice. (The use of consecutive cases is just one common example.)
Even staying within the bounds of RCTs, however, it is possible to discern that a number of studies strongly suggest that something is going on. It's not possible to prove absolutely, ever, that any medicine of any kind results in the effects it obviously does. But if reasonable rigour and statistical power combine to suggest that any medicine is effective, then they're sufficient basis too for assessing homoeopathy. There is one caveat that must be made with respect to analysis of any multicause–multieffect relationship to be examined through an RCT: that the RCT's design must, if it is to be at all suitable, take into account the particulars of the complexity of the relationship it is seeking to test. The application of such caution in research design to homoeopathy as well as to any other similarly complex relationship is a necessity, not a luxury, in doing good science. Those who beg off the work of appropriate design do themselves and their field no service even if they obfuscate their negligence by simplistically asserting that a design appropriate to assess single cause – single effect relationships must suffice for assessment of a multicause–multieffect one.
(c) The contention that clinical research on homoeopathy is not worth doing.
This contention always seems to be based in the first one: that homoeopathy cannot possibly achieve results. When one removes that premise, one is left with a contention that appears to be completely baseless.
(d) The contention that no serious research on clinical results of homoeopathic treatment has been done (except, possibly, a single meta-analysis or a single clinical trial claiming negative results).
For a bookful of such studies, see Bellavite and Signorini's Homeopathy: A Frontier in Medical Science. A more accessible summary of meta-analyses and reviews appears at the Faculty of Homeopathy's site, at , which concludes that 10 of 24 systematic reviews focusing on RCTs of homoeopathy "in specific clinical areas" yielded conclusions of positive effect, nine were inconclusive, and six found no evidence of effectiveness. For those interested, I'm happy to cite individual peer-reviewed papers.
One meta-analysis I will particularly mention is the 2005 Shang et al. meta-analysis (The Lancet), based on just eight clinical trials of homoeopathy that were selected from 21 high-quality trials. The meta-analysis purported to show that those trials demonstrated no effect. Subsequent analyses (e.g. Lüdtke & Rutten, "The conclusions on the effectiveness of homeopathy highly depend on the set of analysed trials", J. Clin. Epidem. 2008) of the authors' data showed that the arbitrariness of their selections had played an undue role in the analysis, and others (e.g. Rutten & Stolper, "The 2005 meta-analysis of homeopathy: the importance of post-publication data", Homeopathy 2008, http://www.sciencedirect.com/science/ar ... 1608000891) showed that the quality of all the homoeopathy trials was higher than that of those of traditional medicine and that Shang et al. had mismatched the homoeopathy trials to the traditional ones.
Another critique (Fisher, "Homeopathy and The Lancet", Evid Based Complement Alternat Med. 2006 March; 3(1): 145–147, www.ncbi.nlm.nih.gov/pmc/articles/PMC1375230) showed that the authors had predefined some outcomes for certain failure. Even the programme that had commissioned the study criticised it heavily for its methodological failures (http://www.bag.admin.ch/kv/forschung/f/ ... ht_PEK.pdf), and its methodology has come under heavy fire by many respected researchers, some of whom have shown that corrected meta-analysis of Shang et al.'s eight trials showed that their results had in fact supported an effect significantly beyond placebo (e.g. Fisher, Berman, et al., "Are the clinical effects of homeopathy placebo effects?", The Lancet 366(9503): 2082–83, http://www.thelancet.com/journals/lance ... 8/fulltext; Rutten & Stolper, "'Proof' against homeopathy in fact supports homeopathy", Homeopathy Jan 2006; 95(1) 57–61, http://www.homeopathy.org/research/edit ... Rutten.pdf).
----------
http://www.pulsetoday.co.uk/comment-blo ... udoscience
Sincerely,
UBM Medica Liferay Administrator
do-not-reply@ubmmedica.com