A great deal of noise has been made over various things the authors of the PACE study have claimed in publications or spoken statements. What has received much less attention is the remarkable extent to which they have remained silent on very significant aspects of that trial. The continuing stream of papers based on the flawed primary study prompts me to explore this neglected aspect.
Did anyone notice the extent to which those PACE results required substantial selection of patients other doctors considered "CFS patients"? No matter what the authors think about diagnosis of "CFS", what other doctors think will continue to determine which patients are shunted into this treatment option. This makes the question of how you separate the patients for which the suggested treatment is useful, from those for whom it is not, very pertinent, particularly when considering costs. Most of the presently-available information can be found in Figure 1 of that paper.
They chose only about 898 out of an initial intake of 3158 patients recommended for possible inclusion. Of these only 641 agreed to participate, and one withdrew consent after the trial started. I'm making a reasonable assumption that patients who refused to participate in the trial would be especially unlikely to benefit. Even with that degree of selectivity, the authors themselves appear to say, in some public statements, that some 40% of patients did not respond to their favored treatments. (It is hard for anyone else to classify responders and non-responders, based on released data.) This might mean that the PACE results only apply to 384 patients out of 3158, assuming it means anything. What do they recommend for the rest of those patients?
The authors were not terribly forthcoming on the subject of exactly what physiological conditions should be tested for as exclusionary. The cost of this testing is nowhere apparent in this sequence of publications, including an economic analysis.
One can only assume this testing and these costs are shifted to regular NHS doctors and service. (Externalizing costs is one popular way to improve your bottom line, ask any accountant.) The same publication assumed the benefits of PACE had been shown to be substantial and lasting. I've already expressed my opinion on the negligible clinical significance of the reported results, but stopped short of considering lasting effects, for which no data at all are available. The idea that significant positive effects are lasting is based on nothing more than wishful thinking, which always helps economists touting nostrums for economic ills. A naive reader of that economic analysis would never guess this without digging much deeper than most professionals are likely to go.
This reliance on regular NHS doctors and services, even extending to the suggestion that CFS patients can be treated by "less senior doctors" (less competent ones?) to reduce costs by half, is an apparent contradiction to the suggestion by some of the same authors in a paper on misdiagnosis which finds rates of diagnostic error around 30%, (though the authors are convinced that missed psychiatric diagnoses are twice as common as making them in error.)
If we are to start comparing rates of diagnostic error between competing teams of psychiatrists we can have a field day, even when we consider better-established diagnostic categories like schizophrenia. The implicit assumption that the PACE authors have an error rate of 0%, without ordering any physiological tests, when even pathologists examining actual tissues removed from patients have an error rate of 1% is ludicrous. What are they doing, reading patient auras?
A publication by Byron Hyde is more forthcoming on the subject of undiagnosed physical conditions found in patients sent to him as CFS patients, and indicates he had no trouble finding undiagnosed and untreated disease, even including syphilis. Note that Dr. Hyde did not require a team of collaborators and millions of pounds to study an unbiased sample of patients referred to him as CFS patients.
One patient in an unfiltered sequence of 53 was even found to be missing a substantial chunk of brain nobody had previously noticed. (How he was able to fit "false illness beliefs" into the remaining volume will no doubt be explained in later publications by the PACE authors.)
Dr. Hyde did find 16% of patients with psychiatric disorders, though not necessarily the one called CFS. (My numbers above indicate that he may actually have sent a higher percentage of patients to psychiatrists than the PACE group.) This does not rule out secondary physical problems, which are virtually ignored in PACE. Must be somebody else's problem.
If you examine a group of diagnosed psychiatric patients without finding undiagnosed and untreated physical illness in 9% or so, you are not trying very hard to find it.
By the way, while we're on the topic of omissions, has anyone ever found a source for the 70% capacity rule used in the APT arm of PACE, other than the PACE authors themselves? That should also count as a very curious omission since pacing was allegedly at the center of this trial. In the absence of such a source we must assume APT was created purely to serve in a straw-man argument.
A final omission is the fact that at the time of the study one specialist in statistics listed as an author was a PhD. candidate with a dissertation which depended on PACE. One might compare the composition of the committee which recommended awarding that degree with the list of authors. This is not a person in a position to make judgements independent of senior authors. Was independent review of statistical practices any more significant than checking arithmetic applied to data declared valid by those senior authors?
Did anyone notice the extent to which those PACE results required substantial selection of patients other doctors considered "CFS patients"? No matter what the authors think about diagnosis of "CFS", what other doctors think will continue to determine which patients are shunted into this treatment option. This makes the question of how you separate the patients for which the suggested treatment is useful, from those for whom it is not, very pertinent, particularly when considering costs. Most of the presently-available information can be found in Figure 1 of that paper.
They chose only about 898 out of an initial intake of 3158 patients recommended for possible inclusion. Of these only 641 agreed to participate, and one withdrew consent after the trial started. I'm making a reasonable assumption that patients who refused to participate in the trial would be especially unlikely to benefit. Even with that degree of selectivity, the authors themselves appear to say, in some public statements, that some 40% of patients did not respond to their favored treatments. (It is hard for anyone else to classify responders and non-responders, based on released data.) This might mean that the PACE results only apply to 384 patients out of 3158, assuming it means anything. What do they recommend for the rest of those patients?
The authors were not terribly forthcoming on the subject of exactly what physiological conditions should be tested for as exclusionary. The cost of this testing is nowhere apparent in this sequence of publications, including an economic analysis.
One can only assume this testing and these costs are shifted to regular NHS doctors and service. (Externalizing costs is one popular way to improve your bottom line, ask any accountant.) The same publication assumed the benefits of PACE had been shown to be substantial and lasting. I've already expressed my opinion on the negligible clinical significance of the reported results, but stopped short of considering lasting effects, for which no data at all are available. The idea that significant positive effects are lasting is based on nothing more than wishful thinking, which always helps economists touting nostrums for economic ills. A naive reader of that economic analysis would never guess this without digging much deeper than most professionals are likely to go.
This reliance on regular NHS doctors and services, even extending to the suggestion that CFS patients can be treated by "less senior doctors" (less competent ones?) to reduce costs by half, is an apparent contradiction to the suggestion by some of the same authors in a paper on misdiagnosis which finds rates of diagnostic error around 30%, (though the authors are convinced that missed psychiatric diagnoses are twice as common as making them in error.)
If we are to start comparing rates of diagnostic error between competing teams of psychiatrists we can have a field day, even when we consider better-established diagnostic categories like schizophrenia. The implicit assumption that the PACE authors have an error rate of 0%, without ordering any physiological tests, when even pathologists examining actual tissues removed from patients have an error rate of 1% is ludicrous. What are they doing, reading patient auras?
A publication by Byron Hyde is more forthcoming on the subject of undiagnosed physical conditions found in patients sent to him as CFS patients, and indicates he had no trouble finding undiagnosed and untreated disease, even including syphilis. Note that Dr. Hyde did not require a team of collaborators and millions of pounds to study an unbiased sample of patients referred to him as CFS patients.
One patient in an unfiltered sequence of 53 was even found to be missing a substantial chunk of brain nobody had previously noticed. (How he was able to fit "false illness beliefs" into the remaining volume will no doubt be explained in later publications by the PACE authors.)
Dr. Hyde did find 16% of patients with psychiatric disorders, though not necessarily the one called CFS. (My numbers above indicate that he may actually have sent a higher percentage of patients to psychiatrists than the PACE group.) This does not rule out secondary physical problems, which are virtually ignored in PACE. Must be somebody else's problem.
If you examine a group of diagnosed psychiatric patients without finding undiagnosed and untreated physical illness in 9% or so, you are not trying very hard to find it.
By the way, while we're on the topic of omissions, has anyone ever found a source for the 70% capacity rule used in the APT arm of PACE, other than the PACE authors themselves? That should also count as a very curious omission since pacing was allegedly at the center of this trial. In the absence of such a source we must assume APT was created purely to serve in a straw-man argument.
A final omission is the fact that at the time of the study one specialist in statistics listed as an author was a PhD. candidate with a dissertation which depended on PACE. One might compare the composition of the committee which recommended awarding that degree with the list of authors. This is not a person in a position to make judgements independent of senior authors. Was independent review of statistical practices any more significant than checking arithmetic applied to data declared valid by those senior authors?
