Fibromyalgia appears destined to remain the elephant in the room if this month’s Rheumatology edition of Australian Family Physician is anything to go by – not a mention of it. So what do healthcare specialists offer? In 2006 Drs Littlejohn and Guymer published, through their work as Monash Director of Rheumatology and trainee respectively “Fibromyalgia Syndrome: Which Antidepressant Drug Should We Choose”. Linking their Medical Centre with the University seems a worthy approach so as to redress physician misconceptions – indeed, shortly thereafter the Journal of Clinical Rheumatology published a survey of Southeast Asian rheumatologists. 87% of them believed fibromyalgia incorporated aspects of psychological illness and only 40% of those associated with an institution reported inclusion of FM in their undergraduate training.
Dr Littlejohn then contributed to the 2009 yearlong study on effectiveness of duloxetine/Cymbalta led by Eli-Lilly employee Amy Chappell, which was discredited by exclusion on quality grounds from the independent Cochrane musculoskeletal group’s systematic review of studies on SNRIs . Their supposedly * impartial conclusion drawing upon another five, unbiased assessments was that 10% more people reported significantly reduced pain with duloxetine than those duped by a placebo, however another 9% discontinued treatment due to side-effects. Withdrawal in itself is harmful, the US FDA has published a safety advisory for Cymbalta Discontinuance Syndrome. Agreements posted in 32,000 blogs are somewhat alarming! Any benefits obtained are ‘figure-atively’ outweighed by weightgain reducing propensity to exercise, consensus being that moderate exercise is beneficial for fibro.
2009 was a bad year for pharma. Eli-Lilly pleaded guilty to illegal marketing of anti-psychotic drug Zalprexa for off-label use, and was fined $1.4bn. A recent out-of-court settlement was made with the family of a boy who suicided whilst on Cymbalta, whose claim being that another suicide during drug trials should have prompted warnings. And Pfizer paid a record $2.3bn for fraudulent marketing of painkillers, including Lyrica. Nonetheless these two medications remain highly profitable with an estimated $18bn in sales in 2012…. which trivialises the settlement for $43m with US attorneys generals that year for once again marketing Lyrica other than for an approved purpose. With such monetary power, when federal NHMRC funding was approved for only 17% of applications for 2014 is it any wonder that Littlejohn and Guymer are reliant on consultancy fees paid by Eli-Lilly and Pfizer declared (inconsistently*) under their conflicts of interest? “The medical profession—in the US, Europe, and beyond—remains heavily reliant on industry funded continual medical education, and many doctors have accepted substantial hospitality and consultancy fees. Very few have been prosecuted. Disclosure remains patchy and inconsistent. Yet it is their decisions that ultimately determine if medicines are reaching patients for whom they are not suitable. If drug companies need to change their attitude, so do prescribers“. Andrew Jack, Financial Times correspondent writing in the BMJ July 2012.
Evidence update: Two books elaborate these concerns from a practitioner’s perspective. Dr Ben Goldacre’s ‘Bad Pharma’ and leader of the Nordic Cochrane Centre, Prof Peter Gøtzsche’s ‘Deadly Medicines and Organised Crime: How Big Pharma Has Corrupted Healthcare’. Peter’s alarmist chapter title: ‘Very few patients benefit from the drugs they take’ explains that “… apart from such scientific misconduct, insufficient blinding can also make us believe that ineffective drugs are effective.” In damning the exalted gold-standard comparison – the randomised control trial, he warns of assessor bias if they’re aware to whom they gave placebo. To give weight to this claim that doctors lie, his colleague Asbjørn Hróbjartsson is cited … the effect was exaggerated by 36% when evaluated by nonblinded observers. Wow, this Cochrane review ‘Observer bias in randomised clinical trials with binary outcomes: systematic review of trials with both blinded and non-blinded outcome assessors.’ needs a read. Ooops, the title omits the keyword ‘subjective’, and 71% of the studies were surgery or the like. Figure 1 meta-analysis shows that two weighty transmyocardial laser revascularisation outcome reports by Oesterle et al and Burkhoff et al dragged the conclusion from one of no significant difference to a non-blinded bias Odds Ratio of 0.64! So, if you ask F.I.G.J.A.M. whether they THINK the patient’s angina has improved recently, there’s a leaning towards affirmation if they’re aware that surgery has been done? Cardiologists genuinely believing that their interventions are beneficial can hardly be extrapolated to mandating that all trials must be blinded in order to avoid falsification of data, but that’s what’s happened. Peter’s Mentalaz speaking tour claimed that all anti-depressants were ineffective – arguing that severe side-effects results in unblinding. And the effect size was typically <36%, whereby he proves all benefit of the med resulted from bias! And these are the watchdogs? When Cochrane’s doctors stretch the truth about the 36% shift in truth by other research doctors, they’re all damned.
Cochrane pioneers Gotzsche and Chalmers are much alike in their evidence fudging. The Handbook warns that sneaky research doctors will try to break the blind and fudge the facts. The last para of ‘Rationale for concern about bias’ cites a study by Schultz, Altman, and Sir Iain Chalmers et al in Feb 1995 JAMA. Just one review, covering studies particular to pregnancy & childbirth actually contradicts their own argument: “Trials with inadequate sequence generation yielded estimates of treatment effects that were similar to those derived from trials with adequate sequence generation, after adjusting for the other three methodological dimensions”. But on pg410 a subgroup analysis limited to those trials reporting adequate allocate concealment only managed to find a statistically insignificant p=0.07 (ie close, but no cigar) exaggeration in effect size as a result of poor sequence generation.
This cracks me up. That’s a negative outcome elsewhere than the Cochrane Collaboration, and it relied upon data dredging to obtain a semi-significant conclusion. But it underpins one of their criteria for downgrade of all studies – sanctimonious dogma used thereafter in box-ticking exercises, which actually detracts from quality analysis of evidence trustworthiness. This recalls the classical Ouroboros, the snake eating itself – as a symbol of perpetuation.
*Aust Family Physician Oct 2013: “Competing interests: None”
<Jan 2016 update: an Austrian survey found 89% of GPs would refer fibromyalgia to a rheumatologist, but only 12% of those wanted to treat the patient. With great neuro research nearby by Uçeyler, Sommer & Hauser you’d hope for more than just a handpass: http://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0146149 >