Sense about Science: First Admit No Harm

Editorial Note: This is a second post exploring Sense about Science. The first post Follow the Rhetoric is here.

Anyone interested in Pharma will know about its ability to Astroturf – to create patient organizations whose role is to promote an illness or subvert an existing one. Creating awareness of conditions sells drugs.

On a higher astral plane

Less well known is what happens at a higher Astral level. You can’t sell a product that gets a bad reputation or is removed from the market. The marketing mission at this meta-level is to risk manage by influencing the debate on Risk.

You do this by setting up think tanks, hiring ex-regulators, academics and others, capturing the regulatory system, and working with a body with a name like Sense about Science.

The regulators advise you on how to get an indication for a drug even though it doesn’t work for that. The academics advise on how to do trials that use a problem your drug causes to hide a problem your drug causes. The lawyers advise on the trials or studies that need doing in order for you to be able to defend the product in academic and legal settings.

All are involved years, perhaps a decade, before any sign of a legal action or public fuss. You are able to brief them on the likely legal actions or publicity you could face because the right hand already knows what problems your drug causes, even though the left hand never concedes anything – even after the drug has been removed from the market. It would be irresponsible to your shareholders not to have defences like this in place.

Sinister operations

But the key thing is to have groups help change the wider climate of debate – their efforts will make you seem mainstream. These are not the right hand (dextrous) experts you engage to take the known problems and hide them but the left hand (sinister) experts that take it as read you have been acting in good faith and don’t want to see you irrationally victimized.

Your, and accordingly their, mission is to transform poisons into fertilizers or even into vaccines. To transform inevitably risky chemicals into something where the greatest risks lie in non-use.

The climate changing operations of Pharma and its Sinisters have been so successful that two hundred years after Pinel’s famous medical dictum:

“The art of medicine lies in treating when possible but it is an even greater art to know when not to treat”,

medicine has been turned inside out and the essence of medical wisdom has become:

“The art of medicine lies in getting people on as many drugs as possible even though most of these will not help them; non-treatment is not acceptable in this day and age”.

And the Hippocratic Oath has become:

“First Admit no Harm”.

Clinical trials – the perfect marketing tool

In this scenario, RCTs, which initially seemed like a problem to industry, incorporated into the regulatory apparatus by Louis Lasagna, became a godsend (See The Tragedy of Lou Lasagna).

There should have been no excuse for not smelling a rat in the 1960s when industry began to use controlled trials to market their drugs.

The same applies to Evidence Based Medicine. Industry rapidly became one of the strongest proponents of EBM, and adherence to the guidelines it spawned. Most of the continuing medical education doctors have had on EBM has come from industry – with a particular focus on the hierarchies of evidence that put controlled trials on the top.

EBM can’t be blamed for trials that use surrogate outcomes, and are of inadequate duration, or for the fact that regulators will license products on the basis of two positive RCTs even if there are 10 or more negative studies, and the fact that pretty well all the publications that stem from these two positive studies are ghost-written. And Cochrane researchers have been among those who have established that companies can pull up to 50 publications out of just one trial.

But in a wonderful symbol of how Cochrane lost its moral (yes moral) compass, a 2009 Cochrane review by Cipriani and colleagues endorsed Zoloft as likely to be the most effective antidepressant when in fact it has debatable if any effectiveness. The Cochrane review was based on the published literature – when everyone except the most naïve must have known at that point that the published literature they were dealing with was likely to be close to entirely ghost-written and they didn’t know how many negative trials there might be. A year previously Erick Turner had published a widely cited article showing just how bad Zoloft really was.

Cochrane though is not one thing. In 2009, another set of Cochrane reviewers, Tom Jefferson and later Jefferson and Doshi, began publicly questioning whether the evidence would support the use of Tamiflu if the evidence were fully available. Their questioning culminated this year when Jefferson, Doshi and colleagues pretty well said flat out that it is not possible to do a Cochrane review on any treatment for which the data is not fully available.

Cochrane without data is like a fish with a bicycle

The wonder is that it has taken the field so long to nudge toward the importance of data.

Fourteen years earlier I lost a job on this very point. I gave a lecture in the U of Toronto which is widely cited as a lecture in which I claimed that antidepressants cause suicide when in fact the claim was:

“I believe that antidepressants cause suicide. Most people in this room likely don’t believe it. The problem we both have is that I have seen data that you haven’t. If you cannot see the data, this is just not science. And in a world where the data is not available conflict of interest becomes a big issue.”

Long before that I had written the Antidepressant Era (1995) which endorsed a recently published initiative from Iain Chalmers (1992), the founder of the Cochrane Collaboration, to have systematic reviews which he I imagine, and I certainly, at the time would have thought meant a review based on access to all the data.

But Iain and others veered off in a different direction. Iain was keen on trial registration, so that the field at least knew how many trials there had been and what remained out there unpublished.

This might make sense if RCTs are absolutely sacramental – the only form of evidence worth having. They aren’t. They are the gold standard way to hide adverse events – see A Black Box Warning for Clinical Trials.

Clinical trials are like men – they find it difficult to do more than one thing at a time. Ask them to focus on efficacy and they lose track of safety.

If Cochrane was an enterprise associated with bringing adverse events to light, and standing up for patients, it might be one thing to go along with proposals born from success at delivering the goods. But it’s not. Aside from Tamiflu, at the moment Cochrane is better known for declaring the statins to be side effect free. Against this background, where patients in the real world are concerned, making the registration of trials a key goal is primarily a box ticking exercise. Something for clinical trial collectors.

Bad medicine

This is where Bad Pharma comes in. By putting a premium on RCTs – everything would be okay if the RCTs were just done properly – its message is everything GSK and other companies could have wanted. See Not So Bad Pharma.

How could a book that catalogues industry failings and is called Bad Pharma be a win-win for Pharma?

There is nothing unprecedented about a book cataloging Pharma sins working wonderfully for Pharma. Some of the guiding spirits behind Sense about Science were once linked to a periodical called Living Marxism (see Follow the Rhetoric), which despite the title often delivered a message supporting corporate interests. It may have achieved its purposes because of its title.

It wouldn’t be any more breath-taking for good Pharma marketers to write a book called Bad Pharma. But the standard way to do these things is to support someone else who does it. This is a hazard for all of us. When Pharmacia wanted to market reboxetine – a non-SSRI – they were happy to have me on a platform to lay ouf the hazards of SSRIs.

A message that says RCTs are all that count works for Pharma because at the end of the day the only people in town able to play the RCT game are Pharma. Sure there were errors in how Pharma did trials, some of them egregious, but that was in the past. Sure there was a lack of transparency but hey we’re about to solve the perception of that one right now, further eroding medical discretion in the process.

What is data?

In the course of the debate about access to clinical trial data triggered by Peter Gøtzsche and the European Ombudsman, AllTrials emerged and called for trials to be registered and for data to be made available. Data however wasn’t defined and it seems to have morphed from data into Clinical Study Reports (CSRs).

CSRs aren’t data. CSRs are drawn from Clinical Record Forms (CRFs). CRFs are almost the data. But AllTrials is not calling for these because of clinical confidentiality issues.

CSRs that have been redacted are even further from the data. CSRs that have been redacted and are read through a periscope – the GSMA-ESK model – are further away again. A periscope that the GSK seagull regularly sits on obscuring the view. Is this what AllTrials has been maneuvered into supporting?

AllTrials and GSK and now EMA also endorse a model whereby it is only people with an analytic plan that should be let access anything. This is the kind of qualification that would be put forward by anyone who believes in RCTs and statistical significance – its not an approach associated with bringing adverse events to light. It is almost in breach of the legal settlement GSK have signed up to with US States.

The clinical trial blender

RCTs work for industry because they focus on one of the ninety-nine effects a drug has. Focusing on this one effect they nevertheless claim that they have told us all we need to know about this drug. If, despite being ignored, another effect of the drug pushes its way into view, standard RCT approaches will claim that if the effect wasn’t present to a statistically significant extent, it didn’t in fact happen.

Behind closed doors industry’s right hand dextrously files reports of adverse events that have been thoroughly assessed by clinical researchers and deemed linked to the drug. They will even in some cases override a researcher who doesn’t make the link to the drug and will say our drug caused this. But once the sinister left hand steps into the public domain the industry spokesperson will say there is absolutely no good scientific evidence our drug has caused this – by which they mean no statistically significant controlled trial evidence that our drug has caused this.

And if we don’t do the trial to look at this problem – who will? We of course can’t do such trials because it wouldn’t be ethical to do a trial to look for adverse events.

So even if all trials were registered, and none were ghostwritten, if you put poisons into the clinical trial blender you get fertilizer out the far end.

Doctors are increasingly faced on the one hand with people with clear and sometimes profound cognitive problems or profound weakness and fatigue on statins, or they are seeing people stop their statins and come back to life, while on the other hand the clinical trial evidence says these drugs have no problems. If doctors don’t prioritize the evidence of their own eyes over the clinical trial data they are going to go Blind. If the message they spin to patients is too at odds with reality, they are going to lose their patients trust – see Nearly Invisible. Evidence Based Medical Nemesis.

The very minimum ask for doctors and scientists is all the data (the CRFs). There can be no compromise here. Settling for redacted CSRs is a worse outcome than absolutely no access at all, because the redactions are primarily adverse events. Redacted CSRs will reinforce the message of efficacy and lead to greater prescribing. when the simple fact of redactions should bring prescribing to a full stop.

An unparalleled scandal

It’s difficult to see why we wouldn’t boycott their brands if companies don’t make the adverse events available.

Companies withhold the data on the basis of informed consent forms patients have signed. This cuts two ways. No doctor prescribing a branded pharmaceutical at present can be doing so with informed consent. They simply cannot tell the patient to whom they are giving a brand what the risks of taking that brand might be.

There is no need to boycott all drugs – one key brand might be picked.

Company clinical trials could be boycotted, unless their consent forms required data access. In the absence of access to the data, these are not the scientific exercises they once were. In the absence of access to the data everyone participating in a clinical trial puts anyone who might later take that drug in a state of legal jeopardy. See Healy D (1999). Clinical trials and legal jeopardy. Bulletin of Medical Ethics 153, 13-18.

This withholding of data on the part of Andrew Witty (AW1) and colleagues is a far far greater scandal than Andrew Wakefield (AW2) ever was.

AllTrials is probably still the group best placed to call for a boycott. But the ambiguities come to a head here – it has a track record of picking on a minnow like AW2 while partnering a science troll like AW1.

Dan Markingson

Meanwhile, if you are concerned about these issues a great petition to sign, or cause to donate to, is the Dan Markingson petition.

Dan Markingson was killed in a University of Minnesota clinical trial – but his death we are coming to realise is just the tip of an iceberg of RCT deaths and injuries. These are the new Disappeared in a Dirty War. See Brand Fascism. This is what Pharma are most keen to hide,

Dan Markingson should be the note to end this post on, except this would let GSK and Astra-Zeneca off the hook, so we’ll return to DM at the end the series of posts.

Follow the patient

Meanwhile we need an analysis of how the mess we live in has arisen and an account of what this means in patients’ lives. A good analysis would point to things that can be changed. It would get us beyond questions about whether clinical trial data should be downloaded or viewed through a GSMA-ESK scope.

There needs to be a recognition that patient confidentiality is a code for adverse events. Companies have little interest in your confidentiality or mine, they are primarily concerned to avoid information about adverse events coming to light.

Follow the money is a great lead for analyzing business. Follow the data is the key to analyzing science. In medicine, the patient is the data and the message is Follow the Patient.