Scientific progress is not
linear. Newer does not automatically equate to better.
Rapid uptake of treatments with decreased critical evaluation is driven by a number of factors including patient demand, industry and ‘gizmo idolatry’– the new, shiny drug or other technology is considered by many to be a better bet than the alternatives. These new gizmos are more enticing – and not just on a profit basis- the latest, the greatest, they are laden with assumptions of advancement and laced with hope for the most desperate. Imagine turning a drug down, a new drug, with the hype and the sparkle and the plausible science, where accepting it means the newest, the best, surely? accepting it means hope?
Gizmo – Gremlins (1984)
Drugs that sound like they make sense and work in animal models may have catastrophic effects in healthy people – TGN1412 is just one example. Even in drugs which have completed Stage 1, 2 and 3 trials, true adverse effects and true outcomes may remain unknown – still relatively few people will have taken the drug overall. The results are often unpublished, undisclosed even to regulators, advisory or commissioning authorities and outcomes are tampered with to spin a drug company’s data for patenting and marketing purposes. In the population however, of unwell people with comorbidities, polypharmacy and physiological and pharmacogenetic differences, a drug that appears to be a logical, new, shiny upgrade on the old one may be anything but. Large scale accurate trials of a drug and the open publication of all results are essential to understand a drug’s real spectrum of effects in vivo. The sooner that the data is available and may be seen transparently the sooner a drug’s therapeutic and all other effects will be determined.
The consequences of not having this data are clear. The drug (or any medical technology/device/procedure) will be in use in the population for longer before the effects become apparent. In a new drug likely to be used in a significant number of people this is, obviously, a problem – the sooner a drug is introduced and taken up on a larger or ad-hoc scale, the longer that drug is in use without adequate assessment and data, the more people there will be who take the drug and over a longer period of time and yes, these people may be quite happy (for a while). They might have campaigned vociferously for access to the newest drug to treat their condition.
It’s shiny. It’s new. It’s better than the others. Hope and hype are all powerful and entirely understandable. But that drug might harm. It might have many effects which were not apparent in the data presented to NICE or the EMA or MHRA that was likely funded and chosen by the pharmaceutical company manufacturing it, it won’t feature in ghost written and false-meta-meta analyses of a selected group or even the genuine articles, for it’s simply not available; and it certainly may harm those patients who are taking it. If a drug has an effect, side effects are normal and idiosyncratic reactions always possible. A wider responsibility exists to know what any risks are at least most likely to be before dishing the drugs out to great fanfare to big groups, and a responsibility to report adverse reactions appropriately thereafter through reporting schemes that are in place.
A responsibility to not create unnecessarily or disproportionately heightened risks of harm – to treat the individual in their own best interests while maximising patient safety as much as possible. To first, do no (entirely avoidable, unjustifiable and reckless) harm. Surely that’s what patients would be campaigning vociferously for, too?
It is an unfortunate fact that many millions of people have taken drugs that were harmful and they did not find out until long after the harm was done. Others may have had an inkling before their last breath but it’s quite unlikely they realised they were being killed by their medicine. We’ve all heard of thalidomide yet there are so many more examples of drugs that, when taken up too quickly, have had devastating effects. Many others still have prolonged unnecessary expense or increased relative deterioration by being no more or considerably less effective than existing options.
Take Stevie Wonder. He was given oxygen as a neonate. Makes sense right? Sure it made sense to give high-dose oxygen to premature neonates. Unfortunately, this was in place for over 12 years before a causal link was made to ROP- Retinopathy of Prematurity (aka RLF). Due to this delay and lack of a link, the therapy harmed many infants, causing permanent blindness in more than 10,000 – including Stevie.
Women with breast cancer petitioned, in their tens of thousands, for bone marrow transplant to accompany high dose chemotherapy. It made sense – it would allow better toleration of a higher dose of treatment, right? The pressure was enormous from patient advocates, lawyers and even their doctors. Some launched legal actions to gain access to this new treatment. Unfortunately despite the hype and intense demand, it was much more expensive yet conferred no benefit at all. Worse, it was far more toxic- leading to greatly impaired life quality and hastened deaths. More than 42,000 women received the treatment before it was stopped – at a cost of $3.4bn to the US.
Baycol (cerivastatin) was a hugely profitable snazzy new statin taken by more than 6 million people. Unfortunately, although a new upgrade, it conferred no benefit compared to any existing treatment. Actually, it transpired that it had a serious adverse risk profile, including things like rhabdomyolisis. Including things like death. It was directly implicated in many deaths – in reality thought accountable for hundreds, thousands of deaths and withdrawn by Bayer amidst legal disaster. Other disasters include Diethystilbestrol (DES), given in pregnancy to prevent miscarriage and widely used for other purposes, becoming routine. Unfortunately as there was no reliable data it took decades to link DES as the cause of cancer and other illness in the children of pregnant women, in which time thousands of women and children were harmed and/or died. The anti-arrythmics Encainide and Flecainide caused excess mortality – with 50,000 deaths that were entirely unexpected and unexplained – 50,000 people who died when the drug companies had failed to undertake trials to observe any potential benefit, and surrogate markers and assumptions had been relied upon. The ALLHAT trial finally revealed that, while a drug may confer little benefit over another for one variable, in this case BP, it can be very different in its effects on another- here, its effects on preventing MI. Huge numbers of patients died unnecessarily through the delay in realising that they were being prescribed a drug that did not work as well as the other on the outcome they should have been interested in, rather than a surrogate. While they were dying, the drug company were cashing in as the newer much snazzier drug was much more expensive.
Something else- something that really made sense- was giving corticosteroids to patients who had acute head injuries. People did this, but there hadn’t been a trial. It just made sense – let’s reduce that swelling- and it had been the case for more than 30 years. A review came about in 1997 and suggested perhaps this treatment conferred a 1-2% reduction in the risk of death. Many doctors deployed this treatment. Others were less convinced and did not, but nobody knew. A trial was finally launched. The CRASH1 trial recruited more than 10,000 patients. It was hard to get approval, randomising unconscious people (who would have been randomised anyway in real life). The results showed an excess mortality in patients treated with IV methylprednisolone compared with placebo – what was immediately clear was that the treatment conferred no survival benefit at all, no advantage for these patients. Indeed, the steroids given to so many had been positively harmful. While a quarter of patients died from their injuries no matter what treatment they received, there were 2.5 extra deaths per 100 for those treated with steroids. ‘The risk of death was higher in the corticosteroid group than in the placebo group (1248 [25.7%] vs 1075 [22.3%] deaths; relative risk 1.15, 95% CI 1.07-1.24; p=0.0001), as was the risk of death or severe disability (1828 [38.1%] vs 1728 [36.3%] dead or severely disabled; 1.05, 0.99-1.10; p=0.079).’ In real world numbers of patients, that’s a lot of extra death. So unequivocal was the result, the trial was stopped early. ‘The lethal effects we have shown might have been found decades ago’ wrote the study’s authors, lamenting a failure to ‘provide robust evidence’ to prove that ‘ prescriptions are likely to do more good than harm’.
Lord Saatchi’s Bill as he proposed would not benefit patients who want access to a treatment – it provides no right or means of access, and would not benefit doctors who are able to innovate responsibly under the current law. Its selling of the register of uncontrolled experimentation (that didn’t exist) and promises of trying something out would prolong the time taken for reliable evidence to become available, reducing the evidence quantity and quality, knowingly promoting an archaic method which lengthens time and heightens risks sustained from any technologies that are unsafe.
Treating innovatively in a patient’s best interests, even where a doctor doesn’t know which is the better method, or has no data, is really not going to be found negligent under the current law, unless it’s so totally and ridiculously illogical and unsupported by any strand of reality. But any treatment currently so so bad that it would be considered negligent would be defensible under the Saatchi Bill’s proposals -and what of cases of product liability and serious adverse drug effects? No redress. No ethics committee equals no REC insurance cover. No implications at all for a pharmaceutical company with a harmful product?
Interventions should be widely introduced in a critical, carefully targeted manner and yes, reliable evidence is required to ensure there is no unnecessary, excessive risk of harm to a patient or a larger group of patients. Without accurate and systematic monitoring and reporting there will be diminished evidence regarding a drug’s safety or efficacy, widening the gap in knowledge required also for properly informing and gaining consent from a patient. Rapid adoption of treatment with decreased critical evaluation has led to countless adverse events in recent history. While many advances offer real advantages, others can be aggressively marketed and championed yet offer little, if any benefit; cause alarming side effects, serious harm, suffering and even death.
There are problems with aspects of research and regulatory frameworks, there are problems with data access and publications, there are problems with bureaucracy and huge problems with funding. A broken system perhaps. But you cannot mend a broken plate by smashing up your entire kitchen, destroying an aga and removing the sink. While it’s possible to advocate for a number of methods to assess an intervention, it is impossible to rationally, or innovatively, advocate that anecdote, uncontrolled and irresponsible experimentation is anything other than a timehop straight back to an era we’d rather not repeat. This imprecise, outdated approach is markedly out of step with current practice, as well as out of step with increasing precision approaches, targeting a drug to genomic features which are increasingly in use and can be incorporated into trial methodologies worldwide for some patients. The Bill would have been an inappropriate step backward that places current and future patients at considerable risk of harm, delaying beneficial data as well. It’s the opposite of innovative, the opposite of an advancement and patients deserve better. Patients deserve their lives to be valued. There is no excuse for this today.
The Medical Innovation Bill, full of hope and packaged with hype- would have harmed, not helped.