Just how effective is modern medicine?

Only one in 10 of the drugs and treatments your doctor prescribes has solid proof it will work – and 20 percent of medicine may not work at all, with very poor evidence that it is effective.

Applying the most exacting of measures, researchers from Oxford University have discovered that most of medicine is supported by only moderate to poor evidence, which throws into question the vast sums invested in healthcare around the world every year – the US spends $3.6 trillion and the UK £197 billion, for example.

Just 10 percent of the scientific papers the researchers reviewed met the criteria of an evaluation system known as GRADE (grading of recommendations, assessment, development and evaluation), while 37 percent had “moderate” evidence they could work, 31 percent had “low” evidence, and 22 percent had “very low-quality” evidence – and yet all are treatments or drugs that are used daily by doctors.1

Even of the 15 studies that met the criteria, only two had “statistically significant” results, meaning the drug or treatment tested could be used with confidence by clinicians, who would know that the treatment or drug would very likely work. If this most extreme assessment were applied across all of medicine, just 2 percent of treatments would have absolute proof they work.

More is less

The Oxford researchers evaluated 154 reviews that had been carried out by the independent Cochrane Collaboration group between 2015 and 2019; this was a follow-up of an analysis the same team had carried out in 2016. In that earlier report, the researchers concluded that 13.5 percent of the 608 reviews they assessed had “high-quality evidence” – but in the three years since, the 154 reviews that had been updated had even poorer evidence the treatment worked. In other words, more research wasn’t producing greater proof of effectiveness.

As medicine prides itself on being a science, the trend is worrying. Lead researcher Jeremy Howick identifies several factors at play, and especially the pressure in academia to “publish or perish,” which leads to low-quality papers being produced. In PubMed, a database of medical studies, more than 12,000 new clinical trials are published every year, and the quality of research inevitably suffers, he says.2

“The evidence-based medicine movement has been banging a drum about the need to improve the quality of research for more than 30 years, but, paradoxically, there is no evidence that things have improved, despite a proliferation of guidelines and guidance,” he writes.

Another problem is the GRADE system itself, which sets the bar for evidence extremely high. For instance, it detects whether a study has been “blinded” – when participants don’t know whether or not they are getting the real drug or treatment. If a study isn’t blinded, the placebo effect can be more dramatic because a participant can think he is feeling better because he knows he is receiving the therapy, despite there being no actual improvement.

The GRADE system also looks exclusively at the primary outcome; for example, if a painkiller is being reviewed, the reduction of pain is the primary outcome being assessed. It doesn’t take into account any secondary benefits, such as better sleep or improved mobility.

But, says Howick, only a bad carpenter blames his tools, and criticism of the GRADE system shouldn’t mask the reality that much of medicine isn’t supported by reliable evidence.

Research fraud and spin also affect the quality of research. Most research is funded by the manufacturer, who benefits from a positive outcome. Dr Marcia Angell, former editor-in-chief of the New England Journal of Medicine, once estimated that more than 70 percent of all medical research was fraudulent.3

Researchers from the BMJ Evidence Centre – since rebranded BMJ Best Practice – came to similar conclusions when they also used the GRADE system. They calculated that just 11 percent of medicine was proven to be effective, and 23 percent was “likely to be beneficial,” with the majority either being unproven or even harmful, when they reviewed more than 3,000 therapies and drugs.4

Limited evidence, limited benefit. Is medicine a science or a guessing game? Just 10 percent of treatments reviewed by researchers at Oxford are supported by strong evidence, while over half have little or low-quality evidence of working. Meanwhile, only about a third of treatments reviewed by the BMJ are likely to benefit patients. Over half have no solid evidence for or against them, and the remainder may do more harm than good.

So why do it?

If so little of medicine has been proven to work, why do doctors still practice unproven or even dangerous treatments? The answer can be complex, but it boils down to two factors: that’s what they’ve been taught in medical school and that’s how their specialty has always done things, irrespective of the evidence, says medical researcher Eric Patashnik. It’s not helped that the evidence-based movement is weak and often fails to get its voice heard in a profession dominated by “standard care,” drug company influence and cash.

As a result, it’s the patient who suffers. “Millions of people are receiving drugs that aren’t helping them, operations that aren’t going to make them better, and scans and tests that do nothing beneficial for them and often cause harm,” said Atul Gawande, a surgeon and healthcare researcher.

One example where practice trumps evidence is the use of stents to open blocked arteries, says Patashnik. A major study discovered the procedure works no better than a placebo, or “sham” surgery, for non-emergency chest pain – but this hasn’t slowed the tide of operations, which cost as much as $41,000 each in the US.5

Another major research study failed to prompt even a rethink in orthopedics when it discovered a common knee operation didn’t relieve pain or improve mobility in osteoarthritis sufferers. The procedures – arthroscopic debridement, where damaged cartilage or bone is removed, and arthroscopic lavage, where any debris is then washed out – were also no better than a sham operation, the researchers concluded.6

Despite these findings, millions of the procedures are still carried out every year and earn the physician $5,000 a shot. Orthopedics didn’t only ignore the study; it kicked back against the findings. Surgeons and associations that represent the specialty disputed the study and even lobbied insurance companies to continue paying out.

The specialty was similarly unmoved by a pair of studies that discovered the most common procedure carried out by surgeons – arthroscopic surgery to repair a torn meniscus of the knee – was no better than a sham operation and that physical therapy was just as effective.7

The trouble is that the procedure is virtually an industry, as it’s carried out around 70,000 times a year in the US alone at a cost of $4 billlion to the healthcare system. “Those who do research have been gradually showing that this popular operation is not of very much value,” said Dr David Felson, a professor of medicine and epidemiology at Boston University.

As to the question of whether the surgeon is accomplishing anything, he replies, “I think often the answer is no.”8 Evidence on the effectiveness of other procedures, such as spinal fusion to ease pain caused by worn discs, and subacromial decompression, which is supposed to ease shoulder pain, is also in short supply, but the lack of evidence has barely registered or inspired even a review of the practices.

Let’s just stop

Evidence-based medicine is one initiative to stop the use of ineffective drugs and treatments; another has been Choosing Wisely, which was launched in 2012 to promote the use of therapies that are supported by evidence, do not repeat other tests, are free from harm and are truly necessary.

But as Patashnik observes, the campaign has had little impact on entrenched practices.

One review of the campaign’s effectiveness discovered it had made negligible changes to medical practices, even when there was clear evidence they were ineffective. One common, but unproven, procedure is to carry out imaging tests for cases of uncomplicated headache – but despite the group’s campaigning, the use of scans fell only slightly, from 14.9 percent to 13.4 percent. Slight as the success was, it was the best the campaign achieved; five other routine tests that are unnecessary or unproven are performed at the same levels today.9

Patashnik is pessimistic that anything will change. Politicians who control the purse strings are loathe to curb the excesses of medicine, partly because they get generous kickbacks from pharmaceutical lobby groups, but also because they know they are far less popular and trustworthy than doctors in the eyes of the electorate.

The other pressure point is the public itself. The patient wants the doctor to do something rather than nothing, even when there is no evidence to support the intervention. The fact that the consequences of the use of unproven treatments costs the US healthcare system an additional $226 billion every year doesn’t register when the alternative is for the doctor to shrug and say there’s nothing he can do.

The doctor wants to act; the patient wants something done. It’s the human factor, and that’s unmoved by the scientific evidence – or lack thereof.

References

1

J Clin Epidemiol, 2020; doi: 10.1016/j.jclinepi.2020.08.005; theconversation.com, Sep 2, 2020

2

The Truth about the Drug Companies: How They Deceive Us and What to Do about It (Random House, 2005)

3

“What Conclusions Has Clinical Evidence Drawn about What Works, What Doesn’t Based on Randomised Controlled Trial Evidence?” BMJ Evidence Centre, 2012; Mayo Clin Proc, 2013; 88: 790-8

4

Lancet, 2018; 391: 31-40

5

N Engl J Med, 2002; 347: 81-8

6

N Engl J Med 2013; 368:1675-1684; N Engl J Med, 2013; 369: 2515-2524

7

New York Times, December 25, 2013

8

JAMA Intern Med, 2015; 175: 1913-20