The Life-Saving Medical Innovation You’ve Never Heard About

This tool allows researchers and clinicians to assess whether an intervention is beneficial or not!

In May of 1999, the Food and Drug Administration of the United States approved a brand new anti-inflammatory drug called Vioxx. Vioxx would become a successor to the less-specific anti-inflammatory drugs. These previous treatments caused safety complications.

Since clinicians reasoned this treatment would also benefit those with rheumatoid arthritis, another trial began. It aimed to determine whether or not it would treat this disease better than the gold-standard drug, naproxen. Naproxen functioned by stopping the action of a protein called cyclooxygenase to reduce inflammation. Unfortunately, this increased the serious-risk of gastrointestinal bleeding.

Differentiating itself from drugs such as aspirin by its specificity to a molecular target, it would not cause gastrointestinal bleeding. Since the mechanism of action was clear cut, the drug was approved while clinical trials to assess its safety and efficacy for rheumatoid arthritis were ongoing.

However, Vioxx came with other unexpected dangerous side-effects as it increased the risk of a heart attack. Even after the publication of their initial trials in the 2000s, the rich dataset of efficacy and side effects was unavailable to other scientists. The lack of transparency prevented other biostatisticians and researchers from independently vetting the data. Unfortunately, a lot of deaths occurred in part because the data was unavailable.

By 2004, the manufacturer Merck voluntarily withdrew Vioxx from the market after it had caused an additional estimated 88 000 heart attacks in the United States alone. The approval of Vioxx proved a deadly mistake, emphasizing the necessity of due diligence and a framework of evidence-based medicine for future treatments and trials. Common sense and knowledge was insufficient for calculating the risks of Vioxx treatment.

The Importance of Evidence-Based Medicine

Until relatively recently, we did not know what caused disease or how to treat it. Physicians often suggested courses of treatment without access to modern scientific knowledge. Sir William Osler, one of the founders of the renowned Johns Hopkins Hospital, remarked:

The practice of medicine is an art, based on science. Medicine is a science of uncertainty and an art of probability.

Although science and technology have greatly increased our understanding of the human body, there is still much to be discovered. It’s no surprise that following ‘logical’ predictions to inform the course of treatments is problematic. Sometimes, physiological responses contradict our expectations, putting patients in harm’s way. This concept is behind the push for evidence-based medicine.

Vioxx is an unfortunate example where our overconfidence with a logical mechanism allowed a risky treatment onto the market. Clinicians must have timely information to confirm these mechanisms while also ensuring the side-effects associated with treatments do not exceed its benefit. To provide an unbiased estimation of a treatment’s effectiveness, clinical researchers conducted a randomized clinical trial (RCT). The RCT is the backbone of evidence-based medicine.

Archie Cochrane was a Scottish epidemiologist born in the early 1900s. He was one of the greatest proponents of RCTs, hoped to establish clear guidelines for collecting evidence for different courses of treatment. Working in a hospital during World War II, he often found that many of the available treatments were not evidence-based. Feasible logical mechanisms and the discretion of the physician, rather than evidence, determining treatment. This proved especially frustrating for him, likely inspiring his advocacy of RCTs later in life.

I remember at the time reading one of those propaganda pamphlets, considerable suitable for POW medical officers about ‘clinical freedom and democracy.’ I found it impossible to understand. I had considerable clinical choice of therapy: my trouble was that I did not know which to use and when.

— Archie Cochrane, Effectiveness and efficiency: random reflections on health services. London: Nuffield Provincial Hospitals Trust; 1973.

Through the use of RCTs, he emphasized the necessity of randomizing participants into a control or treatment group. Flipping a coin or using another method for generating a random number to assign patients to groups independent of any human actions. Randomization effectively reduced the influence of unconscious biases and personalized-patient characteristics on treatment outcomes in the trial. At the end of the RCT, clinicians could calculate whether or not a treatment benefited participants. In 1971, Cochrane published a book called Effectiveness & Efficiency, outlining the importance of randomization in clinical trials while reiterating the need for evidence-based medicine.

Despite the utility of RCTs, resources are often limited across different hospitals and laboratories conducting studies. As a result, a study might be unable to recruit enough participants to find a beneficial effect. After all, clinically-relevant effects are missed due to this. Since it is difficult and costly to increase the sample size, many RCTs remained statistically underpowered.

Many under-powered studies on their own would allow for a better assessment of treatment. By combining data and patient information/outcomes from dozens of RCTs, researchers could better estimate the efficacy of treatment. Such a rigorous study is called a Systematic Review.

Systematic Reviews and the Cochrane Collaboration

The importance of these life-saving ideas cannot be understated. In the late 1980s, Crowley and colleagues found that different obstetricians and neonatal experts disagreed on a certain preemptive treatment for pre-term births. Specifically, they differed in opinions on corticosteroid treatment before birth to improve breathing difficulties.

While 12 studies had been conducted, each of them featured a small cohort, making it difficult to determine if indeed there were benefits. By combining these studies, researchers now worked with 3000 participants to determine the overall effect. Even though the studies individually did not show a clinical benefit, when taking all of these studies into account together, Crowley and colleagues found a life-saving benefit. Their findings, published in the British Journal of Obstetrics and Gynaecology in 1990, proved highly influential. The widespread adoption of this intervention improved lung development in pre-term babies, saving at least thousands of lives.

In 1993, the Cochrane Collaboration formed. This charitable, British organization mobilized scientists and clinicians to conduct more systematic reviews. The organization aimed to standardize and promote the effort to determine if treatments tested in multiple RCTs were beneficial. Time and time again, the Cochrane Collaboration produced systematic reviews of treatments/surgeries with higher quality and rigour than many other groups (see here, here and here).

While one individual study may be flawed, underpowered or methodologically-limited, systematic reviews allow for an overall assessment of a specific treatment. Rather than cherry-picking, relying on a single clinical trial, researchers are recognizing the importance of these systematic reviews.

Such collaborations emphasize the need for proper reporting of clinical trials, the need for data to be available to other groups as well as the overall benefits of RCTs.

While this intervention is not nearly as famous or well-known as antibiotics or hand-washing, it is nonetheless essential in determining the course of treatment.