EN / DA
Disease and treatment

Doctors’ standard treatments are rarely based on good scientific evidence

Much of the patient care in the acute care settings in hospitals is not based on good scientific evidence but relies mostly on tradition and doctors’ intuition. Danish researchers say that this is a huge problem.

Should a person arriving at a hospital’s emergency department be given oxygen, antibiotics, pain relievers or fluid – and how much?

Doctors in Denmark’s hospitals ask themselves these questions every day.

Doctors mostly follow the protocols they have always followed, but these are rarely based on solid scientific evidence.

Danish researchers believe that this is a major problem.

“The paradox is that we have often used treatment approaches for the past 50–60 years without examining whether they are actually effective and whether they benefit some people but harm others. More scientific studies are desperately needed to investigate the standard treatments we use in hospitals today,” explains Anders Perner, Professor, University of Copenhagen and Staff Specialist, Department of Intensive Care, Rigshospitalet.

Oxygen therapy increases the risk of death for some people

Anders Perner recently wrote an editorial on this topic in Intensive Care Medicine.

The editorial covers several new trials that are finally investigating whether the current practice of giving supplementary oxygen to patients with sepsis (blood poisoning) makes sense.

The trials conclude that this requires a balancing act that depends on the underlying disease.

For example, the new trials seem to indicate that people with a stroke or thrombosis in the heart do not benefit from oxygen therapy.

It may even harm those who have had cardiac arrest.

Conversely, people with trauma and those with sepsis seem to benefit from oxygen therapy, but this is still uncertain.

“We clearly need to be critical of the treatments we use on patients with these illnesses so that we will not unknowingly harm them ,” explains Anders Perner.

Intravenous fluid killed people with sepsis

Anders Perner conducts research on the standard treatments that are used in hospitals today.

His research results have shown that hydroxyethyl starch, a fluid that was previously standard care in many acute illnesses, actually harmed patients with sepsis as compared with a saline solution.

This discovery changed practices worldwide. In Denmark, this has saved 1000 lives and an estimated DKK 50 million in healthcare costs.

“In general, simple treatments are rarely tested, but the potential benefits of getting them right are high, because they can save people’s lives and save the healthcare system millions of kroner,” says Anders Perner.

Investigating oxygen therapy for people with acute respiratory failure

Anders Perner believes that more money is needed for clinical trials to investigate the standard treatments in the healthcare system.

He is conducting a trial in which researchers from Aalborg University Hospital and Rigshospitalet in Copenhagen in collaboration with hospitals in both Denmark and abroad are testing the most optimal way to use oxygen therapy for patients with acute respiratory failure.

The results will be ready in May 2020, and Anders Perner expects them to change standard practice.

“I really expect that the effectiveness of oxygen therapy depends on the underlying disease. I anticipate that Denmark’s clinical guidelines will be updated. The current guidelines also recognize that the clinical evidence behind the guidelines is quite weak,” he adds.

The lungs cannot process concentrated oxygen

Ander Perner explains that one can easily imagine why oxygen therapy can harm the body.

The lungs naturally evolved to process air with 21% oxygen, but supplementary oxygen can raise this to 50% or even 100%.

The lungs are not designed for this and, depending on the underlying disease, this can provide benefits or cause harm.

“The rational explanatory models underlying clinical practice focus almost exclusively on the fact that some people may benefit from oxygen therapy, but these forget the balance between effectiveness and side-effects. This is what we are trying to elucidate,” says Anders Perner.

Less than half of all standard treatments are based on research

Annmarie Touborg Lassen is Clinical Professor, University of Southern Denmark and Chief Physician, Department of Emergency Medicine, Odense University Hospital. She was not involved in producing Anders Perner’s editorial, but she has read it and fully agrees that more clinical trials are obviously needed.

Annmarie Touborg Lassen says that less than half the treatments doctors carry out today are based on some type of evidence.

In addition, this is extremely complex for many diseases.

For example, the clinical goal is to always strictly control people’s blood glucose. The problem is that the body may need to adjust blood glucose in a disease situation, and then this strict control can be harmful.

Oxygen is another example. A person clearly dies without oxygen, and oxygen can therefore save lives.

However, this does not necessarily mean that more oxygen is better. Drawing such conclusions uncritically can cost lives.

“We have a huge problem, since half the treatments to which we expose people are not evidence based. Some are good and some will probably turn out to be bad one day. Doctors are trying to do the best they can, but supportive clinical trials are lacking, especially for well-established treatments such as fluids and oxygen,” says Annmarie Touborg Lassen.

Major risk of publication bias related to standard treatments

Annmarie Touborg Lassen also says that publication bias is a completely separate problem.

When researchers examine the standard treatments in the clinic, they often focus on publishing data that reveal something surprising.

For example, if researchers analyse 40 standard treatments and only one does not work, this study has a greater chance of being published than any of the other 39.

This gives the impression that every time researchers examine standard treatments they find something wrong with them. The ones that work are not supported by published evidence.

“The single study showing lack of effectiveness is often the one published and not the other 39. Then determining effectiveness and publication bias is even harder. Conversely, if researchers find that a treatment is effective, then there is less incentive to complete the research and publish,” says Annmarie Touborg Lassen.

Society will not invest in acutely ill people

Annmarie Touborg Lassen and Anders Perner agree that the only way forward is to do more research and then even more research.

Trials are needed to reveal how each standard treatment works for each person with each individual disease.

But this raises another problem – money.

The pharmaceutical industry funds the vast majority of large and expensive clinical trials and can make money by showing that a treatment works.

For oxygen and saline solutions, however, industry cannot earn substantial profit, and therefore there is a reluctance to invest in this kind of research.

“It is a little strange that society will not invest in optimal treatments for the sickest patients. Unfortunately, the investment in research declines the closer we get to patients, especially the ones who are acutely ill. But this is where the efforts should be directed in future, and that is what the editorial focuses on,” says Anders Perner.

Trials on oxygen supplementation in sepsis: better late than never” has been published in Intensive Care Medicine. In 2019, the Novo Nordisk Foundation awarded a Distinguished Investigator grant under its Research Leader Programme to Anders Perner for the project IMPROVE-ICU – Improving Intensive Care through Clinical Trials.

Anders Perner
Professor
Anders Perner is a senior staff specialist in Intensive Care, Rigshospitalet and professor in intensive care at Copenhagen University Hospital. He chairs the Scandinavian Critical Care Trials Group and the strategic research programme ‘New resuscitation strategies in patients with severe sepsis’. This ongoing program will randomize 2000 patients with severe sepsis in three RCTs of resuscitation in 30 ICUs in all the Nordic countries. The first RCT, the 6S trial has been published and the second, the TRISS trial, has been finalized and the third, the CLASSIC trial, is protocolized.