When can we do research without consent? The tricky ethics of the AIROPLANE study

Examine, a free weekly newsletter covering science with a sceptical, evidence-based eye, is sent every Tuesday. Below is an excerpt – sign up to get the whole newsletter in your inbox.

This morning we published a story on a controversial new trial being run on some premature babies at Victorian hospitals known as AIROPLANE. The trial is designed to test the optimum level of oxygen support to give premature babies, who often need assistance to breathe in the first few minutes of life. Currently, different hospitals provide different oxygen concentration levels, ranging from 21 to 30 per cent – a range recommended by medical guidelines. However, there is no evidence showing either concentration is superior. The AIROPLANE trial is testing both ends of that spectrum to determine whether 21 or 30 per cent is best for the babies’ health.

The trial is scientifically important. The team running it, led by Associate Professor Louise Owen, are regarded by their peers as high-quality scientists with tremendous experience running similar research.

A new clinical trial testing different oxygen levels on preterm babies in Victorian hospitals – without telling their parents – has raised concerns among ethics experts and hospital staff.Credit:iStock

But it’s controversial because it’s going ahead without seeking consent from parents. Indeed, parents whose children were enrolled may never find out – there’s no plan to directly inform them, despite a requirement under the federal government’s research ethics policy to inform people they have been in a study “as soon as reasonably possible”.

The study raises several thorny ethical questions. Like: how do we balance the importance of research with the importance of consent? Is it ethical to do a study without consent if the risk is zero? And can you feel harmed – even if science says you aren’t?

Consent, paternalism

Modern scientists take consent very seriously. This is a big historical swing; in times past, doctors tended to have a paternalistic attitude. They knew best. This led most (in)famously to the Tuskegee experiment (1932-72), in which African American men with syphilis had their diagnosis concealed from them to study the long-term health effects of the disease – despite a highly-effective treatment being available.

AIROPLANE is nothing like that study. The scientists behind it have gone to great lengths to design a study they believe is ethical. I only mention the history because it’s important – it tells you why the barriers to doing research without consent are so high.

Consent protects patients from harm. Scientific philosophers also think it protects their personhood, their right to choose what happens to them (or their children).

“The idea is the person is worthwhile as a human being, and therefore you ought not to do something against their consent,” says Dr Diego Silva, senior lecturer in bioethics at the University of Sydney.

In general, human research in Australia needs to be approved by a Human Research Ethics Committee and AIROPLANE was approved by the Royal Children’s Hospital Human Research Ethics Committee.

Parents involved in the study won’t directly find out they were enrolled even after the study is completed. Why? The researchers tell me the study is so low-risk that’s an appropriate course of action. But the trial’s ethical review application hints at another reason: “Although unlikely, public disclosure of the results of this study could potentially raise ethical concerns if parents became retrospectively aware that their child had been involved in the study and this caused distress.”

Merle Spriggs, an honorary senior research fellow at the University of Melbourne and a pediatric bioethics expert, says “There seems to be this fear among medical researchers that criticising [AIROPLANE] will impede future child research. But if people find out that things have been done without consent, you’re undermining people’s trust in research.”

When is ‘low risk’ low risk?

The key reason the trial was allowed to operate without consent is because, according to the trial’s application form, it does not pose a risk to babies.

Why? Because the trial is testing oxygen concentrations of 21 and 30 per cent, and both levels are already used in Victoria at different hospitals. If a pre-term baby needs breathing support they would essentially get oxygen at a concentration level somewhere between 21 and 30 per cent, and the parents would never know the level. And there’s no evidence one concentration is better than the other.

“It would be perfectly reasonable for a hospital to toss a coin to decide whether a baby gets 30 per cent or 21 per cent oxygen,” Professor Paul Komesaroff, director of the Centre for Ethics in Medicine and Society, told me, “and there would be nothing unethical about that.”

But … I’m not convinced the argument quite fits the facts. According to AIROPLANE’s trial documents, about a quarter of Victorian hospitals allow doctors to set the oxygen between 21 and 30 per cent based on their clinical judgement.

This is not allowed as part of the study protocol. So the study does change the practice at some hospitals.

The researchers dismiss this worry. “Without evidence to guide [the doctors], there is little to support the validity of their judgement,” they told me. And they say local doctors at each site will have a choice whether to take part or not. “We cannot and do not impose research on individual clinicians or centres.”

Good argument. But some critics argue anything that takes medical decision-making away from doctors (even if they have chosen to participate) and gives it to researchers is automatically risky to the patient.

“The babies are being randomised,” says Spriggs. “The doctor isn’t looking at the individual child and using their medical judgement.”

But if we took that position, doing research becomes almost impossible, right? One of the great successes of modern medicine is subjecting to scrutiny a bunch of therapies we were “sure” worked – and often discovering they did not work at all. If we were unwilling to test them because we “knew” they were in the patients’ best interest, we’d struggle to do research.

“This is a long-standing debate. If you banned this research, you wouldn’t get evidence for good practice. You can see why waivers of this kind would be justified,” said Associate Professor Andrew Crowden, chair of the University of Queensland’s ethics advisory group.

“But even though they are very good reasons … still consent, or at least a conversation with parents, seems to be the caring, right thing to do. Because it’s not the doctor’s child.”

Subjective harms?

Let’s assume the research really does pose zero objective risk to the children enrolled. That’s still not the end of the story. “Harm,” says Silva, “is in the eye of the beholder.”

It is possible to do a research study that is objectively harmless but still makes a person feel harmed, he says.

A few decades ago, this wouldn’t have been an issue. Science knows best. Our attitude is different now.

What if, asks Dr David Hunter, a medical ethics researcher at the University of Adelaide, a child died and the parent later discovered the child had been enrolled in a trial without their knowledge? Worse, what if the trial found the treatment they had been randomly given was inferior to the other treatment? They might not be objectively harmed, but they would still feel harmed.

But you can also flip that argument: what if the parent gave consent and then the child died? The parent might feel responsible. “That’s a potential harm of seeking informed consent,” says Hunter.

Where does all this leave us? When my partner and I went to our maternity appointments, we consented to several trials. On balance, I think we would have consented to this one had we been asked. But the lack of transparency leaves a bad taste in the mouth.

The Examine newsletter explains and analyses science with a rigorous focus on the evidence. Sign up to get it each week.