[tt] NS: Should researchers tell over bad news brain scans?

Wed Jul 9 00:48:23 UTC 2008

Should researchers tell over bad news brain scans?
http://www.newscientist.com/article.ns?id=mg19926633.800&print=true
8.7.6
Peter Aldhous

SARAH HILGENBERG gave the "informed consent" paperwork only a
cursory read before she volunteered for a brain-imaging study in
September 2002. She had just started as a medical student at
Stanford University in California and was keen to help her new
friend Matt, who needed a few more volunteers to have an MRI scan
for his PhD on memory. She gave little thought to the warning that
"on occasion the investigator may notice a finding in a MRI scan
that seems abnormal".

The first clue that something was wrong came when Matt asked her
some odd questions. Was she having headaches? Any problems with her
eyesight? A few hours later, an agitated Matt called her at home.
"We found something on your scan," he said. Hilgenberg was rushed in
for tests, which revealed a knot of blood vessels called an
arteriovenous malformation. It was a potential time bomb, putting
her at risk of a brain haemorrhage. She was quickly scheduled for
surgery.

Thankfully, the operation was successful, and today Hilgenberg is a
paediatrics resident at the University of Washington in Seattle.
Still, she wonders if the incident could have been better handled.
"Receiving the news over the phone was hard," she says. After
consulting with a neurologist, Matt decided to bring Hilgenberg into
the emergency room. But was a worried phone call the best approach?
And what would he have done if the results had been more ambiguous?

Now there is a road map to help scientists and research volunteers
negotiate this ethical minefield. In the latest Journal of Law,
Medicine & Ethics (vol 36, p 219), a panel of bioethicists, lawyers
and scientists suggest ways to manage such situations. Their
recommendations include a guide on how to decide whether an
"incidental finding", such as a brain abnormality, is serious enough
to tell the volunteer about (see Table). They stress the importance
of developing a plan to assess incidental findings and communicate
them to volunteers, and the need to make this plan clear during the
informed consent process - in which volunteers are made aware of the
purpose of the study and any risks, and sign a form to that effect.

"This is the first major consensus statement on how to handle
incidental findings," says Susan Wolf, a specialist in health law at
the University of Minnesota in Minneapolis, who led the panel.

It is a difficult issue to tackle. Research projects are not
generally set up to diagnose disease, yet brain researchers are
increasingly carrying out MRI scans as part of their studies - and
they frequently discover abnormalities with medical implications
(see "Ill-prepared to communicate bad news"). Similar issues are
likely to emerge in genetics research, where studies are now
accumulating vast quantities of information from across each
volunteer's entire genome (see "Editorial: Dilemmas posed by chance
research results").

It would be straightforward if every case were like Hilgenberg's,
with a clear and positive outcome. But they are often more
complicated - as one anonymous neuroscientist revealed in 2005 in a
letter to Nature (vol 434, p 17). Like Hilgenberg, he volunteered
for an MRI study giving little thought to the consequences. The scan
revealed a golf-ball-sized tumour, creating a huge dilemma.

The neuroscientist was told the tumour could be removed by surgery,
but that there was a risk he would lose the ability to speak. At the
time, his wife was expecting her first child, and the couple were
applying for insurance to support the family should one of them lose
their job though injury or disease. After some deliberation, he
reported the scan results, and as a result lost the insurance
policy. The neuroscientist was left facing an operation that might
force him to leave his job, yet lacking loss-of-job insurance. "I
should have thought about the consequences of volunteering more
thoroughly," he concluded.

Such cases are especially troubling, because doctors do not always
know whether it is better to operate on a tumour identified on an
MRI scan before any symptoms appear, or to wait and only plan
surgery if the tumour starts causing problems.

Sometimes, abnormalities that show up in research MRI scans seem
benign, but the volunteer is monitored, just to be on the safe side,
which takes a financial and psychological toll. Bradley Peterson, a
neuropsychiatrist at the Columbia College of Physicians and Surgeons
in New York, had this experience with one of his research
volunteers. More than six years later, she still has occasional MRI
scans. "She's got to live with this worry of not knowing, and the
cost and the hassle," Peterson says. "Health insurance often doesn't
cover it in full."

What's more, many researchers are not trained to make clinical
diagnoses, and work with information that is not optimised for the
task. Research MRI scans, for instance, usually lack the definition
needed for proper diagnosis. This is usually explained in the
informed consent form, but volunteers do not always take this
message on board. In 2006, Judy Illes and her colleagues at Stanford
University surveyed more than 100 volunteers in MRI studies and
found that 54 per cent of them expected the scans to detect
abnormalities if they existed (Journal of Magnetic Resonance
Imaging, vol 23, p 205).

Some MRI researchers ask neuroradiologists to review research scans
that seem suspicious. And at the National Institutes of Health in
Bethesda, Maryland, all MRI research volunteers are routinely given
a clinical-grade scan. But this bumps up costs. Clinical scans cost
around $1500, compared to about $500 for a research scan. And Illes,
now at the University of British Columbia in Vancouver, Canada,
argues that clinical scans are not appropriate for many labs because
they blur the line between research and medicine, effectively
expecting scientists to have the same duty of care as doctors.

Given these complex ethical issues, Wolf accepts that her team's
recommendations will not be the final word. According to Illes, who
helped write the guidelines, the priority for now should be to
improve the informed consent process, so that volunteers clearly
understand that research projects cannot be expected to deliver
clinical diagnoses.

With the current ambiguity, some lawyers see an opportunity to
pursue researchers for damages. "I maintain that MRI is a medical
procedure," says Alan Milstein, an attorney who handles bioethics
cases at Sherman, Silverstein, Kohl, Rose & Podolsky in Pennsauken,
New Jersey. "The poor, uneducated mother of a child is going to see
someone in a white coat and they will expect care that is rendered
conforming to standards of using that machine."

Milstein argues that researchers who do brain imaging and fail to
spot health problems in their volunteers could face substantial
liability. He also raises the hypothetical case of a child who is
found to have a brain tumour on a research scan but whose family
cannot afford the operation needed to remove it. "There is a moral
and an ethical obligation to care for that child," claims Milstein,
who says he would be prepared to sue the researchers to make them
pay.

The Human Brain - With one hundred billion nerve cells, the
complexity is mind-boggling. Learn more in our cutting edge special
report.

Ill-prepared to communicate bad news

It's an ethical minefield, but how often do brain-imaging studies
throw up information that could affect the health of research
volunteers?

Estimates vary, but after scanning more than 600 volunteers during
the course of his research, Bradley Peterson of the Columbia College
of Physicians and Surgeons in New York found that about a third of
the scans contained potential abnormalities. Most were not deemed
worth following up - only 3.7 per cent were referred for further
evaluation (Journal of Law, Medicine & Ethics, vol 36, p 305).

In 2004, Judy Illes and her team at Stanford University in
California emailed a survey to researchers doing MRI studies. Of the
66 who replied to a question on incidental findings, 80 per cent
said they had found something untoward at least once (Journal of
Magnetic Resonance Imaging, vol 20, p 743). Despite this, only 53
per cent had standard procedures in place to handle such situations
and communicate the results to volunteers. There was also huge
variation in the extent to which neuroradiologists were called in to
help interpret the scans with just 13 per cent of respondents saying
that an expert reviewed every single one.

Related Articles

Diary of a human 'lab rat'
http://www.newscientist.com/article.ns?id=mg19626330.300
08 December 2007
Interview: clinical trials of life
http://www.newscientist.com/article.ns?id=mg19526231.800
26 September 2007
MRI-enabled brain surgery robot revealed
http://www.newscientist.com/article.ns?id=dn11631
18 April 2007

Weblinks

Susan Wolf, University of Minnesota
http://www.law.umn.edu/facultyprofiles/wolfs.html
Bradley Peterson, Columbia College of Physicians and Surgeons
http://childpsych.columbia.edu/brainimaging/Peterson.html
Judy Illes
http://www.chairs.gc.ca/web/chairholders/viewprofile_e.asp?id=2178
Alan Milstein, Sherman, Silverstein, Kohl, Rose & Podolsky
http://www.sskrplaw.com/attorneys/milstein/
Journal of Law, Medicine & Ethics
http://www.aslme.org/pub/jlme/index.php