A US-based push to diagnose stroke on radiological findings alone is ‘seeping’ into Australian clinical practice, putting patients at risk of both over and under-diagnosis, an Australian stroke expert is warning.

 Since the early 2000s, the American Heart and Stroke Associations have put forward several proposals to define stroke and Transient Ischaemic Attack (TIA) based solely on brain imaging results.

 The move –  claimed to be needed to address challenges in acute stroke care – is dangerous and ‘unscientific’ and should not be adopted in Australia or elsewhere, according to Monash University Associate Professor Anne Abbott and colleagues in a new article in Frontiers in Neurology.

Associate Professor Abbott and her co-authors argue the proposals are encouraging the misdiagnosis and over-diagnosis in stroke in anyone who has a focal imaging abnormality (such as a white spot) show up on an MRI or CT brain scan.

While the proposed new definitions have not been adopted in the Australian clinical guidelines, Professor Abbott told the limbic she had witnessed neurologists using the proposed new model.

 “They are using the imaging findings to say that even if a deficit fully resolves in less than 24 hours, or less than a couple of minutes, because there was a ‘spot’ on your brain imaging you had a stroke,” said Professor Abbott, a neurologist from Knox Private hospital in Melbourne.

 “There has been no official acceptance of the definition, it’s just sort of seeping through.”

“However, we need to resist that. There are many limitations in using imaging to diagnose stroke. Plus, just about everything we know about how to best prevent and treat stroke is based on studies where stroke was defined using the clinical status of the patient first, then usually with assistance from imaging findings.”

 “For example,  we don’t need to see a ‘white spot’ on brain MRI before we thrombolise or do clot extraction for someone with sudden loss of body function due to a threatened stroke from a blocked brain artery. We just need to use imaging to rule out other causes of such a deficit, like a bleed, and rule out evidence of already established extensive brain damage. In fact, generally speaking, the more normal the brain imaging, the better.”

 “So anyone who’s trying to use the new proposed definitions is at high risk of working outside the evidence base because so little research relates specifically to identifying ‘positive’ imaging evidence of the reduced brain blood flow that occurs during a stroke or a TIA.”

Professor Abbott said the diagnoses of stroke and TIA using the presence and duration of clinical features (symptoms and examination findings), and subcategorising using imaging, is serving medicine well. Advances in neuroimaging are welcome and certainly help but they should not be relied upon in relative isolation.

“There are people in Europe who are trying to resist the push for change and we should do the same.”