Issue 41 / 22 October 2012

A SIMPLE clinical tool that could help non-specialists to predict the prognosis for acute ischaemic stroke patients has been welcomed by Australian doctors.

The bedside prediction rule for death and severe disability, developed by a team of researchers from Ireland and Canada, who based their calculations on data from patients included on the Registry of the Canadian Stroke Network, is described in the Archives of Internal Medicine. (1)

The PLAN score — based on nine clinical variables at the time of admission including preadmission comorbidities, level of consciousness, age and neurologic deficit — can be used at the time of admission.

The authors said the benefits of the tool compared to similar tools was that it was memorable, easy for non-specialist clinicians to use and possessed adequate discrimination to predict clinically important outcomes.

An accompanying editorial, said the PLAN score, while not perfect, would be of use to clinicians. (2)

“The PLAN rule can be calculated by nonspecialist clinicians and, unlike other prediction rules for outcome after ischemic stroke, does not require specialized skill in interpreting brain scans or assessing the severity of the stroke”, the editorial said.

However, it said future studies were needed to test the rule’s accuracy on a different set of patients to check if it enabled better decision making and planning for patients with ischaemic stroke.

Professor Graeme Hankey, from the department of neurology at Royal Perth Hospital, applauded the study as methodologically sound and novel, with important results.

“Although there are many other clinical prediction equations for death and dependency after stroke, the best being that developed by Counsell et al , they are not widely used, mainly because they are not simple enough”, he said. (3)

Professor Hankey said the beauty of PLAN was the simple scoring system that could be used by generalists.

However, he noted that it had not been validated outside the Canadian data set from which it has been derived.

“That needs to be done to see if it works in our own population of stroke patients, but one would suspect it would, as many of its features are shared by other prediction models in other populations”, he said.

Professor Craig Anderson, the senior director of neurological research at the George Institute, described PLAN as useful for generalists and for determining broad policy decisions.

However, he questioned its use in individual patients. In clinical practice, decisions needed to be made in the first few hours regarding thrombolysis, admission to the intensive care unit and withdrawal of life support.

“Stroke patients in Australia are now usually managed by a neurologist using advanced imaging and other investigations. We would probably use a lot of other variables to assess prognosis”, Professor Anderson said.

Even so, studies had yet to show whether these advances had translated into improved clinical outcomes, he said.

Associate Professor Mark Parsons, a neurologist at John Hunter Newcastle, agreed it was still very hard to predict prognosis in the first few days after stroke, but noted that advanced imaging tools offered much clearer predictive information.

“My view is that all acute stroke patients should have at least their initial care in a specialised stroke unit and, ideally, have access to advanced brain imaging”, he said.

While prognosis could be tricky, even when doctors did know which patients would do poorly, the problem often lay in acknowledging it, discussing this with the family and getting a consensus.

“If [it] encourages doctors to talk to families about end-of-life care, that would be a positive result as we don’t do that well in all areas of medicine”, he said.

– Amanda Bryan

1. Arch Intern Med 2012; Online 15 October
2. Arch Intern Med 2012; Online 15 October
3. Stroke 2002; 33: 1041-1047

Posted 22 October 2012

Leave a Reply

Your email address will not be published. Required fields are marked *