A simple tool to predict dementia risk after stroke

 Are you that blitheringly stupid you think predictions are useful rather than EXACT PREVENTION PROTOCOLS? Don't bother answering, yes you ARE THAT STUPID!

You're fired!

A simple tool to predict dementia risk after stroke

A new calculator tool offers clinicians a quick, evidence-based way to identify patients at higher risk of post-stroke dementia, supporting earlier intervention and, ultimately, helping stroke survivors maintain independence and quality of life for as long as possible.

A stroke occurs every 11 minutes in Australia. It remains one of the nation’s biggest killers, but for the estimated 440,000 Australians living with the effects of stroke, survival marks the beginning of a long and often challenging recovery. While advances in acute treatment are saving lives that would once have been lost, many survivors are left with long-term disability. One of the most consequential and most overlooked challenges is post-stroke cognitive decline.

Dementia has overtaken heart disease as Australia’s leading cause of death, with an estimated 425,000 Australians currently living with the condition, a number projected to more than double in the next two decades. Alzheimer’s disease and vascular dementia are the two most common forms, with vascular dementia accounting for 15%–20% of cases. Stroke plays a major role in the development of vascular dementia, with up to 30% of stroke survivors developing dementia within five years.

Dementia is only the end point of a much broader continuum of cognitive decline after stroke. Around 70% of stroke survivors experience some form of cognitive deficit – difficulties with memory, attention, processing speed, or decision-making. These changes affect everyday life, from managing medications to returning to work, driving, or living independently. Yet these cognitive changes often slip under the radar. Overstretched stroke clinics rarely have the time or resources for routine cognitive testing. While rehabilitation services do an excellent job supporting physical recovery, cognitive changes tend to be more subtle, less visible, and therefore easily missed. Cognitive decline can also continue to worsen years after the initial event. As a result, post‑stroke dementia is often under‑recognised, under‑diagnosed and under‑managed.

A practical tool built for clinicians

Recognising that post‑stroke cognitive deficits are common but often not formally assessed in busy clinical settings motivated us to look for a practical solution. We didn’t want to create another complex scoring system requiring MRI scans or additional testing. We wanted a tool a GP or a stroke nurse could use in under a minute, and one tailored specifically to stroke survivors, given that general dementia risk scores designed for Alzheimer’s disease perform poorly in this population.

Working with the global Stroke and Cognition Consortium (STROKOG), we pooled data from 12 international studies, creating a dataset of 2 663 stroke survivors across Australia, Africa, Asia, Europe, and the US. Patients (average age 67 years; 40% women) were followed with detailed annual cognitive assessments for up to a decade to see who developed dementia.

We analysed factors already collected in routine care. The final model relies on six patient factors: age, sex, years of education, diabetes status, history of previous stroke and stroke severity. It performed strongly, achieving a C-statistic of 0.81, comparable to the widely used Framingham risk score for cardiovascular disease.

To ensure the model is clinically useful, we built it into a simple Excel‑based calculator. It takes less than a minute: answer seven questions, and it generates a five‑year dementia risk estimate. This allows clinicians to quickly identify patients who may need closer monitoring, comprehensive cognitive assessment, or a structured follow‑up plan. It serves as a fast “red flag”, particularly in busy clinical environments.

Like any new prediction tool, it requires further validation before being recommended for routine clinical use. Importantly, we need to test it in contemporary Australian stroke cohorts, including culturally diverse communities, to ensure it performs reliably across settings.

Why this tool matters for stroke care  

Current Australian and European guidelines recommend screening for cognitive impairment before discharge, but in busy stroke clinics this can be difficult to implement consistently. A quick, practical way to identify patients at higher risk can help bridge this gap and support clinicians to:

• identify patients who may need detailed cognitive assessment;
• guide decisions about monitoring and follow-up after discharge; and
• plan early interventions, supports, or referrals for those at high risk.

Closing the gaps in post‑stroke cognitive care

Alongside this tool, we see opportunities for simple system‑level improvements. Ideally, stroke clinics would be resourced to provide dedicated cognitive assessment, and rehabilitation centres equipped to deliver cognitive training and support. We also hope to see guidelines expanded to include long-term cognitive monitoring, ensuring clear care pathways are in place for patients identified as high-risk.

Finally, prevention remains key. Up to 80% of strokes are preventable through lifestyle measures such as regular physical exercise, a healthy diet, and managing high blood pressure, cholesterol, and blood glucose. Aggressively managing vascular risk factors, particularly in mid-life, remains one of our most effective strategies for reducing Australia’s long-term burden of stroke and dementia.

What’s next for this new tool

We are actively seeking collaborators to validate this tool in new settings. Please contact the author if you are interested in using the tool.

This tool is an important step toward more personalised cognitive care after stroke. It offers clinicians a quick, evidence-based way to identify patients at higher risk, supporting earlier intervention and, ultimately, helping stroke survivors maintain independence and quality of life for as long as possible.

Please note that this tool should be used alongside the accompanying Neurology article, which provides important guidance on suitable populations and cautions for interpreting results. Users are encouraged to read the full paper via Neurology before using the tool.

Jess Lo is a Research Associate at UNSW Sydney with expertise in biostatistics and more than a decade of research focused on post‑stroke cognitive decline and dementia. She coordinates the international Stroke and Cognition Consortium (STROKOG), guiding research that supports better recognition and understanding of risk factors for vascular dementia and the cognitive outcomes of stroke survivors.

The authors do not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and have disclosed no relevant affiliations beyond their academic appointment.  

The statements or opinions expressed in this article reflect the views of the authors and do not necessarily represent the official policy of the AMA, the MJA or InSight+ unless so stated.