Adoption of a tissue-based definition modestly increased stroke incidence by 6.5% and improved post-stroke survival by 3.9%, with greater impacts at younger ages, highlighting the importance of consistent definitions in stroke surveillance.
Key Findings
Results
The age-standardized incidence rate of first-ever stroke using the tissue-based definition was 161 per 100,000 person-years in the Normandy region of France.
Among 3,826 first-ever strokes meeting the tissue-based definition, median (IQR) age was 76 (64-85) years and 1,932 (50.5%) were female.
Age-standardized incidence rate (95% CI) was 161 (156-166) per 100,000 person-years using the ESP2013 standard.
Ischemic or undetermined stroke incidence was 132 (128-137) per 100,000 person-years.
Intracerebral hemorrhage incidence was 22 (20-24) per 100,000 person-years.
Subarachnoid hemorrhage incidence was 7 (6-8) per 100,000 person-years.
The study period spanned May 15, 2017 to December 31, 2022, in Caen-Normandy-Metropole, France.
Results
Transitioning from a time-based to a tissue-based stroke definition increased overall stroke incidence by 6.5%.
The tissue-based definition includes events with symptoms lasting less than 24 hours but with neuroimaging evidence of ischemic injury, in addition to events with symptoms lasting 24 hours or more.
The time-based definition required symptoms of at least 24 hours duration.
The 6.5% increase in incidence reflects the addition of cases previously classified as transient ischemic attacks under the time-based definition.
The study used multiple overlapping sources to identify stroke cases.
Results
The impact of the tissue-based definition on stroke incidence was greater at younger ages compared to older ages.
For all strokes, the tissue-based definition increased incidence by +8.5% in patients younger than 55 years versus +6.0% in patients 55 years or older.
For ischemic stroke specifically, the increase was +12.5% in patients younger than 55 years versus +7.3% in patients 55 years or older.
The age-dependent difference suggests younger patients more frequently present with short-duration symptoms but neuroimaging-confirmed ischemic injury.
Results
Three-year post-stroke mortality was 36.4% under the tissue-based definition, representing a 3.9% reduction compared to the time-based definition.
Three-year mortality was 36.4% (95% CI 34.9-38.0) under the tissue-based definition.
The tissue-based definition reduced 3-year mortality by 3.9% compared to the time-based definition.
The reduction in mortality is attributable to the addition of less severe cases (short-duration symptom events with neuroimaging evidence) that have better outcomes.
This survival difference has implications for temporal trend comparisons in stroke surveillance studies.
Discussion
The ICD-11 tissue-based stroke definition may introduce age-related differences in temporal trends when compared to studies using the traditional time-based definition.
The ICD-11 (11th edition of the International Classification of Diseases) incorporates a tissue-based definition of stroke.
Because the impact of the definition change is greater at younger ages, comparisons of incidence across time periods or between registries using different definitions may be disproportionately affected in younger age groups.
The authors highlight 'the importance of consistent definitions in stroke surveillance.'
Implementation of ICD-11 may create artificial changes in apparent temporal trends that are actually definitional rather than epidemiological.
Methods
The Normandy Stroke Study used a population-based design with multiple overlapping sources to identify stroke cases.
The study was conducted in Caen-Normandy-Metropole, France.
Events were classified using both time-based (symptoms ≥24 hours) and tissue-based (symptoms ≥24 hours or <24 hours with neuroimaging evidence of ischemic injury) definitions.
The study period ran from May 15, 2017 to December 31, 2022.
A total of 3,826 first-ever strokes met the tissue-based definition over the study period.
What This Means
This research examined how changing the definition of stroke affects the number of stroke cases counted and patient survival rates. Traditionally, a stroke was defined by symptoms lasting at least 24 hours. The new international disease classification system (ICD-11) uses a 'tissue-based' definition, which also counts cases where symptoms lasted less than 24 hours if brain imaging shows actual tissue damage. Using data from over 3,800 patients in Normandy, France collected between 2017 and 2022, researchers found that switching to the new definition increased the number of strokes counted by about 6.5% and made survival statistics look slightly better (3-year mortality dropped by 3.9%), because the newly added cases tend to be less severe.
A particularly important finding is that the impact of this definitional change is not equal across age groups. In patients younger than 55 years, the new definition increased ischemic stroke counts by 12.5%, compared to only 7.3% in those 55 and older. This means that when researchers or public health officials compare stroke rates over time or between countries, they need to be careful about whether they are comparing data collected under the same definition, especially when looking at trends in younger adults.
This research suggests that the shift to tissue-based stroke definitions—enabled by increasingly widespread brain imaging—will meaningfully affect how stroke statistics look, even if the underlying disease burden has not changed. Health authorities tracking stroke trends over time need to account for this definitional shift to avoid mistakenly concluding that stroke rates are rising or that survival is improving when the change may simply reflect how cases are being classified rather than true changes in disease patterns.
Nehme A, Schneckenburger R, Watrin M, Bouchart J, Boulanger M, Cogez J, et al.. (2026). Incidence and Post-Stroke Survival in the Normandy Stroke Population-Based Study (2017-2022): Impact of Time-Based Versus Tissue-Based Definitions.. Neurology. https://doi.org/10.1212/WNL.0000000000218457