Pediatric Stroke Rates May Be Double Previous US Estimates

Susan Jeffrey

September 18, 2009

September 18, 2009 (UPDATED September 21, 2009) — A new analysis from the Kaiser Pediatric Stroke Study searching both International Statistical Classification of Diseases and Related Health Problems, 9th revision (ICD-9), codes and imaging data finds a stroke incidence among children roughly double that of previous US estimates.

The report was published online September 17 and will appear in the November issue of Stroke.

Previous annualized estimates of stroke incidence among children, most based on ICD-9 codes alone, range from 0.54 to 1.2 per 100,000 children, but using a radiology search of the database in addition to diagnostic codes, researchers found an incidence of 2.4 per 100,000 person years.

"We found that the ICD-9 codes were pretty insensitive, particularly for neonatal strokes," which are instead often coded for encephalopathy, seizures, or eventual diagnoses of cerebral palsy, corresponding author Heather Fullerton, MD, from the Department of Neurology at the University of California–San Francisco, told Medscape Neurology.

In addition, because of issues with search terms, the authors were only able to fully establish the incidence of ischemic strokes, suggesting that the overall incidence of stroke including hemorrhagic events would be even higher.

"Essentially, it means now that we have a more accurate incidence estimate for ischemic stroke in particular and that that estimate is twice what prior estimates have been in the United States," Dr. Fullerton said.

Insensitive Measure

The researchers had suspected that there were issues with using diagnostic codes to determine incidence in pediatric stroke cases for several reasons, they write. For example, some previous reports have documented limited and variable accuracy of stroke ICD-9 codes even when applied to adults, and one such report showed the same limitation in the pediatric population.

The only prospective study looking at pediatric stroke incidence that would have avoided the problem with retrospective search terms was from a population-based study in Dijon, France, Dr. Fullerton noted. Using this stronger methodology, these researchers reported the highest incidence of pediatric stroke to date of 7.9 per 100,000 person-years (Giroud M, et al. J Clin Epidemiol. 1995;48:1343–1348).

Finally, in her own experience, Dr. Fullerton reports that there seems to be a general reluctance to use stroke diagnostic codes in children, "because there's often a perception that stroke is not a pediatric disease. I've even had insurance companies call me when I've seen children in my stroke clinic and done my billing, and say there's a problem with the coding I used because the code can't be applied to children."In the current paper, the researchers used data from the Kaiser Pediatric Stroke Study, a retrospective cohort study with the aim of identifying all stroke cases among children aged newborn to 19 years enrolled in the Kaiser Permanente Medical Care Program managed care plan. They searched records first, using ICD-9 codes suggestive of stroke or cerebral palsy, and next they searched radiology reports for terms suggestive of infarction.

Cases were confirmed through chart review. In all, from the Kaiser population of 2.3 million children, they identified 1307 potential cases using ICD-9 codes, and 510 potential cases from the radiology search. Of these, 205 ischemic stroke cases were confirmed, for an incidence of 2.4 per 100,000 person-years.

The radiology search had higher sensitivity than the ICD-9 search (83% vs 39%), although both had a low positive predictive value, the authors write. For perinatal stroke, the sensitivity of the stroke ICD-9 codes was even lower, at 12% vs 57% for the combined stroke and cerebral palsy codes, and the radiology search was again more sensitive at 87%.

"I think just in general, more emphasis needs to be placed on prospective studies," Dr. Fullerton said. "The issue with prospective studies is that they are also generally more costly. Basically, it means there needs to be a larger investment in pediatric stroke research, because the research that we try to do retrospectively to save money might not give us the right answers."

More Potential for Recovery?

Asked by Medscape Neurology for some perspective on these findings, Stephen Ashwal, MD, from the division of child neurology at Loma Linda University School of Medicine, California, speaking on behalf of the American Academy of Neurology, called this an important paper.

"It's an important paper because it shows that a problem that was thought to be very rare is really much more common," he said. The discovery is largely due to new advances in neuroimaging that allow closer examination of the brain when a child or infant is admitted for acute neurologic insult.

"Identification of strokes in this population is critical both because children may face a lifelong disability from stroke damage and because, as their brains are still developing, they may be able to achieve much better functional recovery than would an adult patient with the same intervention, he added. "Lots of people are looking at enrichment therapies or constraints therapies in adults, but they're also starting to do this in young children, and it may be that a young child who has a stroke has a better potential to recover."

Epidemiological studies like this one, then, are a critical first step in addressing this issue, Dr. Ashwal concluded. "Children from newborns to adolescents can have strokes," he said. "It's important for us to accept that and understand that, and then try to figure out what causes that as well as finding better ways of treating it."

The study was supported by the American Heart Association Scientific Development Grant, Neurological Sciences Academic Development Award and Independent Scientist Award. The authors have disclosed no relevant financial relationships.

Stroke. Published online September 17, 2009.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: