Predictive Risk Model Determines Likelihood of Stroke in Patients

A simple predictive risk model helped forecast stroke risk in adult patients who have migraine with aura, according to research funded by the National Institute on Minority Health and Health Disparities.

Some people who have migraines also experience neurological symptoms that impact vision, known as visual aura. These symptoms can include flashes of light and blind spots or tingling in the hands and face. In adults, these indicators often come on before the headache itself, and younger women are at higher risk for migraine with aura.

“People who have migraine with aura are at increased risk for an ischemic stroke,” said Souvik Sen, MD, MPH, study co-author, and professor and chair of the neurology department at the University of South Carolina School of Medicine in Columbia, South Carolina. “With our new risk-prediction tool, we could start identifying those at higher risk, treat their risk factors and lower their risk of stroke.”

The team reviewed data from the Atherosclerosis Risk in Communities Cohort (ARIC), a community-based group of people from four US localities who have been followed since 1987. They assessed the data of 429 adults who had a history of migraine with aura who were in their 50s at their first visit. By the end of the study, participants were in their 70s.

Researchers identified five risk factors for stroke and developed a prediction model with each factor assigned a number of points proportional to its influence: Diabetes (seven points), age greater than 65 years (five points), heart rate variability (three points), high blood pressure (three points), and gender (one point).

Because the study group contained fewer men and migraines primarily affect women, the researchers added fewer points for females. After calculating scores for each person, researchers split participants into one of three groups: low-risk (zero to four points), intermediate-risk (five to ten points), and high-risk (eleven to 21 points).

Over an average of 18 years of follow-up, 32 people experienced a stroke. At the end of 18 years, three percent in the low-risk group had a stroke, eight percent in the intermediate-risk group had a stroke, and 34 percent in the high-risk group had a stroke. Compared to the low-risk group, migraine with aura sufferers in the high-risk group were about seven times more likely to have a stroke.

The group noted that they will need to test the risk model in a larger population before using it in a clinical setting.

Researchers have consistently explored the use of risk scores and predictive analytics to improve chronic disease management and preventive care delivery. A team from Intermountain Healthcare recently used predictive analytics to develop risk scores for patients with COPD, helping clinicians provide better care at the end of life.

The Laboratory-based Intermountain Validated Exacerbation (LIVE) score predicts all-cause mortality, morbidity, and hospitalization rates for patients with COPD. The risk scores can also predict which patients need palliative care referrals and advanced care planning resources.

“By exploring the association of palliative care referrals and LIVE score risk, this study is a step forward in understanding how the LIVE score may be used to target appropriate patient care,” said Denitza Blagev, MD, the study’s lead author, and a pulmonary and critical care physician at Intermountain Medical Center, who serves as medical director for Quality, Specialty Based Care at Intermountain Healthcare in Salt Lake City.

“Our findings lend more insight into how we can use these laboratory-based scores at the bedside to ensure that patients are receiving the most appropriate care. This doesn’t mean everyone with high risk needs to be referred to palliative care, but it shows potential opportunities to improve care for patients in that highest risk group.”