To develop a method to help clinicians differentiate stroke from preeclampsia with severe features (PSF) in pregnant and postpartum patients presenting with acute symptoms.

Stroke is a leading cause of maternal morbidity and mortality, but diagnostic delays are common in this population due to maternal stroke’s relative rarity, and symptomatic overlap with other acute pregnancy-related syndromes such as PSF, which commonly presents with neurological symptoms.

We reviewed 71 consecutive cases of pregnant and postpartum patients with confirmed stroke of any type admitted to our academic medical center. Our control group consisted of 102 patients admitted with acute PSF. We compared clinical characteristics and presenting signs and symptoms between the two groups. We then used multivariable logistic regression models to develop a discriminatory score. Score validity was then tested by dividing the same cohort into derivation (n=140) and validation (n=33) datasets. Key variables were used to develop a bedside tool to distinguish acute stroke from PSF. 

Beta coefficients from the multivariable regression model yielding the best c-statistic (0.9) were used to develop the GO-SCAN score, which included Gestational diabetes (-1), Obesity (-1), Sudden headache (+7), Chronic hypertension (-1),  Altered mental status (+2), and Neurological focal deficits (+1).  The score range was −3 to +10. A score of ≥1 had 99% specificity and 54% sensitivity for diagnosing stroke, while a score of ≤-2 had 98% specificity for PSF without stroke. Scores between -1-0 had 27% specificity and 98% sensitivity for distinguishing stroke from PSF. 

Stroke in postpartum patients can be accurately distinguished from PSF using a combination of clinical signs and symptoms as well as patient medical history. The GO-SCAN score may be a useful clinical decision-making tool, but should be validated in larger, multicenter cohorts.