“…Potentially eligible studies for the meta‐analysis (n = 43) were retained, after screening both the titles and abstracts of all studies. After retrieving the full‐text version of the aforementioned 43 studies, 8 studies were excluded because they were retrospective cohort studies, included mixed population (non‐IS/TIA patients as the majority of study population), or evaluated exclusively nonstroke patients . The remaining 35 studies, comprising 3,067 total patients (89.7% with cryptogenic IS/TIA, mean age = 50 years, 57.3% males), were included in both the qualitative and quantitative synthesis (Fig ).…”