Spontaneous intracerebral hemorrhage (ICH) affects over 2 million adults every year with one-month mortality of 40% and considerable morbidity in survivors. Accurate prognostication of neurologic outcome is critical for determining goals of care and appropriate medical or surgical interventions. Several risk scores can guide prognostication, such as the ICH score or Functional Outcome in Patients with Primary Intracerebral
Hemorrhage (FUNC) score. While these have proven clinically useful, they have inherent limitations resulting from the necessary simplicity of their calculation. With more advanced statistical methodology, significantly more accurate and informative patient-specific
prognostication would be possible. To accomplish this task we leverage the processing speed of modern mobile phones to power a predictive algorithm for the iPhone platforms. We first harmonized 3,562 patients from the ATACH-2, ERICH, CLEAR-III,
and MISTIE-III trials, using de-identified publicly available datasets and included patients with a pre-morbid modified Rankin Scale (mRS) ≤3 and no withdrawal of care or comfort care. Using regression models and marginal effects, the app will predict follow-up
mRS, both as a binary good outcome (0-3) versus poor outcome (4-6), split into quality-of-life informed categories (0/1, 2/3, 4, and 5/6), and the binary outcome of death.