Premium
Quantitative Predictive Models for the Degree of Disability After Acute Ischemic Stroke
Author(s) -
Lim HyeongSeok,
Kim Seung Min,
Kang DongWha
Publication year - 2018
Publication title -
the journal of clinical pharmacology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.92
H-Index - 116
eISSN - 1552-4604
pISSN - 0091-2700
DOI - 10.1002/jcph.1039
Subject(s) - medicine , stroke (engine) , modified rankin scale , logistic regression , magnetic resonance imaging , diabetes mellitus , lesion , odds ratio , neuroimaging , covariate , cardiology , surgery , radiology , ischemic stroke , ischemia , mechanical engineering , statistics , mathematics , psychiatry , engineering , endocrinology
Although stroke is a leading cause of disability, the quantitative relationship between baseline clinical and imaging characteristics and long‐term disability outcomes has rarely been studied. Prospectively collected clinical data from 405 patients with acute ischemic stroke including brain magnetic resonance images (MRIs) and disability outcomes assessed using the modified Rankin Scale (mRS) 3 month after the onset of disease were analyzed using a proportional odds cumulative logit model implemented in NONMEM. The relationship between the difference in lesion volume (DLV) — lesion volume measured by brain MRI 5 days later — lesion volume at the onset of the disease, and the mRS measured at 3 months (mRS3) was modeled first, and the potential covariates were tested. The E max model best described the relationship between DLV and the logit probability of each mRS3. DLV, baseline stroke severity, age, and diabetes mellitus were identified as significant predictors of the probabilities of mRS3. The quantitative model constructed in the current analysis will enable us to predict the long‐term disabilities of the patients with acute ischemic stroke using the patient‐specific MRI and other clinical information, which will be useful for individualizing therapies and for making the clinical development of a novel drug more efficient.