Abstract Details

Title Validation of Models That Predict Mortality and Functional Outcomes at 3 Months after Ischemic Stroke

Topic Cerebrovascular Disease and Interventional Neurology

Presentation(s) S09 - (-)

Poster/Presentation
Number 005

Objective

Background We validate the models using two unique data sets, revealing the relative importance of various factors.

Design/Methods Our original models used only data clinically available in the first 48 hours from 451 subjects with ischemic stroke in 1999. We examined the effect of comorbidities in multiple, non-exclusive categories to further adjust the model. We subsequently collected two additional cohorts of ischemic stroke subjects: 460 and 496 in 2005 and 2010 respectively. In validation models, we held the variable structures constant but allowed the parameter estimates to vary and examined model performance. Mortality modeling was logistic multivariable regression, whereas outcomes models used linear multivariable regression treating modified Rankin scale (mRS) as a continuous variable.

Results The final mortality model included only age and post-stroke mRS. The C-statistic for these models was 0.80 in 1999, 0.86 in 2005, and 0.83 in 2010 respectively. In each period, these variables remained statistically significant. When predicting mRS at 3 months, the R[sup2] was 0.48, 0.60, and 0.50 in 1999, 2005, and 2010 respectively. Model variables included age, diabetes, severe periventricular white matter disease (versus less than severe), pre-stroke mRS, post-stroke mRS, and the retrospective NIHSS (per Williams et al.). When allowing parameter estimates to vary, the pre-stroke mRS, post-stroke mRS, and retrospective NIHSS were always statistically significant, and diabetes was significant in 1999 and 2010. This suggests that these variables most consistently impact the ultimate mRS. As in previous periods, subsequent comorbidities contributed to the predictive value of the model. For 2010, life threatening, psychiatric, infectious, and neurovascular comorbidities all significantly added to the predictive model as seen in prior periods.

Conclusions Our prediction models have proven to be robust, in consistently predicting mortality and functional outcome in subsequent study periods.

Authors/Disclosures
Brett M. Kissela, MD, MS, FAAN (University of Cincinnati Hospital) PRESENTER	The institution of Dr. Kissela has received research support from NIH/NINDS.
	No disclosure on file
Kathleen Alwell	No disclosure on file
	No disclosure on file
	No disclosure on file
Laura J. Balcer, MD, MSCE, FAAN (NYU Grossman School of Medicine)	Dr. Balcer has received personal compensation for serving as an employee of North American Neuro-Ophthalmology Society. An immediate family member of Dr. Balcer has received personal compensation in the range of $5,000-$9,999 for serving as a Consultant for Children's Hospital of Philadelphia.
	No disclosure on file
Daniel Woo, MD, FAAN (University at Buffalo)	The institution of Dr. Woo has received research support from NIH.
Matthew L. Flaherty, MD	Dr. Flaherty has received personal compensation for serving as an employee of Sense Diagnostics, Inc. Dr. Flaherty has received personal compensation in the range of $500-$4,999 for serving as a Consultant for Boeringher Engelheim. Dr. Flaherty has received personal compensation in the range of $5,000-$9,999 for serving on a Speakers Bureau for CSL Behring. Dr. Flaherty has received personal compensation in the range of $0-$499 for serving on a Speakers Bureau for Alexion. Dr. Flaherty has received personal compensation in the range of $10,000-$49,999 for serving as an Expert Witness for various law firms. Dr. Flaherty has stock in Sense Diagnostics, Inc. The institution of Dr. Flaherty has received research support from NINDS. Dr. Flaherty has received intellectual property interests from a discovery or technology relating to health care.
Pooja Khatri, MD, FAAN (Univ of Cincinnati/Dept of Neuro)	The institution of Dr. Khatri has received personal compensation in the range of $10,000-$49,999 for serving as a Consultant for Lumosa. Dr. Khatri has received personal compensation in the range of $10,000-$49,999 for serving as a Consultant for Bayer. The institution of Dr. Khatri has received personal compensation in the range of $500-$4,999 for serving on a Scientific Advisory or Data Safety Monitoring board for Diamedica. Dr. Khatri has received personal compensation in the range of $5,000-$9,999 for serving on a Scientific Advisory or Data Safety Monitoring board for Basking Biosciences. Dr. Khatri has received personal compensation in the range of $10,000-$49,999 for serving as an Editor, Associate Editor, or Editorial Advisory Board Member for American Heart Association. The institution of Dr. Khatri has received research support from Cerenovus. Dr. Khatri has received publishing royalties from a publication relating to health care.
Opeolu Adeoye	No disclosure on file
Simona Ferioli, MD (UCMC)	Dr. Ferioli has nothing to disclose.
Dawn O. Kleindorfer, MD, FAAN (University of Michigan Department of Neurology)	Dr. Kleindorfer has nothing to disclose.

��ɫ��

��ɫ��