Abstract 54: A Novel Technique in Controlling for Severity of Subarachnoid Hemorrhage and Measuring Outcomes in the Analysis of the Nationwide Inpatient Sample: The NIS-SAH Severity Score and NIS-SAH Outcome Measure
Introduction: The Nationwide Inpatient Sample (NIS) is the largest publicly available inpatient care database. Its use in subarachnoid hemorrhage (SAH) outcome studies, has been limited by an inability to control for SAH severity, and unverified outcome measures. To address this, we developed the NIS-SAH Severity Score (NIS-SSS) and NIS-SAH Outcome Measure (NIS-SOM).
Methods: Three patient populations were used. Pop-I (N=148,958) and -II (N=147,395) were derived from the 1998-2009 NIS. Pop-III (N = 716) was derived from our institution. Using Pop-I, diagnosis codes likely to predict poor outcome were used to create the NIS-SOM. To create the NIS-SSS, diagnoses likely to predict SAH severity were entered into a regression model assessing poor outcome; model coefficients of significant factors were summed to generate the NIS-SSS. NIS-SOM was validated in Pop-III against modified Rankin Score (mRS) using inter-rater reliability; NIS-SSS was validated against Hunt-Hess (HH) grade using ANOVA; and the ability of NIS-SSS to predict poor outcome by mRS was analyzed using logistic regression. Using Pop-II, the ability of NIS-SSS to predict poor outcome was compared to other common measures of disease severity (APR-DRG, APS-DRG, and DRG).
Results: A strong correlation between NIS-SOM and mRS was found, with agreement and Kappa statistic of 85% and 0.63 when poor outcome was defined by mRS > 2 and 95% and 0.84 when poor outcome was defined by mRS > 3. The NIS-SSS significantly correlated to HH grade (Fig. 1A) and strongly predicted poor outcome (Fig. 1B).
When compared to APR-DRG, APS-DRG, and DRG, NIS-SSS was more accurate in predicting SAH outcome (AUC = 0.69, 0.71, 0.71, and 0.79, respectively).
Conclusion: Our results indicate that: NIS-SOM is a valid measure for SAH outcome, NIS-SSS is a valid measure for SAH severity, and NIS-SSS outperforms other measures of disease severity in predicting SAH outcome. We believe NIS-SOM and NIS-SSS should be included in future NIS analyses.
- © 2012 by American Heart Association, Inc.