Abstract W MP86: Improving the Accuracy of the ABC/2 Estimation Technique in Spontaneous Supratentorial Intracerebral Hemorrhage
Introduction: The ABC/2 formula is the most common, rapid, and effective method to estimate intracerebral hemorrhage (ICH) volumes. However, ABC/2 overestimates clot volume with increasing error for larger hematomas. We used linear and nonlinear regressions to find if adjusting C, adjusting the denominator, or adjusting both would improve ICH volume estimation.
Materials and methods: We assessed the diagnostic computed tomography (CT) scans of 373 patients enrolled in the MISTIE II (N=100) and CLEAR III (N=273) clinical trials using a standard central laboratory ABC/2 technique compared with planimetry volumetric calculations. Linear and nonlinear regressions were used to find optimal adjustments for C and the denominator, both individually and jointly. Results were stratified by study population to determine generality.
Results: In MISTIE II (M2) patients, mean hematoma volume was 45.72 ± 27.1 (SD) cm3 using ABC/2 vs. 39.48 ± 19.58 cm3 using planimetry. In CLEAR III (C3) patients, mean hematoma volume was 11.11±8.30 (ABC/2) vs. 9.58±7.10 cm3 (planimetry). The optimal denominator for the ABC approximation without adjusting C was 2.39 in M2 patients (95% CI 2.33, 2.46) and 2.37 in C3 patients (95% CI 2.31, 2.42). When these samples are combined, the optimal denominator was 2.39 (95% CI 2.35, 2.42). Without any correction to the denominator, the optimal adjustment for C was a decrease of 0.8352 in M2 patients (95% CI -0.976, -0.694) and a decrease of 0.5277 in C3 patients (95% CI -0.595, -0.46). In regression models adjusting C and the denominator, the optimal adjustment for C was an increase of 0.5707 in M2 patients (95% CI -0.215, 1.635), and an increase of 0.0485 in C3 patients (95% CI -0.383, 0.235). The optimal value of the denominator in these models was 2.64 in M2 patients (95% CI 2.29, 3.13) and 2.39 in C3 patients (95% CI 2.21, 2.62).
Conclusion: Empirical evidence from a large international population of ICH patients suggests that ABC/2 consistently over-estimates ICH volumes. Shrinking the C-axis of the ellipse improves fit, but the optimal adjustment depends on ICH size. Using a denominator of 2.4 is a simple, objective way to improve rapid ICH assessment with the existing measurement paradigm, which is consistent in both large and small hematomas.
Author Disclosures: W. Ziai: None. S. Nekoovaght-Tak: None. J.F. Betz: None. J. Muschelli: None. R.N. Fisico: None. X. Li: None. D.F. Hanley: Research Grant; Significant; MISTIE sponsored by NINDS, R01NS046309, CLEAR III sponsored by NINDS, U01NS062851. Other Research Support; Significant; Donations from Genentech, Inc. &Codman, Inc..
- © 2014 by American Heart Association, Inc.