Related Experiment Video
Updated: Aug 6, 2026
Pre-Chiasmatic, Single Injection of Autologous Blood to Induce Experimental Subarachnoid Hemorrhage in a Rat Model
Published on: June 18, 2021
Evaluating Utility-Weighted Modified Rankin Scale as an Outcome for Clinical Trials in Subarachnoid Hemorrhage: A
Ofer Sadan1, Zhuoran Yu2, Tianwen Ma2
1Department of Neurology and Neurosurgery, Division of Neurocritical Care, Emory University School of Medicine, 1365 Clifton Rd NE, Atlanta, GA, 30322, USA. ofer.sadan@gmail.com.
Introduction:
The utility-weighted modified Rankin Scale (UW-mRS) is an emerging alternative to the traditional ordinal mRS as a patient-centered measure of functional outcome in clinical trials. As a quasi-continuous parameter, the UW-mRS has been shown to better represent patient-centered outcomes and enable more efficient power calculations in acute ischemic stroke studies. However, its utility has not yet been tested in patients with subarachnoid hemorrhage (SAH).
Methods:
In a secondary analysis of the aneurysmal SubArachnoid Hemorrhage - Red blood cell transfusion And outcome (SAHaRA) randomized clinical trial, the UW-mRS, where higher scores reflect better outcomes, was computed by mapping each mRS category using utility scores reported by the patients, and in addition, using previously published utility weights to validate these weights in this population. The primary and secondary outcomes of SAHaRA were analyzed with the UW-mRS. Power calculations were compared across three possible outcome analytic approaches: binary mRS, ordinal mRS, and UW-mRS.
Results:
A total of 725 participants were included. Recalculated utility-weights from SAHaRA closely matched those previously published in ischemic stroke trials. At 12 months, 33.5% of patients in the liberal transfusion arm achieved the predefined unfavorable outcome (mRS ≥ 4) versus 37.7% in the restrictive arm (odds ratio [OR] 0.88, 95% confidence interval [CI] 0.72-1.07, p = 0.24). Using the UW-mRS, the average ± standard deviation in the liberal arm was 0.55 ± 0.39 compared with 0.52 ± 0.39 (linear regression β = 0.04, 95% CI - 0.02-0.09, p = 0.22). Power calculations, assuming 80% power and α = 0.05, indicated that replicating SAHaRA with dichotomized mRS would require 1900 participants per group, 1450 per group with the full-ordinal mRS, and 1460 per group with the UW-mRS. The UW-mRS demonstrated better goodness-of-fit to secondary patient-reported outcomes.
Conclusions:
The UW-mRS yielded results consistent with dichotomized mRS while preserving more patient-centered information, improving statistical efficiency, and showing stronger agreement with patient-reported outcomes.