Modeling the Cost Effectiveness of Neuroimaging-Based Treatment of Acute Wake-Up Stroke.

BACKGROUND

Thrombolytic treatment (tissue-type plasminogen activator [tPA]) is only recommended for acute ischemic stroke patients with stroke onset time <4.5 hours. tPA is not recommended when stroke onset time is unknown. Diffusion-weighted MRI (DWI) and fluid attenuated inversion recovery (FLAIR) MRI mismatch information has been found to approximate stroke onset time with some accuracy. Therefore, we developed a micro-simulation model to project health outcomes and costs of MRI-based treatment decisions versus no treatment for acute wake-up stroke patients.

METHODS AND FINDINGS

The model assigned simulated patients a true stroke onset time from a specified probability distribution. DWI-FLAIR mismatch estimated stroke onset <4.5 hours with sensitivity and specificity of 0.62 and 0.78, respectively. Modified Rankin Scale (mRS) scores reflected tPA treatment effectiveness accounting for patients' true stroke onset time. Discounted lifetime costs and benefits (quality-adjusted life years [QALYs]) were projected for each strategy. Incremental cost-effectiveness ratios (ICERs) were calculated for the MRI-based strategy in base-case and sensitivity analyses. With no treatment, 45.1% of simulated patients experienced a good stroke outcome (mRS score 0-1). Under the MRI-based strategy, in which 17.0% of all patients received tPA despite stroke onset times >4.5 hours, 46.3% experienced a good stroke outcome. Lifetime discounted QALYs and costs were 5.312 and $88,247 for the no treatment strategy and 5.342 and $90,869 for the MRI-based strategy, resulting in an ICER of $88,000/QALY. Results were sensitive to variations in patient- and provider-specific factors such as sleep duration, hospital travel and door-to-needle times, as well as onset probability distribution, MRI specificity, and mRS utility values.

CONCLUSIONS

Our model-based findings suggest that an MRI-based treatment strategy for this population could be cost-effective and quantifies the impact that patient- and provider-specific factors, such as sleep duration, hospital travel and door-to-needle times, could have on the optimal decision for wake-up stroke patients.