LnEASIX is a simple, robust, and continuous predictor of short- and long-term mortality in critically ill patients with ischemic stroke, and integrating this zero-cost biomarker with established intensive care scoring systems provides significant incremental value for early prognostic risk stratification.
Key Findings
Results
A per-unit increase in lnEASIX was independently associated with elevated 28-day mortality risk in critically ill ischemic stroke patients.
OR 1.47 (95% CI 1.36–1.59, p < 0.001) for 28-day mortality as the primary endpoint
Analysis was conducted on 3063 patients with ischemic stroke from the MIMIC-IV database
lnEASIX was calculated from admission LDH, creatinine, and platelet values
Multivariable logistic regression was used to establish independent association after adjustment
Results
lnEASIX was independently associated with all secondary mortality endpoints including in-hospital, 90-day, and 180-day mortality.
Secondary outcomes included in-hospital, 90-day, and 180-day mortality
Cox proportional hazards regression was used for time-to-event secondary outcomes
The association was consistent across all time points evaluated
Results
Restricted cubic spline analysis confirmed a strictly linear, monotonic dose-response relationship between lnEASIX and mortality.
P for nonlinearity > 0.05, confirming linearity of the relationship
The relationship was described as 'strictly linear, monotonic dose-response'
RCS analysis was used to assess the functional form of the association
Results
lnEASIX demonstrated moderate standalone discriminative ability for mortality prediction, comparable to SOFA score.
AUC ranged from 0.64 to 0.67 across mortality endpoints
Standalone discrimination was described as 'moderate' and 'comparable to SOFA'
This AUC range was observed across multiple mortality time horizons (28-day, in-hospital, 90-day, 180-day)
Results
Integrating lnEASIX with baseline SOFA score yielded significant incremental predictive value over SOFA alone.
Net Reclassification Improvement (NRI) = 0.208 (p < 0.01)
Decision curve analysis showed superior net clinical benefit when lnEASIX was added to SOFA
Advanced clinical utility analyses including NRI, IDI, and decision curve analysis were all utilized
Results
Subgroup analysis showed that the prognostic impact of lnEASIX was markedly more pronounced in patients without pre-existing cancer.
A statistically significant interaction was found for cancer status (p for interaction < 0.05)
Prognostic utility was described as consistent across most other subgroup strata
The cancer subgroup was the primary modifier of the lnEASIX-mortality association
Methods
The study analyzed 3063 critically ill ischemic stroke patients from the MIMIC-IV database using multiple analytical approaches.
Sample size: 3063 patients with ischemic stroke
Data source: MIMIC-IV database
EASIX was natural log-transformed (lnEASIX) for analysis
Analytical methods included multivariable logistic regression, Cox proportional hazards regression, RCS, NRI, IDI, and decision curve analysis
What This Means
This research suggests that a simple blood-test-based score called EASIX (Endothelial Activation and Stress Index), calculated from three routine laboratory values — lactate dehydrogenase (LDH), creatinine, and platelets — can predict the risk of death in critically ill patients who have suffered an ischemic stroke. Using data from over 3,000 stroke patients in a large hospital database (MIMIC-IV), the researchers found that higher lnEASIX scores at hospital admission were consistently linked to higher rates of death at 28 days, during hospitalization, and at 90 and 180 days after admission. Importantly, this relationship was linear, meaning that every incremental increase in the score was associated with progressively higher mortality risk.
While lnEASIX performed similarly to the established SOFA score (a common ICU severity measure) on its own, the real benefit came from combining the two. Adding lnEASIX to SOFA meaningfully improved the ability to correctly classify patients by risk, as shown by significant improvements in reclassification and discrimination statistics. This matters because EASIX requires no additional testing — it is derived entirely from standard admission labs, making it a 'zero-cost' tool that can be immediately applied in any clinical setting.
This research suggests that monitoring endothelial (blood vessel lining) dysfunction through EASIX could help clinicians identify high-risk stroke patients earlier and more accurately. The findings were broadly consistent across different patient subgroups, though patients with pre-existing cancer appeared to respond differently, which warrants further investigation. Overall, this study supports the potential use of lnEASIX as an easy-to-calculate complement to existing ICU scoring systems for stroke prognosis.
Bai W, Tao M, Li S, Shi W, Jiang K, Tian L. (2026). Prognostic Value of the Endothelial Activation and Stress Index (EASIX) in Critically Ill Patients With Ischemic Stroke: A MIMIC-IV Database Analysis.. Brain and behavior. https://doi.org/10.1002/brb3.71708