Admission rSIG Tracks Early Hematoma Expansion Risk in Spontaneous Intracerebral Hemorrhage
A low reverse shock index multiplied by Glasgow Coma Scale score at admission was linked to a higher likelihood of early hematoma expansion in spontaneous intracerebral hemorrhage, adding a simple bedside marker to the earliest phase of hemorrhagic stroke care. In a retrospective single-center cohort spanning nearly a decade, the score separated patients who later demonstrated radiographic growth from those who did not, even after adjustment for other clinical factors.
The finding is clinically relevant because hematoma expansion is one of the most damaging early events in intracerebral bleeding. Once the clot enlarges, intracranial pressure rises, perihematomal tissue injury worsens, and neurological decline can accelerate before any delayed intervention has a chance to help. A score built from three first-contact variables — Glasgow Coma Scale, systolic blood pressure, and heart rate — is attractive precisely because it can be calculated at the bedside without waiting for advanced testing.
A Large Cohort With a Clear Early Hemorrhage Signal
Investigators reviewed consecutive adults admitted with spontaneous intracerebral hemorrhage between July 2016 and June 2026. Of 2,450 screened patients, 2,035 met inclusion criteria, and 156 patients, or 7.7%, developed early hematoma expansion. Expansion was defined rigorously as an absolute hematoma-volume increase greater than 6 mL or a relative increase greater than 33% on follow-up non-contrast CT within 8 hours.
That endpoint matters because it captures the vulnerable early window when the clot is still biologically active and the injury trajectory can still worsen quickly. Follow-up imaging within hours, rather than days, aligns with the natural history of hemorrhage growth and makes the analysis more directly applicable to emergency and stroke-unit decision-making.
Each 5-point decrease in rSIG was independently associated with higher odds of early hematoma expansion in the fully adjusted model, with an adjusted odds ratio of 1.783 and a 95% confidence interval of 1.560 to 2.038. The p value was reported as significant, reinforcing the robustness of the association across multivariable modeling.
Why the Score Makes Physiologic Sense at the Bedside
rSIG combines markers that reflect neurologic status and hemodynamic reserve. Glasgow Coma Scale captures the severity of acute cerebral dysfunction, while the reverse shock index incorporates systolic blood pressure and heart rate, two variables that move in response to stress, compensation, and cardiovascular instability. In hemorrhagic stroke, a low rSIG may therefore identify patients in whom the bleed is already exerting a stronger systemic and neurologic effect, or in whom the physiology is less able to buffer further enlargement.
The appeal of such a composite tool is practical. A low score does not explain the bleeding mechanism by itself, but it can flag a patient whose initial presentation suggests a more dangerous trajectory. In a condition where the first few hours can determine whether a patient stabilizes or deteriorates, any validated marker that helps separate lower-risk from higher-risk presentations deserves attention.
Modeling Strategy Strengthened the Association
The analysis did not rely on a single statistical approach. Associations were examined using hierarchical logistic regression, restricted cubic splines, quartile analysis, bootstrap resampling, and a 1:2 matched sensitivity analysis. That combination of methods is important because it tests whether the signal persists across different analytic frameworks rather than depending on one narrow specification.
Restricted cubic splines are particularly useful when the relationship between a biomarker and outcome may not be strictly linear. In acute neurologic injury, physiologic markers often behave nonlinearly: a modest drop in a composite score may have little consequence until a threshold is crossed, after which risk rises faster. The use of spline analysis suggests the investigators were looking for exactly that kind of pattern, not assuming a straight-line relationship from the start.
The cohort itself was substantial for a single-center study, and the follow-up imaging endpoint was clinically direct. Even so, the design remains retrospective, and bedside scores can only complement, not replace, imaging, neurological examination, and established hemorrhage-risk markers. A prognostic association does not automatically translate into a treatment target.
What This Means for Emergency Stroke Triage
In real-world stroke care, the earliest question is often not whether hematoma expansion can be proven retrospectively, but which patient needs the most urgent monitoring, repeat imaging, and escalation of care. A simple score derived from the first set of vital signs and neurologic assessment fits that workflow better than complex biomarkers that are not immediately available.
The present data suggest that lower admission rSIG identifies a subgroup at higher risk for radiographic growth within hours. If validated externally, the score could support triage decisions in emergency departments, stroke units, and intensive care settings where early deterioration can be difficult to predict from blood pressure or consciousness alone. Because the components are already routinely collected, the barrier to implementation would be low.
A bedside index is most valuable when it changes behavior. In this setting, a low rSIG could prompt faster repeat CT, closer neurologic surveillance, and earlier discussion of prognosis with families, especially when the initial scan shows a bleed pattern already concerning for expansion.
Limits of a Single-Center Retrospective Study
Several constraints remain. The work came from one center, and retrospective single-center data can reflect local practice patterns, imaging timing, and admission thresholds. The cohort also depended on first-admission measurements, which may vary with prehospital care, ED resuscitation, or the timing of the initial vital signs relative to symptom onset.
The study established association, not causation. A low rSIG may be a marker of early severity rather than a modifiable driver of expansion. External validation will be essential before the score is used as anything more than a risk stratification aid. Different populations, imaging protocols, and stroke systems of care may shift its operating characteristics.
Still, the signal is strong enough to warrant attention. In spontaneous intracerebral hemorrhage, where early enlargement can quickly narrow treatment options, a readily available admission score that independently tracks expansion risk deserves a place in the discussion at the bedside and in future prospective validation studies.