Temporal Increase in Strict Spontaneous Intracerebral Hemorrhage Admissions During the First March Following Direct Israel–Iran Hostilities: Preliminary Single-Center Findings from a Decade-Referenced Neuroscience Services Cohort

Objective: On 28 February 2026, Israel entered direct large-scale hostilities with Iran under Operation Roaring Lion. The opening phase was characterized by repeated missile alerts, civilian protected-space instructions, and rapid reorganization of hospital activity into protected areas. We observed an apparent increase in strict spontaneous intracerebral hemorrhage admissions during March 2026 within our linked neurology/neurosurgery services dataset. The aim of this preliminary single-center study was to determine whether March 2026 was temporally associated with a higher proportional burden of strict spontaneous intracerebral hemorrhage admissions compared with March cohorts from the preceding decade and whether this pattern was also observed for acute ischemic stroke or non-traumatic subarachnoid hemorrhage. Methods: We performed a retrospective observational cohort study of all unique March admissions captured within a linked neurology/neurosurgery services dataset from 2016 through 2026. Hospitalizations were deduplicated by admission number. March 2026 was treated as the first full March occurring after the onset of direct Israel–Iran hostilities on 28 February 2026. Strict spontaneous ICH was defined using diagnosis-text phenotyping that included intraparenchymal or intracerebral hemorrhage terminology while excluding trauma, subarachnoid hemorrhage, subdural hematoma, aneurysm, arteriovenous malformation, tumor-related hemorrhage, cavernoma, venous sinus thrombosis, dissection, and other clearly secondary etiologies. Comparator phenotypes included acute ischemic stroke and non-traumatic subarachnoid hemorrhage (SAH). Results: Across 3855 unique March admissions, 68 met criteria for strict spontaneous ICH. In March 2026, 9 of 223 admissions (4.0%) were classified as strict spontaneous ICH, compared with 59 of 3632 admissions (1.6%) across March 2016–2025, yielding a rate ratio of 2.48 (95% CI 1.25–4.94; p = 0.015). Patients with strict spontaneous ICH in March 2026 were older (mean age 72.3 vs. 65.8 years), and 7 of 9 cases (77.8%) occurred in patients aged ≥70 years compared with 25 of 59 (42.4%) historically (p = 0.073). Acute ischemic stroke did not increase in March 2026 (7.6% vs. 9.4%; p = 0.475), and non-traumatic SAH showed only a non-significant numerical increase (2.7% vs. 1.4%; p = 0.147). Sensitivity analyses showed a directionally consistent but statistically non-significant increase when March 2026 was compared with March 2025 alone (4.0% vs. 1.2%; rate ratio 3.36, 95% CI 0.92–12.27; p = 0.076) and with a rolling 3-year March baseline from 2023 through 2025 (4.0% vs. 2.1%; rate ratio 1.93, 95% CI 0.88–4.23; p = 0.143). In-hospital mortality among strict spontaneous ICH patients was 1 of 9 (11.1%) in March 2026 versus 4 of 59 (6.8%) in March 2016–2025. Conclusions: In this preliminary single-center neurology/neurosurgery services cohort, March 2026 showed a higher proportional burden of strict spontaneous intracerebral hemorrhage admissions than March cohorts from the preceding decade, while acute ischemic stroke did not increase. Sensitivity analyses using March 2025 alone and a rolling 3-year March baseline were directionally consistent but did not reach statistical significance. These findings should therefore be interpreted as a hypothesis-generating temporal association rather than evidence of causality or population-level incidence. Wartime-related psychological stress, sleep disruption, altered healthcare access, blood pressure dysregulation, and medication nonadherence are biologically plausible contributors, but individual-level blood pressure, medication exposure, body mass index, time-to-admission, direct stress exposure, and detailed outcome data were not available in the present dataset. Multicenter, hospital-wide, and registry-based validation with seasonal and systems-level sensitivity analyses is required.