Transformation into chronic subdural hematoma following acute subdural hematoma: Incidence, outcomes, and predictors

Acute subdural hematoma (aSDH) and chronic subdural hematoma (cSDH) have traditionally been considered distinct clinical entities. However, a subset of patients with aSDH may subsequently develop cSDH, representing a poorly characterized clinical phenomenon with unclear incidence, outcomes, and risk factors.

To quantify the incidence and outcomes of cSDH-related readmissions following aSDH, identify independent predictors of transformation, and develop and validate a prediction score for risk stratification.

Retrospective cohort study SETTING: Nationwide Readmissions Database (2016-2022).

Adult patients (aged ≥18 years) hospitalized with primary aSDH were included. Exclusions included pre-existing subacute or chronic SDH, in-hospital mortality, prolonged hospitalization (≥14 days), comfort care status, middle meningeal artery embolization during index hospitalization, or insufficient follow-up (<30 days).

The primary outcome was cSDH-related readmission within 180 days. Secondary outcomes included functional decline, mortality, surgical intervention, and hospitalization costs. Predictors were identified using least absolute shrinkage and selection operator (LASSO) regularization in a Cox proportional hazards framework, and the Post-acute Evolution Risk Score to Identify Subdural Transformation (PERSIST) was developed and validated.

Of 189,905 patients meeting inclusion criteria (median age 73 years; 42% female; 95.4% had documented head trauma; 82.5% underwent open craniotomy), 4836 (2.5%) experienced cSDH-related readmissions, representing a cumulative incidence of 2.8% at 180 days. Among patients readmitted with cSDH, 26.0% had functional decline, 33.6% required surgical evacuation, 4.3% died, and the median readmission hospitalization cost was $71,474. In a dedicated training cohort (n = 113,942), LASSO identified 21 independent predictors, including absence of concomitant traumatic pathology, liver disease, male sex, and coagulopathy-associated conditions. In validation (n = 37,983), PERSIST demonstrated adequate discrimination (C-index 0.69 [95%CI 0.67-0.70]) and excellent calibration (integrated calibration index 0.4%). The highest-risk decile (PERSIST ≥48) exhibited 180-day transformation incidence of 7.1%, 10-fold higher than the lowest-risk decile (0.7%) and 1.5-fold higher than the second-highest decile (4.7%).

In this nationally representative cohort, cSDH-related readmissions occurred in 2.8% of aSDH patients within 180 days and were associated with substantial morbidity. The PERSIST score demonstrated adequate discrimination and excellent calibration for identifying high-risk patients who may benefit from enhanced surveillance or preventive interventions.