Document resource
Importance Readmission after chronic or subacute subdural hematoma (cSDH/sSDH) hospitalization is common, yet clinical attention and trial design have focused primarily on surgical recurrence. The full spectrum of readmission events, their clinical impact, and the heterogeneity of the affected patient population remain poorly characterized.Objective To characterize the incidence, diversity, and outcomes of 90-day readmissions after cSDH/sSDH hospitalization and to identify patient phenotypes with distinct readmission risk profiles using machine learning-driven clustering.Design, Setting, and Participants Retrospective cohort study using the Nationwide Readmissions Database, 2016-2022. Adults with non-elective primary cSDH/sSDH admissions were included. Patients who died during index admission, had do-not-resuscitate orders, or underwent middle meningeal artery embolization were excluded.Main Outcomes and Measures First readmission within 90 days classified into 7 mutually exclusive categories. Readmission outcomes included length of stay, cost, in-hospital mortality,functional decline, new disability, and inability to return home. Machine learning-based phenotyping using Shapley Additive Explanations (SHAP) values from a multinomial gradient-boosted model and K-means clustering identified patient subgroups with divergent readmission patterns.Results Of 22,387 patients (mean [SD] age, 70.8 [13.0] years; 29.6% female), 6,497 (29.0%) were readmitted within 90 days across diverse causes. Surgical SDH recurrence accounted for only 22.5% of readmissions, and fewer than half (44.0%) were directly SDH-related. Non-SDH readmissions carried substantial clinical impact: infection readmissions had the highest mortality(9.6%), exceeding surgical SDH (2.9%) more than 3-fold, and the highest rate of new disability(45.5%) among patients initially discharged with routine self-care. Machine learning-driven phenotyping analysis identified 5 patient clusters with unique clinical characteristics and diverging readmission patterns: low acuity (39.8%), elderly/frail (16.6%), atrial fibrillation(16.9%), young/healthy (14.5%), and high acuity (12.1%). These phenotypes revealed marked patient heterogeneity, with each cluster exhibiting distinct readmission risk profiles that were not apparent from aggregate analyses.Conclusions and Relevance Readmissions after cSDH/sSDH are diverse and predominantly non-surgical, with non-SDH readmissions carrying equal or worse outcomes than surgical recurrence. Machine learning-based phenotyping uncovered substantial patient heterogeneity, highlighting the need for new therapeutic strategies, expanded clinical trial outcome targets beyond surgical recurrence, and comprehensive post-discharge care models tailored to distinct patient subgroups.Disclosures M. Colasurdo: None. D. Lakhani: None. A. Malhotra: None.Abstract E-077 Figure 1