Background: Intracranial procedures (including craniotomy/ craniectomy, external ventricular drain [EVD], ventriculoperitoneal shunt [VPS], stereotactic craniotomy, and administration of intraventricular alteplase [tPA]) are performed in severe cases of intracerebral hemorrhage (ICH). However, severity-adjusted national estimates of complications after such procedures are not known.
Methods: The Nationwide Readmissions Database contains >14 million discharges for all payers and uninsured in 2013. International Classification of Disease, Ninth Revision, Clinical Modification codes were used to identify index cases of ICH, intracranial procedures, and comorbidities. We summarized demographics, comorbidities, and Charlson Comorbidity Index (CCI) during the index admission. Preventable causes of admission were determined by Prevention Quality Indicators. Multivariable Poisson regression models estimated associations between intracranial procedures and readmission rates up to 1 year. Model 1 adjusted for age, discharge status, and CCI; Model 2 adjusted for All Patients Refined Diagnosis Related Groups severity measures.
Results: Among 26,160 index admissions for ICH with intracranial procedures, mean age (SD) was 68.6 (14.9) years; 48% were female; 28% diabetic; 70% hypertensive; 23% smokers, mean CCI (SD) was 3.5 (2.2). 25% were discharged home; 24% died during index admission; 30.7% had readmission during follow-up, 13.2% of which were preventable. Top reasons for 30-day readmission were acute cerebrovascular disease (21%), septicemia (8%), and renal failure (4%). In unadjusted models, craniotomy, EVD, and VPS were associated with higher readmission rates. In adjusted models, higher readmission rates were seen only after craniotomy (Model 1 rate ratio [RR] 1.13, 95% CI 1.08-1.19; Model 2 RR 1.10, 1.04-1.15). tPA and stereotactic craniotomy showed no significant associations.
Conclusions: This analysis of nationally representative US data of ICH admissions showed a 10-13% increased rate of readmissions after craniotomy, most of which were non-preventable. Further research would clarify causal relationships and suggest ways to reduce readmission risk after this potentially life-saving procedure for ICH.