Time to surgery following chronic subdural hematoma: post hoc analysis of a prospective cohort study

Abstract

Background: Chronic subdural hematoma (CSDH) is a common neurological condition; surgical evacuation is the mainstay of treatment for symptomatic patients. No clear evidence exists regarding the impact of timing of surgery on outcomes. We investigated factors influencing time to surgery and its impact on outcomes of interest.

Methods: Patients with CSDH who underwent burr-hole craniostomy were included. This is a subset of data from a prospective observational study conducted in the UK. Logistic mixed modelling was performed to examine the factors influencing time to surgery. The impact of time to surgery on discharge modified Rankin Scale (mRS), complications, recurrence, length of stay and survival was investigated with multivariable logistic regression analysis.

Results: 656 patients were included. Time to surgery ranged from 0 to 44 days (median 1, IQR 1-3). Older age, more favorable mRS on admission, high preoperative Glasgow Coma Scale score, use of antiplatelet medications, comorbidities and bilateral hematomas were associated with increased time to surgery. Time to surgery showed a significant positive association with length of stay; it was not associated with outcome, complication rate, reoperation rate, or survival on multivariable analysis. There was a trend for patients with time to surgery of ≥7 days to have lower odds of favorable outcome at discharge (p=0.061).

Conclusions: This study provides evidence that time to surgery does not substantially impact on outcomes following CSDH. However, increasing time to surgery is associated with increasing length of stay. These results should not encourage delaying operations for patients when they are clinically indicated.

Keywords: cohort study; outcomes research; real world evidence.

Figure 1

Flow diagram of included patients. CSDH, chronic subdural hematoma.

Figure 2

Distribution of patients by time to surgery, stratified by risk factor with absolute numbers and proportions by time category shown. GCS, Glasgow Coma Scale.

Figure 3

Distribution of patients by time to surgery stratified by neurosurgical center.

Figure 4

Multivariable logistic regression statistical results. Variables are shown with ORs, 95% CIs and p values labeled beneath each variable. Time to surgery is reported relative to an interval of 0 days. Discharge mRS was dichotomized into favorable as 1 (representing mRS of 0–3) or unfavorable of 0 (representing mRS of 4–6). Length of stay was dichotomized based on a median length of stay for the cohort of 7 days. mRS, modified Rankin Scale.