Predicting the Likelihood of Neurosurgical Intervention Prior to Transfer of Spontaneous Intracerebral Hemorrhage (ICH) to Tertiary Care Facilities Using Data From a Retrospective Cohort: The Likelihood of Neurosurgery Score (LoNS)

Abstract

Background

Patients with spontaneous intracranial hemorrhage (ICH) who present to non-tertiary care centers are often transferred to a facility with neurosurgical expertise without consideration of the likelihood of surgical intervention. At our Comprehensive Stroke Center (CSC), a minority of patients transferred for neurosurgical evaluation undergo interventions putting a strain on hospital resources. This study aimed to (1) quantify the frequency of neurosurgical intervention in ICH patients transferred to our hospital and, (2) to develop a tool to aid in transfer decisions.

Methods

Using an IRB-approved retrospective cohort study design, we identified all spontaneous ICH patients transferred to our CSC between January 1, 2016, and May 31, 2023. All patients were reviewed to ensure a primary diagnosis of non-traumatic supratentorial ICH. Odds ratios were calculated using a logistic regression model to identify factors predictive of neurosurgery which were weighted by strength of association. Internal validation was then performed.

Results

Of the 496 participants included in the final dataset, 78 (15.7%) underwent neurosurgical intervention. Age, Glasgow Coma Scale, ICH volume, and intraventricular extension were the greatest predictors of neurosurgery. These factors were used to create the Likelihood of Neurosurgery Score (LoNS), a weighted score used to inform transfer decisions. The score performed well on calibration and discrimination tests.

Conclusion

The LoNS is a new tool to identify ICH patients unlikely to be neurosurgical candidates who could be safely managed at the local level rather than urgently transferred to a tertiary care center. Prospective validation is needed.