Being Safe with SDH – The SafeSDH Tool
June 18, 2024
Written by Megan Hilbert
Spoon Feed
The SafeSDH prediction tool can help to identify patients with subdural hematoma (SDH) who may not necessitate a higher level of care (either ICU stay or interfacility transfer) and can safely be monitored in the community or lower acuity setting.
Transfers are really a pain in the neck head
This was a retrospective chart review of patients who presented with an isolated SDH to 1 of 6 hospitals (predominantly academic). Primary outcome was a composite of neurologic deterioration, neurosurgical intervention, discharge on hospice, and death. The goal was external validation of the SafeSDH tool that was originally published in 2017. Importantly, patients were excluded and not considered low risk if they were on anticoagulant or non-aspirin antiplatelet agent, had a Glasgow Coma Score (GCS) < 14, more than 1 discrete hematoma, hematoma thickness > 5mm, or midline shift. In this validation cohort, the tool demonstrated sensitivity of 99% (95%CI 96-100) and specificity of 31% (95%CI 27-35) with a negative likelihood ratio of 0.03 (95%CI 0.01-0.11). This compares to a sensitivity of 96% and specificity of 29% on internal validation cohort with initial publishing of the tool. There were only two false negatives both of whom had decompensation from a medical standpoint and not the SDH itself.
Limitations exist including the retrospective nature of the data. Also, while this tool could be helpful in determining the level of care a patient needs when being admitted at an academic center, where I think it truly would be worth its weight in gold would be in the consideration of transfer from a community site to an academic center for in-person neurosurgical consultation. There was only one community site included in this validation study, currently limiting its generalizability to that setting.
How will this change my practice
I currently practice at both an academic and a community site. While I like the idea of a validated tool, I’m uncomfortable not calling my neurosurgical colleagues to discuss SDH patients. Not only that, I anticipate a good deal of push back from my inpatient colleagues who may be less than thrilled to keep a patient in the community setting, even if this tool is able to deem them low risk for decompensation.
Another Spoonful
More than half of potentially eligible patients were excluded due to missing data, which creates the potential for significant selection bias. I don’t think the data is quite there to support anything other than transport to a facility with on-site neurosurgical capabilities. ~Millie Cossé
Source
External Validation of a Tool to Identify Low-Risk Patients With Isolated Subdural Hematoma and Preserved Consciousness. Ann Emerg Med. 2024 May;83(5):421-431. doi: 10.1016/j.annemergmed.2023.08.481. Epub 2023 Sep 19. PMID: 37725019.