In the PECARN study, they excluded 969 patients with a GCS score of < 14.
They excluded these patients on the basis that the risk of TBI on a CT scan is > 20% and this risk outweighs the radiation risk from CT, and therefore radiation risk from CT is not controversial.
We should most likely do the same.
Note: this possibly assumes clinical staff would order a CT scan generally when for a GCS < 14, and that these patients wouldn't be assessed with the tool we build.
I'm no longer convinced that we should drop GCS columns. This really reduces the number of cases where a TBI was found, and further imbalances the data.
In the PECARN study, they excluded 969 patients with a GCS score of < 14.
They excluded these patients on the basis that the risk of TBI on a CT scan is > 20% and this risk outweighs the radiation risk from CT, and therefore radiation risk from CT is not controversial.
We should most likely do the same.
Note: this possibly assumes clinical staff would order a CT scan generally when for a GCS < 14, and that these patients wouldn't be assessed with the tool we build.