Very much this. That number in isolation isn’t especially useful. Percentage of injuries of scooterfolk vs bikefolk, pedestrianfolk, or motorcyclefolk would be a good first metric IMHO, and I bet will show that as previously discussed, helmets are probably a good idea.