Table 5.

AUC values with 95% CI for long-term mortality and poor outcome, for each scale and rater. The results from the comparison with the GCS are also shown

Rater	Senior Consultant			Junior Consultant			Resident			Nurse
	AUC	95% CI	p	AUC	95% CI	p	AUC	95% CI	p	AUC	95% CI	p
6-Month mortality
GCS	0.923a	0.845–0.969	–	0.918a	0.838–0.966	–	0.910a	0.828–0.961	–	0.939a	0.865–0.979	–
SMS	0.852b	0.759–0.919	0.003*	0.857b	0.765–0.923	0.040*	0.846b	0.752–0.915	0.062	0.891b	0.806–0.948	0.097
MGMR	0.826b	0.730–0.900	< 0.001*	0.853b	0.760–0.920	0.051	0.839b	0.745–0.910	0.031*	0.864b	0.773–0.929	0.017*
AVPU	0.833b	0.737–0.904	< 0.001*	0.854b	0.762–0.921	0.017*	0.853b	0.760–0.920	0.047*	0.899b	0.816–0.954	0.147
6-Month mRS 3–6
GCS	0.947a	0.876–0.984	–	0.959a	0.892–0.990	–	0.960a	0.895–0.991	–	0.962a	0.896–0.991	–
SMS	0.849b	0.756–0.917	0.002*	0.885b	0.799–0.944	0.009*	0.891b	0.805–0.948	0.021*	0.893b	0.807–0.949	0.005*
MGMR	0.823b	0.726–0.897	< 0.001*	0.884b	0.797–0.943	0.010*	0.874b	0.785–0.936	0.005*	0.875b	0.786–0.937	0.001*
AVPU	0.848b	0.755–0.916	< 0.001*	0.881b	0.793–0.941	0.001*	0.881b	0.793–0.941	0.002*	0.913a	0.833–0.963	0.022*
6-Month GOSE 1–4
GCS	0.955a	0.888–0.988	–	0.970a	0.908–0.995	–	0.971a	0.910–0.995	–	0.969a	0.907–0.994	–
SMS	0.842b	0.747–0.912	< 0.001*	0.907a	0.825–0.959	0.025*	0.863b	0.796–0.942	0.002*	0.899b	0.815–0.954	0.005*
MGMR	0.815b	0.717–0.891	< 0.001*	0.899b	0.815–0.954	0.020*	0.864b	0.775–0.930	< 0.001*	0.890b	0.805–0.948	0.003*
AVPU	0.843b	0.749–0.913	< 0.001*	0.898b	0.814–0.953	0.004*	0.865b	0.787–0.937	< 0.001*	0.931a	0.855–0.974	0.047*

AUC area under the receiver operating characteristic curve, AVPU alert/verbal/painful/unresponsive scale, CI confidence interval, GCS Glasgow coma scale, GOSE Glasgow outcome scale-extended, mRS modified Rankin scale, MGMR modified GCS motor response, p p values for the comparison with the GCS, SMS simplified motor scale

*Denotes statistically significant difference (p < 0.050)

^aDenotes excellent predictive validity

^bDenotes very good predictive validity

^cDenotes good predictive validity