Sequential Pattern Mining to Predict Medical In-Hospital Mortality from Administrative Data: Application to Acute Coronary Syndrome

Table 7

External validation: AURC and Brier Score by context according to the type of trajectory.

Context	DRG trajectories		ICD-10 code trajectories
	AURC	Brier Score	AURC	Brier Score
Man and >65 years and ≤5 stays	0.87	0.11	0.82	0.15
Woman and 45–65 years	0.65	0.12	0.65	0.11
>65 years and ≤5 stays	0.90	0.09	0.88	0.09
Man and 45–65 years and ≤5 stays	0.99	0.02	0.81	0.11
45–65 years and ≤5 stays	0.96	0.05	0.91	0.14
Woman and >65 years and >5 stays	0.77	0.19	0.74	0.19
Woman and >65 years	0.75	0.14	0.81	0.16
Man and ≤5 stays	0.95	0.08	0.89	0.14
≤5 stays	0.94	0.05	0.87	0.16
Woman and >5 stays	0.67	0.21	0.79	0.20
Woman	0.83	0.13	0.82	0.16
Man and 45–65 years and >5 stays	0.80	0.18	0.65	0.24
45–65 years and >5 stays	0.78	0.18	0.57	0.26
Man and >65 years and >5 stays	0.68	0.23	0.75	0.19
Man and 45–65 years	0.80	0.14	0.76	0.24
Man and >65 years	0.76	0.14	0.80	0.18
45–65 years	0.74	0.22	0.81	0.17
>65 years and >5 stays	0.70	0.17	0.70	0.24
>65 years	0.79	0.15	0.72	0.17
Man and >5 stays	0.77	0.20	0.74	0.21
Man	0.76	0.24	0.85	0.17
>5 stays	0.73	0.21	0.80	0.19
General	0.82	0.16	0.79	0.17

Best results are in bold.