Design of a Clinical Decision Support System for Fracture Prediction Using Imbalanced Dataset

Table 6

Association between prescribed inhaled corticosteroid and fracture.

	Fracture		value
	No ()	Yes ()
Medication (inhaled corticosteroids)	10511 (90.3%)	1134 (9.7%)
Cumulative dose (mg)			<0.001
Mean ± SD	171.97 ± 318.13	230.14 ± 442.76
Median	45	54
Mean follow-up daily dose (mg/d) (cumulative dose/follow-up duration)			<0.001
Mean ± SD	1.00 ± 3.91	0.44 ± 2.31
Median	0.07	0.05
Mean exposed daily dose (mg/d)			0.261
(Cumulative dose/cumulative exposed duration)
Mean ± SD	0.55 ± 1.35	0.51 ± 0.81
Median	0.42	0.39
Duration (day), median (range)
Follow-up duration	1370.91 ± 1097.76	1777.17 ± 1039.82	<0.001
Median	1239.0	1862.0
Exposed duration	342.35 ± 503.32	453.79 ± 584.65	<0.001
Median	121.0	199.0