Table 4.

a: Summary of the “moxifloxacin models” and associated odds ratios in the validation sets.

	Learning			Validation
	OR	p	95% CI	OR	p	95% CI
RR(1 msec. inc.)	1.00	0.49	1.00-1.00	1.00	0.067	1.01-1.11
QTcF(lead ev1)(1 msec. inc.)	1.17	<0.0001	1.11-1.23	1.14	<0.0001	1.09-1.18
ERD30%(1 msec. inc.)	1.12	0.032	1.09-1.30	1.15	<0.0001	1.04-1.26

b: Summary of the “LQTS models” and associated odds ratios in the validation sets.
	Learning			Validation
	OR	p	95% CI	OR	p	95% CI
RR(1 msec. inc.)	2.04	0.006	1.23-3.38	3.44	<0.0001	1.97-6.00
QTcB(lead ev1)(1 msec. inc.)	2.79	<0.0001	1.66-4.66	1.56	0.031	1.04-2.34
αL(lead ev1)(1.5μV/ms inc.)	0.46	0.002	0.28-0.75	0.38	0.0002	0.23-0.64

c: Summary of the Cox model for cardiac events in LQTS type 2.
	HR	p	95% CI
RR(1 msec. inc.)	1.00	0.097	1.000-1.011
QTcB(lead ev1)(1 msec. inc.)	1.01	0.045	1.000-1.011
T roundness(10% inc.)	1.15	0.036	1.009-1.310
Female(age ≤17)	1.15	0.74	0.486-2.735
Female(age >17)	4.24	0.0003	1.929-9.320

QTc are associated with different heart rate correction- F: Fridericia and B: Bazett Each parameter is associated with information about its associated lead.

^**the parameters in the LQTS group were normalized using the standard deviation of the non-carrier population. Thus, each point estimate is associated with one increase of standard deviation of these values in non-carrier LQTS patients.