Table 1.

Input parameters for base-case analysis in the model in Taiwan setting.

Variables	Estimates	Standard error/Range	Distribution	References
Transitional probabilities
Hospitalization for heart failure			Beta	EMPEROR-Reduced trial (9)
Add-on empagliflozin	0.008811915	0.002165244
Standard therapy alone	0.012566527	0.002578038
Cardiovascular death
Add-on empagliflozin	0.006589325	0.00187447
Standard therapy alone	0.007131185	0.001947398
All-cause death
Add-on empagliflozin	0.002113141	0.001063894
Standard therapy alone	0.002177683	0.001078827
Utility score
Stable heart failure	0.770	0.016	Beta	(10, 19)
Decrement for age	−0.0016	0.0001
Hospital for heart failure	−0.321	0.02
Monthly costs (US$)
Monthly costs of empagliflozin (10 mg once per day)	35	17.5	Gamma	NHIRD
Monthly costs of stable heart failure	450	225
Mean costs of hospitalization for heart failure	2,887	1,443.5
Costs before cardiovascular death	3,430	1,715
Costs before all-cause death	3,390	1,695

NHIRD, National Health Insurance Research Database.

The monthly transition probabilities were transformed by the following process.

(1) Probability (obtained from the EMPEROR-Reduced trial) to a rate = [-ln (1-p)] ÷ t.

(2) Rate to a probability (monthly transition probability applied in the analyses) = 1- exp(-rt).

Where r = rate, p = probability, and t = time.

Example: the probability of hospitalization for heart failure is 13.2% over the follow-up period in the EMEPROR-Reduced trial.

16-month probability was transformed to one-month rate and then one-month rate was transformed to one-month probability.

one-month rate = (-ln (1-0.132)/16 = 0.00885.

one-month possibility = 1-exp (−0.00885) = 0.008811915.