Table 1. Base-case estimates used within the model.

Parameter	Base-case estimate*	Reference
Male gender, %	75.9	Study database
Mean age ± SD	39.2 ± 9.5	Study database
Prior DM, %	3.8	Study database
Prior CHD, %	1.2	Study database
CHD risk, %		[17]
No prior DM/CHD	0.17
Prior DM	0.82
Prior CHD	3.75
Prior DM/CHD	4.94
Effect of treatment on TC:HDL ratio‡		[18]
LPV/r	–0.17
ATV+r (1 & 2)	–0.40
Risk of CHD being fatal, %	35.4	[19]
Effect of treatment on transition to HS with VL ≥50 copies/mL		[12]
LPV/r	Not Applicable
ATV+r 1	Not Applicable
ATV+r 2†	–19%
AIDS risk (%)	0.09 to 94.35 (states 1 to 8)	Study database
CKD risk (%)		[13]
LPV/r	0.12 to 2.9 (cycle 1 to 40)
ATV+r (1 & 2)	0.12 to 23.3 (cycle 1 to 40)
OI risk (%)	1.76	Study database
Diarrhoea risk (%)		Study database
LPV/r	1.76
ATV+r (1 & 2)	0.0
Hyperbilirubinemia risk (%)		Study database
LPV/r	0.0
ATV+r (1 & 2)	1.47

SD: standard deviation; DM: diabetes mellitus; CHD: Coronary heart disease; TC: Total cholesterol; HDL: High-density lipoprotein; VL: Viral load; HS: Health state; OI: Opportunistic infection; CKD: Chronic kidney disease

^*Estimates gathered from the literature and expressed as incidence rates (i.e., event/person-years) were converted into semestral probabilities using standard formulas

^†Transition probabilities of transitioning to a state with greater viral load (≥ 50 copies/mL) are 19% lesser than LPV/r

^‡Effect on TC:HDL ratio is considered null in second-line treatment