Table 1 Parameter estimates, ranges and sources

Probabilities for clinical parameters

ACT failure rate in a scenario of widespread resistance (P2)

30%

30-80%

A conservative estimate as compared with recent ACT failure rates in Mae Sot, Thailand and those for chloroquine, amodiaquine and sulphadoxine-pyrimethamine in SSA and Asia.

Assumption[10, 14]

ACT failure rates in the absence of widespread resistance (P2)

5%

0-10%

[14]

Treatment failure becomes severe (P3)

2%

0.5-5%

Data from an artesunate efficacy trial in Cambodia and best fit to WHO incidence/mortality data

[15, 24]

Beta distribution

(α = 3 β = 156)

Mortality rate for severe malaria treated with quinine – Asia, EM, LA

22%

Beta distribution

A large multisite in Asia (the trial data were also used to construct the probability distributions for the PSA)

[3]

(α = 164 β = 567)

Mortality rate for severe malaria treated with artesunate – Asia, EM, LA (P5)

15%

Beta distribution

[3]

(α = 107 β = 627)

Mortality rate for severe malaria treated with quinine – SSA (P5)

10.9%

Beta distribution

The largest study of severe malaria treatments in hospitalized patients

[4]

(α = 297 β = 2,416)

Mortality rate for severe malaria treated with artesunate – SSA (P5)

8.5%

Beta distribution

[4]

(α = 230 β = 2,482)

Mortality rate for untreated severe malaria – high transmission (P6)

50%

40-90%

[16, 24]

Beta distribution

(α = 5 β = 5)

Mortality rate for untreated severe malaria – low transmission (P6)

75%

40-90%

[16, 24]

Beta distribution

(α = 7 β = 3)

Access to care

Access to any anti-malarial (P1)

20-100%

Country level data

[24–27]

Access to inpatient care (P4)

40-90%

Determined by GDP per capita

Costs

ACT

$0.8/$1.6

Private sector prices are mostly higher which would imply higher costs for retreatment of failures

[28]

Test

$0.8

$0.5-1.5

[24]

Inpatient care for severe malaria

$65

[18]