Figure 1.
Summary of structure of mathematical model.
A shows the basic model unit with parasite life cycle stages in the human host, antimalarial drug action and immunity. B shows the unit in A repeated three times to track parasites resistant to artemisinins and ACT partner drug. C shows multiple repetitions of B to reproduce the various strategies used in the trial. In A, ‘blood stage’ refers to individuals with asexual stage parasites in the peripheral blood but no gametocytes and ‘infectious stage’ is individuals with gametocytes.
Figure 2.
Model fits to data and validation.
Model predictions are shown as red lines and surveillance data in black (subgroups) or blue (summary). A Model validated with to surveillance data for Kampot Operational District (OD) (2004–2010). The malaria control strategies used are shown. B–E Model fitted to data from the field study (2004–2007). B–C Reductions in P. falciparum asexual stage parasite (B) and gametocyte (C) prevalences in 17 villages in Kampong Speu OD. The strategy used was treatment with ACT plus single dose adjunctive primaquine for three years and a single MDA with ACT and multiple rounds of primaquine MDA. D Reduction in P. falciparum asexual parasite prevalence in 3 villages in Kampot OD with MDA and treatment as above, although with lower coverage, followed by a second MDA of ACT plus single dose adjunctive primaquine with higher coverage (dotted line) at 42 days. E Reduction in P. falciparum asexual parasite prevalence in 4 villages in Kampong Speu OD with the same strategy as D but with the second MDA (dotted line) at 1 year.
Figure 3.
Contribution of each component of the strategies employed on P. falciparum overall parasite prevalence.
Each panel shows the additional effects of adding primaquine to MDA with ACT. A MDA alone and B MDA combined with simultaneous introduction of ACT plus or minus single dose adjunctive primaquine for treatment. Black line is no treatment, blue lines are no treatment (A) or treatment with ACT (B), red/pink lines are treatment with ACT plus primaquine.
Figure 4.
Predicted numbers of symptomatic cases and proportion with parasites in the blood before, during and after the trial. The red line is the percent of the population with blood stage parasites and the black line is the number of symptomatic cases. The * indicates the paradoxical increase in clinical cases despite a decrease in the proportion affected.
Figure 5.
Artemisinin resistance and different elimination strategies.
Effect of an increasing prevalence of resistance (defined as the proportion of infections which are artemisinin resistant) from 0% to 80% on the success of MDA with ACT plus multiple rounds of primaquine and treatment with A ACT or B ACT plus single dose primaquine from 2004–2007. C–D effects of different treatment and MDA regimes on the spread of artemisinin resistance (C: prevalence of resistant infections, D: number of resistant infections), presuming a starting prevalence of resistance of 5%. Interventions ceased in 2007.
Figure 6.
Effect of varying the timing of interventions.
A Varying the season in which interventions are implemented. MDA with ACT+multiple primaquine and Rx with ACT+single primaquine started together at different times during the malaria season: when parasite prevalence is falling, at trough, rising or at peak. B Simultaneous versus sequential interventions and the additional effect of bed nets. Simultaneous introduction of treatment with ACT+single primaquine plus MDA with ACT+multiple primaquine (red) versus optimally timed sequential interventions (black): a. treatment with ACT+single primaquine then b. MDA with ACT+primaquine at 9 weeks before the nadir and c. another MDA with multiple primaquine at the nadir and the same simultaneous interventions with the addition of long lasting insecticide treated bed nets with 30% coverage (blue).
Table 1.
Main policy implications of modelling results.