Table 1.
Model assumptions for dynamics of therapeutic efficacy and decay.
Fig 1.
Illustration of seasonality profiles, intervention timing, and primary outcome measures.
Panel A illustrates the two seasonal profiles modelled, where 70% of malaria cases occur within a four- or six-month period and where the y-axis represents the average annual entomological inoculation rate (EIR). Panel B illustrates interventions deployed according to four seasonal deployment scenarios. Scenario 1. Perfect deployment: All children aged between three and 59 months had perfect compliance to five yearly SMC cycles. Scenario 2. Imperfect seasonal coverage: All children aged three to 59 months had perfect compliance to four yearly SMC cycles but missed the final cycle. Scenario 3. Imperfect deployment: All children aged three to 59 months had perfect compliance to yearly SMC cycles one, three, and five. Scenario 4. Random allocation: 30% to 100% of children were randomly selected to receive each of five yearly SMC cycles. Panel C illustrates the evaluation of our primary outcome measures. We tracked a cohort of children born in the first year of intervention deployment, and evaluated cumulative incidence when children were five and ten years old. We then calculated the reductions in the cumulative incidence of uncomplicated and severe malaria attributed to the combination of SMC and the novel therapeutic. Reductions were evaluated relative to two counterfactual scenarios: regular deployment of SMC alone, and seasonal deployment of a novel therapeutic alone. Shaded regions indicate the minimum and maximum cumulative cases per person observed across ten stochastic replicates of simulations from the individual-based malaria transmission model.
Fig 2.
Predicted cumulative case curves when seasonal deployment of two exemplar pre-liver stage therapeutics are combined with SMC.
The median number of cumulative cases per year of age were calculated when children aged three to 59 months received one of the following interventions: no intervention, three SMC cycles (imperfect deployment scenario), seasonal deployment of a pre-liver stage therapeutic, or the combination of three SMC cycles and the pre-liver stage therapeutic. Top panels include deployment of a pre-liver stage therapeutic with a protection half-life of 354 days, 90% initial efficacy, and decay shape parameter of 1. Bottom panels include a pre-liver stage therapeutic with a protection half-life of 120 days, 50% initial efficacy, and decay shape parameter of 1. The pink shaded regions indicate age-eligibility for SMC and the novel therapeutic. Shaded regions indicate the minimum and maximum cumulative cases per person observed across stochastic replicates of simulations from the individual-based malaria transmission model. Model results are shown for a scenario with transmission intensity corresponding to 32% PfPR2-10, where 75% of malaria cases occur within six months of the year and the probability of seeking first-line treatment for clinical malaria over 14 days is low (10%).
Fig 3.
Gaussian Process regression emulator predictions for the relationship between pre-liver stage therapeutic properties and reductions in cumulative uncomplicated and severe cases relative to SMC alone.
Results show the imperfect deployment scenario, where children aged three to 59 months received three SMC cycles. Transmission is high (32% PfPR2-10), 75% of malaria cases occur within six months of the year, and the probability of seeking first-line treatment for clinical malaria over 14 days is low (10%). Each panel shows the median reduction in cumulative uncomplicated (solid lines) or severe cases (dashed lines) achieved by combining seasonal deployment of a pre-liver stage therapeutic with SMC, relative to cumulative cases when SMC is deployed alone. Median reductions are calculated by dividing the parameter range for the therapeutic property shown on the x-axis (protection half-life, initial efficacy, decay shape) into 51 segments, and calculating the median outcome for each given segment across all other parameter values. Shaded regions represent the 25th and 75th percentiles of the corresponding reductions. Predictions are shown separately for cumulative case outcomes at five (top panels) and ten years old (bottom panels). SMC = seasonal malaria chemoprevention.
Fig 4.
Predicted relationships between pre-liver stage therapeutic properties and reductions in cumulative severe cases relative to SMC alone, compared across SMC deployment scenarios.
Grid squares indicate the predicted reduction if a pre-liver stage therapeutic with the given protection half-life (x-axis) and initial efficacy (y-axis) were deployed in addition to SMC, assuming a long tail of protection (decay shape parameter 0.7). Reductions are rounded to the nearest 5%. Results show a scenario where transmission is high (32% PfPR2-10), 75% of malaria cases occur within six months of the year, and the probability of seeking first-line treatment for clinical malaria over 14 days is low (10%). Panels depict predicted relationships for four different SMC deployment scenarios. In deployment scenario 4, SMC coverage is fixed at 70%. SMC = seasonal malaria chemoprevention.
Fig 5.
Gaussian Process regression emulator predictions for the relationship between blood stage therapeutic properties and expected reductions in cumulative uncomplicated and severe cases relative to SMC alone.
Results show the imperfect deployment scenario, where children aged three to 59 months received three SMC cycles. Transmission is high (32% PfPR2-10), 75% of malaria cases occur within six months of the year, and the probability of seeking first-line treatment for clinical malaria over 14 days is low (10%). Each panel shows the median reduction in cumulative uncomplicated (solid lines) or severe cases (dashed lines) achieved by combining seasonal deployment of a blood stage therapeutic with SMC, relative to cumulative cases when SMC is deployed alone. Median reductions are calculated by dividing the parameter range for the therapeutic property shown on the x-axis (protection half-life, initial efficacy, decay shape) into 51 segments, and calculating the median outcome for each given segment across all other parameter values. Shaded regions represent the 25th and 75th percentiles of the corresponding reductions. Emulator predictions are shown separately for cumulative case outcomes at five (panels A) and ten years old (panels B). SMC = seasonal malaria chemoprevention.
Fig 6.
Impact of combining a therapeutic with both pre-liver stage and blood stage activity with three yearly cycles of SMC on cumulative uncomplicated and severe cases, compared with SMC deployed alone.
Panel A shows the drivers of impact on all cumulative case outcomes for a therapeutic with both pre-liver stage and blood stage activity deployed together with three yearly cycles of SMC, compared with SMC deployed alone. Bars show the total Sobol effect indices for intervention model parameters. Total Sobol effect indices can be interpreted as the proportion of variation in the cumulative case outcome that can be attributed to a small change in each model parameter and its interactions with other parameters. Indices are shown across levels of transmission intensity (x-axis) for a scenario where three yearly cycles of SMC are deployed (imperfect coverage scenario) in a setting where 75% of malaria cases occur within six months of the year and the probability of seeking first-line treatment for clinical malaria over 14 days is low (10%). Panel B shows differences in the median impact on cumulative case outcomes for therapeutics with pre-liver stage activity alone, blood stage activity alone, and multi-stage pre-liver stage and blood stage activity. Results are shown for the same scenario as in panel A. Points in each panel show the median expected reduction in cumulative cases achieved by combining seasonal deployment of a pre-liver stage, blood stage, or multi-stage therapeutic with SMC, relative to cumulative cases when SMC is deployed alone. Median reductions are evaluated across all possible combinations of therapeutic initial efficacy, protection half-life, and decay shape. Bars indicate the 25th and 75th percentiles of the corresponding reductions. SMC = seasonal malaria chemoprevention, PfPR2-10 = Annual P. falciparum parasite prevalence in children aged between 2 and 10 years old.