Uncertainty and sensitivity analysis applied to an in-host malaria model with multiple vaccine antigens

Abstract

Results of uncertainty and sensitivity analysis have significant epidemiological importance in malaria control. In this paper, the efficient technique of latin hypercube sampling (LHS) and partial rank correlation coefficient is applied to an in-host Plasmodium falciparum malaria model. Sensitivity indices of the basic reproduction number are derived using the normalised forward approach. By a theoretical analysis, we show the existence and stability of the model steady states based on threshold value R0 . Our results show that the in-host model is sensitive to variations in the efficacy of malaria vaccines. The rates of parasite to cell invasions, the density of merozoites released per bursting infected erythrocyte and the proportions of merozoites that become gametocytes are highly significant in determining the severity of malaria infection. Moreover, a highly effective vaccine combination is critical for malaria disease elimination goal. This study further shows that the long term precise predictions of the concentrations of infected cells during malaria infection would be difficult until these key parameters are correctly determined. These results are vital in the on-going malaria vaccine development.