Statistical analysis of age-adjusted seroprevalence data using reversible catalytic models.

For P. vivax seropositivity data the best model assumes (i) a single seroconversion rate for every site apart from Sucuriju where a changed in seroconversion rate seemed to have occurred 14 years ago and (ii) two seroreversion rates, one for Anajás, Jacareacanga, Goianésia do Pará, Itaituba and Trairão (ρ = 0.054, 95%CI = (0.033,0.081)), and another one for Sucuriju and Belém (ρ = 0.018, 95%CI = (0.002,0.039)). For P. falciparum data, the best model assumes: (i) a change in force of infection occurred 29 years ago and (ii) a common seroreversion rate for all study sites (ρ = 0.007, 95%CI = (0.001,0.029)). Data of Sucuriju was not analysed due to a low number of seropositive individuals for any P. falciparum antigens. Statistical analysis of age-adjusted seroprevalence data using reversible catalytic models.