MALARIA VACCINE STUDY SITE IN IRIAN-JAYA, INDONESIA - PLASMODIUM-FALCIPARUM INCIDENCE MEASUREMENTS AND EPIDEMIOLOGIC CONSIDERATIONS IN SAMPLE-SIZE ESTIMATION

Malaria epidemiologic and entomologic studies were performed during bo th the high transmission and low transmission seasons to characterize the Plasmodium falciparum malaria transmission at a proposed malaria v accine trial site in Irian Jaya, Indonesia. The study population consi sted of two subsets: native Irianese men with lifelong exposure to mal aria and transmigrants who arrived from a nonmalarious area 2.5 years before the start of the study. All subjects received a radical cure fo r malaria and were then monitored weekly by blood film. Both P. falcip arum malaria attack rates and incidence densities were calculated; tra nsmigrants had a significantly higher rate (P = 0.003) than the Iriane se during the low transmission season study (20-weeks long) but not du ring the high transmission season study (12-weeks long). Lack of expos ure-induced immunity left the transmigrants at a minimum 17-25% greate r relative risk of becoming parasitemic compared with the Irianese dur ing the low transmission season study. During the high transmission se ason study, 50% of the transmigrants were P. falciparum positive by we ek 6 and 50% of the Irianese by week 9. During the low transmission se ason, 50% of the transmigrants were positive by week 10 and 43% of the Irianese were positive by week 17. Entomologic studies showed that An opheles koliensis was the predominant vector (> 98% of anopheline catc h). Entomologic inoculation rates for P. falciparum were 0.018 and 0.3 9 infective bites/person/night for the low and high transmission seaso ns, respectively. New P. vivax cases represented between 16% and 42% o f all initial malaria cases. Factors affecting sample size calculation were evaluated, including causes of loss to follow-up, and sample siz es required to complete the study under several sets of conditions wer e calculated. The confidence intervals around specific sample sizes we re calculated highlighting the need to determine the sample size based not only on predicted vaccine efficacy and estimated subject attritio n but also on a desired width of confidence interval for the observed vaccine efficacy.