The burden of P. vivax malaria from Asia has been under appreciated, despite the region contributing almost 40% of the world's malaria . Plasmodium falciparum and P. vivax are equally prevalent over much of the continent, however P. vivax is usually assumed to be benign and its associated morbidity often overlooked [6, 8]. The situation has been further complicated by the emergence of multidrug resistant Plasmodium. Over the last 50 years, Asia has been the epicentre for the evolution and spread of drug resistant isolates of P. falciparum. The latter have subsequently spread to almost all regions of the malarious world, undermining control programmes and significantly exacerbating the burden of malaria . Antimalarial drug resistance in P. vivax has taken longer to emerge, the first reports of chloroquine resistance coming from the island of New Guinea in 1989 . However, over the ensuing two decades highly chloroquine-resistant P. vivax has become increasingly prevalent in Papua province, with reports of declining efficacy also documented across the Indonesian archipelago as well as in Myanmar, India and South America [6, 16–18]. Despite this threat, few studies have addressed the associated epidemiology of malaria in areas where resistance has emerged to both P. vivax and P. falciparum.
The study presented reviews the epidemiology of malaria in patients attending community clinics and the only referral hospital in southern Papua. In this region cure rates following chloroquine plus sulphadoxine-pyrimethamine have fallen below 50% for both P. falciparum and P. vivax, with high grade resistance a significant problem . The results demonstrate that malaria accounts for a considerable proportion of the total clinic workload amounting to 15–27% of patients attending outpatient clinics with fever and 34% of all hospital admissions. P. vivax was responsible for 42% of malaria treated in the outpatients and 19% of patients admitted to hospital with malaria. In the household surveys, the overall prevalence of malaria was 16%, lower than that reported in West Sumba, Indonesia (31%)  and PNG (51%) , although the proportion of carriage attributable to P. vivax was higher: 39% in Timika compared to 32% in West Sumba and 19% in PNG.
There were significant differences in the age-stratified rates of infection (Figure 2). The workload due to malaria peaked in the 5–14 age group, reaching 24% in hospital outpatients and 42% in inpatients, however this was mainly attributable to an increasing prevalence of clinical infections with P. falciparum. In contrast, the burden and proportion of P. vivax peaked in early childhood with more than half of malaria in infants attributable to P. vivax irrespective of healthcare facility. These findings concur with studies from PNG, Vanuatu and Thailand [20–22]. A notable feature of P. vivax infection is the presence of the hypnozoite stage which in equatorial regions can result in frequent relapse and recurrent symptomatic infection. The predominance of P. vivax in infancy presumably reflects early exposure to infection, high rates of subsequent recurrence and the rapid acquisition of immunity. The additional impact of chloroquine resistance is difficult to gauge, but likely to further exacerbate the recurrence of malaria following partially effective treatment. In contrast to the PNG study, in which children had acquired almost complete immunity by the age of 9 years old , both the present study and a study from Thailand observed a significant number of older children and adults presenting with symptomatic P. vivax infections . In Papua, economic immigration has resulted in a high proportion of the population being non-immune, either Papuans from malaria-free highland areas or non-Papuans from provinces with low endemicity. With little or no prior exposure to malaria these migrants are vulnerable to symptomatic infection into adult life. Indeed the proportion of malaria carriage attributable to non-Papuans rose with increasing age (Figure 4b), reflecting not only their vulnerability to symptomatic infection, but also their higher exposure to infection.
In the prevalence survey, approximately a third of patients with peripheral parasitaemia reported an associated fever. Although there was no difference in this proportion between P. falciparum and P. vivax infections, the pyrogenic thresholds differed considerably: 310 μl-1 for P. vivax and 1,734 μl-1 for P. falciparum. The ability of P. vivax to induce fever at lower levels of parasitemia than P. falciparum is well described [22, 23] and is consistent with a greater host inflammatory response during infections with P. vivax, as evidenced by higher plasma concentration of fever-inducing cytokines, such as TNF, in vivax malaria compared to P. falciparum infections with similar parasitaemia [24, 25]. Whereas the proportion of symptomatic P. falciparum carriage did not change with age, there was a significant reduction in symptomatic carriage with P. vivax, falling from 67% in infants to 28% in adults (Figure 4a). All patients with peripheral parasitaemia were treated and, therefore, the proportion of asymptomatic patients who would have become febrile if left untreated can not be gauged.
Since the surveillance system did not include a prospective cohort an estimate of the incidence of clinical infection was derived from the clinic workload and a treatment seeking behaviour survey. The latter will be reported more fully in a separate paper, however the results highlight that only 39% with symptomatic malaria would have sought treatment at a facility in the surveillance network, the majority either staying at home or seeking treatment through the private sector or pharmacies. Extrapolating the total number of reported malaria cases we estimated an incidence of malaria of 512 and 322 per 1,000 population per year for P. falciparum and P. vivax, respectively.