The proportion of fevers attributable to malaria varies significantly between sites in Papua New Guinea

Malaria is endemic across lowland Papua New Guinea (PNG) and case management has been based on symptomatic diagnosis and presumptive treatment of fever cases with an antimalarial. This study aimed to investigate the prevalence of malaria infection among fever cases presenting to 5 purposely selected sentinel health facilities in order to estimate the proportion of patients requiring antimalarial drugs. A total of 1807 fever patients were screened. Overall, 45% of fever patients had a positive malaria blood slide; 35% were infected with Plasmodium falciparum, 9% with P. vivax and 2% with P. malariae. Slide positivity was highest in Dreikikir (75%) and lowest in Wipim (2%). Among patients aged 1-4 years, 22% had moderate to severe anaemia (Hb < 8 g/dI) and 21% of children 2-9 years of age showed signs of splenomegaly (Hackett score 1-5). Comorbidity differed significantly between study sites and was not closely correlated with malaria infection. Clinical diagnosis by health facility staff was malaria for 67% of all fever cases, including 89% of slide-positive and 48% of slide-negative patients. 70% of rapid diagnostic test-negative cases were treated with an antimalarial. It is estimated that due to the lack of parasitological diagnosis the selected health facilities reported an excess of 18% (Dreikikir) to 98% (Wipim) malaria patients on average each month. In consideration of the significant differences in malaria-attributable fevers between study sites, the implementation of parasitological diagnosis in health facilities and administration of antimalarials only to test-positive patients has the potential to significantly improve the management of fever cases and reporting of malaria. A better tailoring to different settings may increase the effectiveness of malaria control interventions.