Abstract
AbstractDengue virus (DENV) is a serious threat to public health. Having reliable estimates of the burden of dengue is important to inform policy and research, but surveillance systems are not designed to capture all symptomatic DENV infections. We derived the rate of reporting of dengue by comparing active surveillance of symptomatic DENV infections in a prospective community-based seroepidemiological cohort study (N = 1008) of acute febrile illness in Punta Princesa, Cebu City, Philippines, with passive surveillance data from the Cebu City Health Department. Febrile episodes detected in a weekly follow-up of participants were tested for serotype-specific DENV by hemi-nested reverse transcription-polymerase chain reaction (nested RT-PCR) and acute/convalescent blood samples tested by dengue IgM/IgG enzyme immunoassay. We estimated the burden of dengue in the Philippines in disability-adjusted life years (DALYs), and conducted a probabilistic sensitivity analysis using Monte-Carlo simulations to address uncertainty. The results showed a 21% cumulative reporting rate of symptomatic DENV infections, equivalent to an expansion factor of 4.7 (95% certainty level [CL]: 2.2-15.1). Based on surveillance data in the Philippines for 2010-2014, we estimated 794,255 annual dengue episodes (95% CL: 463,000-2,076,000) and a disease burden of 535 (95% CL: 380-994) DALYs per million population using age weights and time discounting and 997 (95% CL: 681-1,871) DALYs per million population without age and time adjustments. Dengue imposes a substantial burden in the Philippines; almost 10 times higher than estimated for rabies, about twice the burden of intestinal fluke infections, and about 10% of the burden of tuberculosis. Our estimates should inform policy makers and raise awareness among the public.