Dengue is the most common arbovirus infection globally, with transmission occurring in at least 128 countries and almost 4 billion people at risk.1 The number of dengue cases reported to WHO has increased steadily from an average of less than a thousand cases per year globally in the 1950s to more than 3 million cases in 2013.2, 3, 4, 5 These reports, however, greatly understate the problem and estimates of the true number of annual apparent infections range from 50 million to 200 million, where apparent infections are defined as all symptomatic infections, including those that are undetected by reporting systems. The most commonly cited range, including by WHO, is 50 million to 100 million apparent cases per year.6, 7 Although estimates of dengue deaths are less often reported, the most commonly cited number is around 20 000 deaths per year.8 To our knowledge, these estimates seem largely to be based on expert opinion. The Global Burden of Disease Study 2010—providing the most recent data-driven estimate of dengue deaths—estimated that more than 14 000 people died from dengue in 2010.9
The disparity between the number of reported cases and estimates of the number of actual cases stems from under-recognition and under-reporting of dengue. Symptomatic dengue infections have a broad range of severity and as many as 70% of patients choose to not seek treatment or treat themselves.7 Even for those who are seen by a health-care professional, the clinical presentation of dengue shares similarities with up to 12 major pathogens, making misdiagnosis common, particularly in areas with a high incidence of febrile illnesses.10 Population-based cohort studies11, 12 have consistently found dengue cases to be greatly under-reported through official passive surveillance and reporting systems. Several studies have attempted to quantify the degree of under-reporting by comparing incidence rates derived from active febrile-illness surveillance with comparable incidence rates derived from official reports. The ratio of these rates is referred to as the expansion factor, and it represents the number by which one would multiply the number of reported cases to derive the number of true apparent dengue infections in a given population. That said, the degree to which dengue is under-reported varies by orders of magnitude across time and space, precluding the use of a simple multiplier. Moreover, many countries where dengue is believed to occur file no official reports, or do so only intermittently, and in these cases there is no number of reported cases to which a multiplier could be applied.
Research in context
Evidence before this study
A PubMed search for “dengue AND (burden OR estimates OR model)”, done in October, 2015, with no restrictions placed on date of publication or language, returned 1418 results. Screening abstracts and titles identified 40 articles reporting original estimates of dengue incidence, mortality, or health gap metrics. Full-text screening eliminated 19 articles that did not provide original estimates for areas larger than a community. We, therefore, identified 17 studies that reported original country-level dengue estimates, and three studies that reported original region-level dengue estimates. Only one study contained global dengue estimates, and this study produced only incidence estimates for a single year. We identified no studies that reported global dengue estimates for multiple years. Moreover, we identified no studies that reported global estimates of either dengue mortality or health gap metrics.
Added value of this study
To our knowledge, this study is the first to provide estimates of dengue incidence, mortality, years of life lost, years lived with disability, and disability-adjusted life-years at country, regional, and global levels, and the first study to provide global estimates of dengue trends. As such, it is the most comprehensive assessment of the burden of dengue so far. Our estimates offer important information about the growing importance of the disease. Estimates of health gap metrics, which account for disease severity and give greater weight to deaths occurring at younger ages, enable comparisons between diseases of differing severity, and give policy makers the information necessary to set objective priorities.
Implications of all the available evidence
Our results suggest that there are about 58·4 million annual symptomatic dengue infections resulting in about 10 000 deaths per year. Although lower than another estimate of 96 million symptomatic infections, the two estimates have overlapping uncertainty intervals, and bracket the commonly cited range of 50–100 million cases. Our study offers more evidence that the true incidence probably falls within that range. Furthermore, our results suggest a dramatic increase in the incidence of dengue in the past two decades, with the number of symptomatic dengue infections more than doubling every 10 years between 1990 and 2013. This increase makes dengue an outlier compared with near universal declines in most communicable diseases. We hope that these improved estimates of dengue incidence and mortality, and their longer term trends, will inform public health officials, scholars, and policy makers to assess and identify cost-effective control strategies to reduce transmission and the burden of dengue.
Attempts to estimate the true incidence of symptomatic dengue must, therefore, address these problems. Bhatt and colleagues7 applied geostatistical methods to the problem: they first developed a global dengue risk map, then geolocated studies of dengue incidence and, finally, modelled the relation between risk and incidence to estimate the incidence for each 5-by-5 km area. Their method yielded an estimate of 96 million (95% credible interval 67 million to 136 million) apparent infections globally. Unfortunately, this method cannot easily be used to estimate changes in dengue incidence over time.
Beyond incidence and mortality, understanding the true burden of dengue demands the estimation of metrics that allow for meaningful comparisons with other diseases that have different severity and duration, and that allow for comparisons between fatal and non-fatal outcomes. Among the most common of these burden metrics are years of life lost to premature mortality (YLLs), which quantifies health loss due to mortality, giving greater weight to deaths occurring at younger ages; years lived with disability (YLDs), which quantifies non-fatal health loss accounting for both the severity and duration of a given condition; and disability-adjusted life-years (DALYs), which is the combination of YLLs and YLDs.
We estimated dengue mortality, incidence, and burden by age, sex, and country, as estimated for the Global Burden of Disease Study 2013. The Global Burden of Disease Study 2013 was an effort to comprehensively and systematically estimate death and disability from 306 causes, producing estimates by year, age, sex, and country for 1990–2013. Although summary results have been published previously,13, 14, 15 here we present previously unpublished details of our modelling approach and results for dengue and discuss these results in the context of independent attempts to estimate the burden of dengue, with reference to dengue-specific literature.