An Estimate of the Global Health Care and Lost Productivity Costs of Dengue

Introduction

To combat infectious diseases globally, it is important to have an accurate understanding of the economic burden that diseases impose. Dengue is one example of a neglected tropical disease for which standardized global cost estimates are unavailable. Country-level cost estimates available in the literature are for only a small fraction of countries in which dengue is endemic, and the methods used to generate these estimates vary considerably and are difficult to generalize.

This report estimates the global costs of dengue fever using recently published estimates of symptomatic incidence by Bhatt et al. (2013). The cost model accounts for two cost categories—direct costs of dengue treatment in the health care system and indirect costs due to lost productivity.

Methods

The sample for our cost estimates started with the 214 countries or territories that are followed by the International Monetary Fund (IMF). We merged our sample with the 139 countries and territories designated as being dengue endemic by Bhatt et al. (2013). Fourteen countries or territories in the dengue incidence data were not found in the IMF set of countries or territories. ^* Of the 125 countries or territories that were successfully merged, 17 were not included in our cost estimates due to either missing wage or health care cost data. ² Our final dataset consisted of 108 countries. Because most of the countries or territories not included in the analysis had relatively small populations, 95 of the 96 million estimated cases (98.9%) were accounted for in the cost estimates.

Symptomatic dengue cases were allocated to one of four outcomes using probabilities based on previous estimates by Shepard et al. (2011). Ninety-four percent of symptomatic cases were assumed to be nonsevere, with 90% of these cases treated on an outpatient basis and the remainder treated as inpatients. Six percent of cases were assumed to be severe dengue fever, with 1.3% of these cases resulting in death. Work/school days lost and health care utilization parameters used for each of the four outcomes were primarily taken from a prospective dengue cost study performed in eight countries (Suaya et al. 2009). These parameters are provided in Table 1.

Information on health center and hospital prices were taken from the World Health Organization's Choosing Interventions that are Cost-Effective (WHO-CHOICE) data (World Health Organization 2008). Costs due to lost productivity were derived from the IMF World Economic Outlook on-line database using per capita gross domestic product (International Monetary Fund 2014). Lost caregiver productivity was accounted for by assuming a 0.5 multiplier for the number of days lost. Death costs were calculated as the summation of the present value of per capita gross domestic product (GDP) per life year lost and discounted at 3% rate per annum (Drummond et al. 2005). Costs and wages were adjusted for inflation and purchasing power using 2011 values (International Monetary Fund 2014) and are expressed in US dollars. To evaluate the sensitivity of our cost estimates due to the uncertainty in the case estimates, we also calculated costs based on the end points of the country-level 95% confidence intervals provided in Bhatt et al. (2013).

Finally, to show the relative dengue cost burden in each of the endemic countries, we estimated that country's total dengue cost per 1000 population and then assigned each to into a quintile category to indicate their position in the distribution of all dengue endemic countries.

Results

For 2011, our cost estimates are based on 95 million symptomatic cases occurring in 108 countries. Of that total, 80 million cases (84.5%) were nonsevere and treated on an ambulatory basis. Nine million cases were nonsevere and treated on an inpatient basis. Almost 6 million cases were allocated to severe dengue (5.9%), of which approximately 74,000 cases resulted in death. More than 60% of symptomatic cases occurred in countries with a per capita GDP of less than $6000.

Global costs due to lost productivity, premature death, and health care utilization were estimated to be $39.3 billion, or approximately $414 per symptomatic dengue case. Temporary productivity loss due to illness accounted for $18.5 billion (47.1% of total costs), health care utilization $9.8 billion (24.9%), and costs due to premature death were estimated to be $11.1 billion (28.2%). Forty percent, or $15.8 billion, of the costs were in the Southeast Asia region, 24.7% ($9.7 billion) were in the Americas, and 21.6% ($8.5 billion) of the costs were in the Western Pacific region. Africa and the Eastern Mediterranean Region accounted for 5.9% ($2.3 billion) and 6.1% ($2.3 billion), respectively. Our total cost estimates ranged from $26.9 billion to $57.3 billion, respectively, based on the lower and upper values of each country's case estimate's 95% confidence interval.

Figure 1 shows the distribution of the dengue cost burden by country. The mean cost per 1000 population from the lowest to highest quintile ranged from $847 (which includes countries such as Afghanistan and Mali) to $53,420 (which includes countries such as Brunei Darussalam and Singapore), respectively.

OPEN IN VIEWER

FIG. 1.

Distribution of dengue costs in 2011, by country. Each color indicates the country's costs per thousand population quintile. Lack of color indicates a nonendemic country. Gray indicates the country is endemic, but cost data is unavailable.

Conclusion

With a total cost of nearly forty billion dollars in 2011, dengue represents a significant financial burden and deserves the attention of global health advocates. However, because more than half of the cases occurred in countries with a per capita gross domestic product of less than $6000 per year, alternative measures that depict the physical burden imposed by dengue (such as disability-adjusted life years) should be used in conjunction with economic estimates when evaluating strategies for disease control. On the other hand, our estimates also provide support that dengue prevention initiatives, such as vaccination or mosquito abatement, could be cost beneficial for middle- to upper-income countries afflicted with endemic dengue.

Several caveats deserve mention. First, the cost estimates presented are higher than those previously reported, due in large part to our reliance on the higher estimates of country-specific dengue incidence in Bhatt et al. (2013) (Suaya et al. 2009, Shepard et al. 2011, Undurraga et al. 2013). Second, we assume that every case receives care in some kind of formal setting, although in reality a substantial proportion of those afflicted may not seek or be able to access care. Third, we also assume that every apparent case of dengue results in a loss of economic activity, which may not be the case in a “milder” form of apparent dengue that does not confine the individual from working. Finally, our estimates do not include the costs of dengue prevention and surveillance activities. In sum, the costs presented in this paper represent a “back of the envelope” estimate using new estimates of dengue incidence and publicly available cost data. Further work refining country-level assumptions about health care use, productivity loss, and other factors will help provide more precise cost estimates for this neglected, but economically important, tropical disease.