Dengue is the fastest-spreading mosquito-borne viral illness in the world, and has become a leading cause of morbidity in many tropical and subtropical regions – including outbreaks in recent years in Mexico, Brazil, Colombia, the Dominican Republic, and other Latin American countries. The direct impact of these outbreaks is exacerbated by the stress they place on public health systems, as hospitals become overwhelmed by a surge in cases of dengue as well as “false alarms” from non-dengue fevers. 

Early-warning systems for outbreaks are becoming an increasingly important tool to mitigate the impact of outbreaks for dengue and other diseases, potentially limiting the spread of the illness as well as enabling public health systems to respond more efficiently and effectively.

Current Imperial College London faculty previously participated in an international, World Health Organization (WHO)-led effort to study predictors of dengue outbreaks and develop an early dengue outbreak warning and response system (EWARS). Public health professionals hoping to make this kind of impact can work with Imperial’s world-class faculty from anywhere in the world through the online Global Master of Public Health. 

From Alarm Variables to EWARS 

The first task of this project was to create a statistical model for dengue outbreaks that met the criteria of having a high sensitivity (number of true outbreak detections) and a low number of false positives. This work was led by Dr. Leigh Bowman, then at Umeå University in Sweden and now a Teaching Fellow at Imperial College as well as Consultant Epidemiologist with the WHO. 

Using data from 2007 to 2013 from Mexico, Brazil, Dominican Republic, Malaysia, and Vietnam, researchers sought to define dengue outbreaks based on numbers of hospitalizations and then retrospectively predict them using changes in three key types of alarm variables:

• Meteorological: Outdoor temperature, rainfall, relative humidity
• Epidemiological: Mean age of infection, circulating type of virus, probable dengue cases, hospitalized dengue cases
• Entomological: Various factors indicating the presence and extent of the specific Aedes mosquito that carries dengue

Logistic regressions were performed on these “alarm variables” to develop coefficients for each to build a predictive model for outbreaks. Traditional epidemiological variables, and probable dengue cases in particular, had the greatest predictive potential, but meteorological variables — particularly mean temperature — were also effective predictors in Mexico and Brazil. Entomological variables, meanwhile, required further study. 

This work provided a foundation for the development of the “Early Dengue Outbreak Warning and Response System” (EWARS), a user-friendly, open-source software tool that provides an “alarm system” for dengue outbreaks based on these indicators. EWARS defines a moving average number of dengue cases within a “normal” state for a given set of local conditions, and then defines an alarm threshold for when these indicators increased the probability of an outbreak beyond a specified level requiring action.

The initial EWARS tool and a revised version (EWARS-R) were developed and tested in Mexico, Brazil, and Malaysia, achieving promising results for both sensitivity and positive predictive values (the proportion of true positives as opposed to false alarms, abbreviated PPV):

• Mexico: 79-100% sensitivity, 50-83% PPV
• Brazil: 83-99% sensitivity, 40-88% PPV
• Malaysia: 50-99% sensitivity, 71-80% PPV

District health managers across all 30 study districts that had used EWARS in the field for seven to 10 months were also surveyed as part of the continuing calibration and development of this tool. Respondents indicated an overall consensus that EWARS is useful overall, but also had a number of suggestions for operational improvements to be included in future versions. This work will continue as the project partners work with stakeholders in dengue-prone countries to scale up and expand the use of EWARS.

Global Collaboration for Global Impact

This research has drawn support from leading international institutions for  its potential to mitigate the impact of dengue outbreaks in different, country-specific contexts worldwide. Primary funding came from the European Commission to the International Research Consortium on Dengue Risk Assessment, Management and Surveillance network.

In addition to Dr. Bowman, these two research papers included global contributors from Europe, Asia, and Latin America. Co-authors included representatives from universities in the United Kingdom and several other European countries; the Ministries of Health of Mexico, Brazil, Malaysia, and the Dominican Republic; and the UNICEF/UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Diseases (TDR).

If you’re looking to make this type of impact on the spread of infectious diseases or other pressing public health issues in Latin America or globally, the online Global Master of Public Health from Imperial College of London’s School of Public Health makes the next step of your career accessible from anywhere in the world. 

The post Imperial Research Teaching Fellow Helps Create “Alarm System” For Dengue Outbreaks appeared first on Coursera Blog.