Watch out for Aedes mosquitoes, the vector that spreads dengue fever.

Limiting dengue outbreaks requires rapid action; early warning systems need to be prompt and efficient. But current dengue reporting systems are often delayed, holding up outbreak responses.

“Detecting a dengue outbreak in its early stages means fewer people get sick,” says Doctor Ronald Skewes-Ramm, Director General of Epidemiology at the Dominican Republic’s Ministry of Health. “Fewer people suffer, and the cost of treating and containing the disease is significantly less.”

Dr. Skewes-Ramm is the Dominican Republic’s collaborator in a global project to develop a new model for dengue surveillance and outbreak response. The initiative is a collaboration between the Special Programme for Research and Training in Tropical Diseases (TDR) and International Research Consortium on Dengue Risk Assessment, Management and Surveillance (IDAMS).

IDAMS International Stakeholder Evaluation Meeting . Photograph by Hanspeter Trefzer, courtesy of Katholische Akademie, Freiburg

Currently in its second phase, the project aims to help dengue-affected countries implement more efficient, affordable, evidence-based outbreak early warning systems. “Our aim is to significantly improve a country’s capacity to detect and respond to dengue outbreaks,” comments Dr. Leigh Bowman, Epidemiologist at the Liverpool School of Tropical Medicine and project research team member.

Phase I: Identifying potential outbreak indicators

The first stage of the project, a ‘retrospective study’ which ran during 2013 and 2014, aimed to identify possible dengue outbreak signs. Five countries took part: Brazil, Mexico, Dominican Republic, Malaysia, and Vietnam.

Local teams on the ground collected data on a wide variety of possible indicators. In the Dominican Republic, for instance, Dr. Skewes-Ramm assembled surveillance data on all local outbreaks since 1997. It included information such as gender, age, date of onset, where the person was treated, symptoms, serotype (if known) and more for around 80,000 dengue cases.

With data from the five countries collected, the research team set about its analysis. “We modeled the historical data, then looked forward to seeing if we could explain the onset of a dengue outbreak by a prior change in one of the indicators,” explains Doctor Bowman.

The team found a combination of just two factors consistently arose before an outbreak: an increase in probable cases and a shift in mean temperature. “We were surprised to discover that certain indicators are predictive of outbreaks irrespective of the country context, especially considering that transmission is driven by many factors that vary from country to countries, such as herd immunity and the person-vector contact ratio,” adds Doctor Bowman.

The dengue analysis helped the team to create the early warning system.

Phase II: Verifying the system

Fast forward about a year to mid-2015. The project team is ready to prove the effectiveness of its prospective early warning system. They’ve developed a simplified Excel version of their model that will be relatively easy for local programme managers and regional epidemiologists in the study countries to test.

Four of the original countries take part in this ‘prospective study’: Mexico, Malaysia, Brazil and Dominican Republic. Twenty districts were identified in each country; ten would act as control areas while the others would use the early warning system. “Our hope is to mitigate or even prevent the number of outbreaks that occur in the intervention districts in comparison to the control districts,” says Doctor Bowman.

The trial will run until mid-2017 to allow the team to capture two rainy seasons in each country. Doctor Bowman explains why: “We must also capture any changes in the variables during the dry season that may be indicative of an outbreak in the following wet season.”

Spreadsheets are currently being completed, and the research team has its first few months of data. Local program managers record interventions taking place: which alarm signals they responded to initially when the intervention started and finished, along with several other factors.

Tailored responses

The project isn’t just about the early warning system detecting potential outbreaks ahead of time; it’s also about delivering appropriate responses when alarm indicators are identified. “If the alarm signals from each of the indicators are relatively weak, an intervention requiring relatively few resources is chosen,” comments Dr. Bowman. “As they strengthen – indicating there’s a higher risk of an outbreak – interventions become more resource-intensive.”

Those interventions, whatever their intensity, are aligned to each local region. Each district has the autonomy to implement interventions as they see fit. “They might decide to run a community-based campaign, and they determine whether it involves leaflets or radio announcements and how long it is going to continue for,” Doctor Bowman says.

International Stakeholder Evaluation Meeting. Photograph by Hanspeter Trefzer, courtesy of Katholische Akademie, Freiburg

An integrated approach

The prospective early warning systems are fully aligned with the World Health Organization’s Integrated Vector Management approach.

“Our plan is to capture dengue, chikungunya, and Zika incident case data,” reveals Dr. Bowman. “That way we can see whether the vector control interventions are working for not only dengue but also for chikungunya and Zika.”

After all, Aedes aegypti and Aedes albopictus – the predominant mosquito vectors of dengue – are also primarily responsible for the transmission of chikungunya and Zika. “We are interested in gaining a better understanding of the impact of vector control, as there are so few studies that robustly capture its effectiveness against both the mosquito and virus transmission,” elaborates Doctor Bowman.

A recent study published by Doctor Bowman et al. demonstrated “the remarkable paucity of reliable evidence for the effectiveness of any dengue vector control method.” It suggests prioritising studies of higher quality to evaluate and compare methods to optimise cost-effective dengue prevention.

More work to do

Further challenges still lie ahead; one being how an outbreak should be defined. Outbreaks are currently characterized by the number of cases relative to previous years – and that can change over time. “Defining a dengue outbreak is important,” notes Doctor Bowman. “As soon as we term a group of cases an outbreak, it starts a cascade of responses.”

Another challenge lies in understanding how to predict the magnitude of any outbreak. The retrospective study – like many models around the world – didn’t quantify the magnitude of an outbreak. Doctor Bowman confirms, “Being able to respond to the correct magnitude and at the correct time means resources can be used more cost-effectively.”

Understanding the stage of the outbreak is also important. The response needs to be intensive during the early stages of an outbreak, more than towards the end when there’s a natural burn out among the population because immunity has increased.

A positive impact

Even though it is still on-going, the project has already had a positive influence on the countries involved. “Previously, we only had a gut feeling of what the indicators for an outbreak were,” Dr. Skewes-Ramm tells us. “If there were a change in temperature or a lot of rainfall, we would expect a lot more mosquitoes. Now we know we have to act if there is also a change in the number of cases or a change in the circulating serotype.”

And whatever the project uncovers, as a bare minimum it will have improved the data collection, surveillance and reporting capabilities of the countries involved. “Our project will help standardize in-country dengue reporting systems and provide a framework that could be rolled out to other countries as well,” concludes Doctor Bowman.

—

Click below to boost dengue fever surveillance, report incidence in your area using Dengue Track.

https://breakdengue.wpengine.com/dengue-track/?utm_source=wordpress&utm_medium=socialpost&utm_campaign=denguetrack