U.S.-Mexico Border Early Warning Disease Surveillance for Zika, Dengue and Chikungunya
Project Name: U.S.-Mexico Border Early Warning Disease Surveillance for Zika, Dengue and Chikungunya / Kidenga App
Project Status: Completed
Point of Contact: Stephen Waterman
Center: National Center for Emerging and Zoonotic Infectious Diseases
Keywords: Participatory surveillance, Zika, Dengue, Chikungunya
Project Description: Problem – Traditional passive surveillance systems relying on the medical community often do not rapidly detect human illness until there is high morbidity in the community or more severe disease presentation. While Zika and dengue are nationally notifiable, symptomatic infection most commonly presents as a mild to moderate, nonspecific, acute, febrile illness, is not always likely to come to the attention of a medical provider, and thus likely to be underreported. Symptoms of chikungunya, most commonly fever and joint pain, tend to be more severe, but may also be non-specific. Chikungunya is not yet nationally notifiable and is not reportable in Texas and Arizona. Earlier detection of febrile illness and Aedes mosquitoes would allow for earlier response at the local, state, and federal levels to implement or enhance more targeted mosquito control activities, and minimize or prevent morbidity due to Zika, dengue and chikungunya viruses.
Solution and Methods – We propose to develop the technical infrastructure for, and pilot, a participatory surveillance system in southern Arizona, south Texas, and west Texas to detect febrile illness and day-biting mosquitoes, such as Aedes aegypti, in near real-time using mobile technology. We will partner with Skoll Global Threats Fund (SGTF) and University of Arizona to modify and enhance existing technologies developed for febrile illness surveillance in the U.S. and Puerto Rico (‘Flu Near You’ and ‘Salud Boricua’) to develop a border surveillance system. Flu Near You, monitors influenza-like illness, as well as Salud Boricua, which additionally monitors dengue, chikungunya, and leptospirosis. These systems allow users to respond to a weekly survey of symptoms using apps for their Android or iPhones, or by email. After making the weekly report, users are directed to the website where they can see a map with aggregate data of activity (or lack of activity) in their area, visualize trends, access influenza or other disease specific news content, and connect with local public health resources and organizations. When participants first register, they must agree to share an email address, zip code, age, gender and Internet Protocol address. All data exchanged with the sites are protected by SSL encryption, the same standard used to protect the transmission of credit card information and other sensitive data over the internet. Submissions are anonymized and only posted in aggregate with other submissions.
University of Arizona will create a similar system for the border region, but with enhancements that incorporate surveillance for the vector. The weekly survey of symptoms would include history of sustaining a mosquito bite during the day. Participants will additionally be able to initiate reports of day-biting mosquitos and breeding sites with associated geolocation data. Aggregate data would be available to participants and stakeholders on a Kidenga webpage. The app would also be used to provide up-to-date health risk communication, alerts, and education. Social media and other marketing strategies would be employed to recruit participants in the surveillance, and SGTF will fully fund the participant recruitment and marketing efforts, estimated at $90,000. We would again leverage the significant experience of SGTF and HealthMap in participant recruitment and marketing of the apps based on lessons learned from Flu Near You and Salud Boricua. The serious threat of dengue from the ongoing outbreak in Mexican border states may be particularly advantageous for participant recruitment efforts in the border region.
Follow up of participatory surveillance reports – An optional feature designed into the application will permit participants to confidentially and automatically notify local public health departments and local environmental health agencies and/or mosquito abatement districts of their willingness to cooperate with laboratory based confirmation of arbovirus diagnosis and identification of mosquito vector species observed on or near a residence. Health departments can request blood samples from reporting febrile individuals and entomology workers can collect larvae from water containing breeding sites or place traps for mosquito adults permitting Aedes aegypti species identification from participatory reports. In this way participatory surveillance can enhance, complement, make more geographically comprehensive and timely, existing laboratory and field based surveillance systems for vector-borne diseases. Such two-way web-based communication between the community and agencies responsible for disease control will also validate and permit better interpretation of data arising from participatory surveillance (such as proportion of reported day-biting insects representing Aedes aegypti). Agencies can use geolocation data even without direct communication from individual participants to target human disease and vector surveillance through health messaging and field visits.
Impact – The immediate product will be a functional platform for a sustainable, low cost, participatory surveillance system for acute febrile illness and for vector exposure and breeding. We expect full implementation and impact to take two years, as outbreaks are not predictable and experience from ‘Salud Boricua’ in Puerto Rico suggests that marketing and socializing the use of the mobile applications will take time. The interest/involvement of many collaborators could also bring more resources to bear to implement further stages of the project such as development of a web-based platform for reporting by persons without Android or iPhones, evaluations of the alerts and health communication, and participants’ health care-seeking behaviors. In the longer term, such a participatory surveillance system would provide near real-time information on the presence of febrile illness and Aedes mosquitoes in Arizona and Texas, and potential Zika dengue and chikungunya outbreaks. With earlier detection, local and state health departments, and CDC, will have improved understanding of the distribution of Aedes mosquitoes along the southern border, and will be poised for timelier vector control efforts and community education, and ultimately minimize morbidity and mortality due to Zika, dengue and/or chikungunya. This project would also provide valuable additional experience in using participatory digital surveillance technology for dengue and chikungunya illness and vector distribution and density in a different part of the North American/Caribbean region Lessons learned will potentially allow for greater use of this type of surveillance in other regions, and will also provide more information about how to use social media to provide up-to to-date health risk communication and education related to arboviruses.
Evaluation – A specific evaluation plan will be developed by stakeholders using in-kind resources. The plan will include early efforts to assess the acceptability of the apps to smartphone users, such as sending a survey to the first wave of participants after they complete their first report. As public health messaging to users is implemented, users could also be surveyed on the usefulness of the alerts or educational information. Syndromic data will be compared to routine reports of notifiable conditions to look for possible correlation. As resources permit, follow-up will be conducted for a subset of reports of suspect cases, such as encouraging the participant to see a doctor for serologic testing, supported by the local health department. Reports of day-biting mosquito activity and breeding grounds will be compared to routine mosquito trapping distribution and density data to look for correlation. As resources permit, follow-up will also be conducted for a subset of reports, with attempts to collect and perform speciation of mosquitos for interpretation and validation purposes, conducted by local vector control programs. Unlike participatory surveillance systems for influenza, we have no gold-standard with which to compare syndromic surveillance data for Zika dengue and chikungunya. Ultimately, a recognized outbreak would need to be identified by traditional surveillance systems in the catchment area to retrospectively evaluate whether this border system provides early warning.