Identifying the Birth Defect Epidemic for Zika Virus: Where Are the Relevant Databases?
A Blog post by Elaine Faustman (WDS Scientific Committee member)
Hello! I am a Professor in a School of Public Health who directs an Institute in Risk Analyses and Risk Communication, and in that role I am frequently asked questions on current health risks. The recent Zika epidemic is a significant example of such a request, and provides an opportunity to illustrate use of databases to answer risk assessment questions for this emergent issue.
In risk assessment for Zika virus, we are interested in identifying specific health impacts—including potential birth defects—that may be associated with exposure. We are also interested in the potency of the virus, duration of infection, and whether the duration of the infection relates to the severity of the health impacts. In this post, we pose the question: what databases and data sources exist for us to examine this epidemic but also to be prepared for potential future epidemics? I share with you example databases that I used to answer these questions in a recent journal club. I have also included a series of comments and conclusions about the utility of these databases for risk assessment questions.
Background on Zika Virus
I’d like to start by providing a little background on Zika virus, as one critical step in risk assessment is hazard identification and characterization. Though Zika virus was first discovered in 1947 in Africa, the first large epidemic was not reported until 2007 in the Pacific Island of Yap (Al- Qahtani et al. 2016). Since then, outbreaks have been reported in French Polynesia (2013), and Brazil and surrounding countries (Chang et al. 2016). The first case of Zika virus in Brazil was reported in May of 2015. Currently, 30 countries in the Americas have reported active cases of Zika virus. Though Zika is usually transmitted through the bite of a mosquito from the Aedes genera (Aedes albopictus and Aedes aegypti), it can also be spread through sexual activities and intravenous infection, such as blood transfusions. For most healthy individuals, infection can lead to mild flu-like symptoms or even be asymptomatic. However, infection (both symptomatic and asymptomatic) during pregnancy can lead to irreparable birth defects that severely impair child development (Kleber de Oliveira et al. 2016).
The most common birth defect associated with Zika virus exposure during pregnancy is microcephaly (Rasmussen et al. 2016). The basic definition of microcephaly is 'the clinical finding of a small head compared with infants of the same sex and gestational age' (CDC 2016). Problematically, there is no universally accepted definition of microcephaly; thus, when tracking cases of microcephaly and Zika viruses across healthcare providers, provinces, states, countries, and regions, the criteria employed can be drastically different. Inconsistencies in data collection techniques frequently limit the ability of Public Health professionals to accurately identify and predict Zika-induced microcephaly cases. To add further complications, microcephaly is not unique to Zika infection, but can be caused by a number of environmental and viral exposures, such as toxicoplasmosis, rubella, cytomegalovirus, herpes, HIV, Syphilis, mercury, alcohol, radiation, as well as genetic and maternal health conditions including poorly controlled material diabetes and hyperphenylalaninemia (CDC 2016).
Figure 1: Visual representation of microcephaly (CDC 2016)
This fast spreading epidemic demonstrates the need for access to global databases tracking the spread of mosquito species, infections, and birth defects, both under current and future climate conditions. Next, I will describe databases and data sources relevant to tackling this multifaceted global health risk.
Mosquitos: Because Zika virus is a vector-born infection, tracking the distribution of both Aedes albopictus and Aedes aegypti under current and future climate conditions will be critical to combating seasonal outbreaks, preventing the geographical spread of current outbreaks, and developing long-term strategic interventions to interrupt the vector-host pathway. HealthMap provides an excellent resource for tracking and predicting the spread of Zika virus with up-to-date interactive maps that show the distributions of both mosquito species and Zika infections on a global scale. Through an automated system, HealthMap updates distributions on a daily basis and provides convenient interfaces in nine different languages. Because the Zika epidemic has spread at such an alarming rate, the availability of data in real-time is critical. In addition to Zika cases, HealthMap also tracks Yellow Fever, West Nile Virus, and Chikungunya, which are related to Zika virus. By co-tracking these better characterized viruses, we may be able to translate lessons learned into Zika research and prevention. The Centers for Disease Control and Prevention (CDC) also tracks mosquito distributions in the United States. These ranges show that while Aedes aegypti distributions are primarily in the southern region of the United States, the Aedes Allopictus distribution reaches as far north as New Hampshire, and extends into the mid-west, reaching Minnesota. While this does not mean that Zika will spread in all of these areas, knowing mosquito distribution patterns can help communities prepare and mitigate risks.
As the global climate changes, mosquito distributions are predicted to expand. Many options exist for predicting mosquito distribution changes alongside increased temperatures and changes in global precipitation patterns (see resources below). Many of these programs have been optimized to describe the changes in malaria infections (e.g., Medlock et al. 2015). By translating lessons learned from malaria surveillance programs that predict changes in disease related to climate change, this will be relevant for Zika epidemic prediction.
Zika Infections: Both the World Health Organization (WHO) and CDC are actively tracking global cases of Zika virus. However, because infection can be mild or asymptomatic, it is expected that these may be underestimates. Additionally, Zika infections occurring in underserved communities may go unreported due to lack of access to healthcare.
Figure 3: Distribution of Zika infections in the United States from CDC found here.
Birth Defects Registries: Both CDC and WHO track incidents of microcephaly at national and global scales, respectively. Generally, birth defects are identified by active or passive surveillance systems. Under active surveillance, Public Health or healthcare professionals seek out birth defect information. For example, the expert goes to hospitals and reviews medical reports to find babies with birth defects. Passive surveillance, on the other hand, relies on doctors or hospitals to send reports to the Public Health Department responsible for tracking birth defects. In this model, doctors and healthcare providers must be able to accurately diagnosis birth defects and report them to the proper Public Health Department. Hybrid approaches are also used, in which the surveillance is passive; however, Public Health professionals will follow-up to confirm birth defect reports. For microcephaly, it is particularly complicated due to discrepancies in how the condition is diagnosed. Comparing countries with active and passive surveillance systems is complex and often introduces biases into the analyses. Additionally, depending on the legal and healthcare environment, women carrying fetuses with known birth defects may terminate their pregnancies before a birth defect can be reported, leading to an underestimation of birth defects. These complexities make international comparisons of birth defects complicated.
Dysmorphology: Efforts to standardize the definitions of congenital abnormalities, including microcephaly, are important in harmonizing data collection at national and international levels. CDC uses the Systematized Nomenclature of Medicine Clinical Terms (SNOMED CT) ontology as a controlled vocabulary for describing congenital abnormalities. Additionally, SNOMED CT has compiled an extensive database of known causes of microcephaly, including genetic abnormalities.
Databases were available that answered all of these questions, and which provided additional details on potential challenges related to data collection. However, separate databases need to be consulted to track microcephaly and Zika cases alongside mosquito populations under current and projected climate scenarios. Some of these databases are automatically updated consistently; however, others have to be manually updated and can become out-of-date relatively quickly. Current projections are what are being accessed to answer questions about the global and local risks associated with the upcoming Olympic Games.
The available databases enabled decision-makers to craft location-specific risk communication advice and also to make predictions of vector spread. As with many emerging risks, more information is always needed, and hence the frequency of database updates directly correlated with the increasing frequency of revised messages. Information sources differed in detail and were dynamic. In particular, with birth defects, getting the message wrong or having access to inaccurate data can result in serious healthcare actions. Most of the databases we accessed to make these assessments were government- and/or agency-based databases, best used for population level predictions rather than for use in individual patient-based decisions. At the population level, these databases were exceptionally helpful.
All in all, we found a wide variety of databases available that are relevant to understanding and predicting risks associated with Zika virus. Some weaknesses include: lack of international standards for diagnosing microcephaly, and difficulties in quantifying prevalence of Zika virus in rural and underserved communities; infrequently updated databases; and 'lack of one-stop shopping'. However, there are many promising tools such as HealthMap, which contains information on both mosquitos and Zika cases and is frequently updated.
Special thanks for M. Smith and D. Pyle with the Institute for Risk Analysis and Risk Communication for their contributions to this blog post.
Climate change models for mosquito spread:
– Medlock, J. M. and S. A. Leach (2015) 'Effect of climate change on vector-borne disease risk in the UK.' The Lancet Infectious Diseases 15(6): 721-730.
– Paz, S. and J. C. Semenza (2016) 'El Niño and climate change-contributing factors in the dispersal of Zika virus in the Americas?' The Lancet 387(10020): 745.
– Sucaet, Y., J. V. Hemert, B. Tucker and L. Bartholomay (2008). 'A Web-based Relational Database for Monitoring and Analyzing Mosquito Population Dynamics.' Journal of Medical Entomology 45(4): 775-784.
– Vector Map
WHO Pan American Health Organization:
– 'Zika Virus Infection'