Zika virus is new to the Americas and is spreading through populations with no preexisting immunity. Physicians astutely noted a correlation between the emergence of Zika virus and the appearance of birth defects in babies1,2 and neurologic complications in otherwise immunocompetent adults.3 The rapid accumulation of knowledge about Zika virus less than 1 year after the first description of Zika virus circulation in Brazil4 is a testament to the endeavor of the global scientific community, including researchers from affected countries.5
The contrast between the early response to Zika virus and human immunodeficiency virus (HIV)/AIDS is notable. Some involved in the initial response to HIV/AIDS lament that more was not done “when the epidemic was most vulnerable to intervention—the moment it was first reported.”6 The first case reports were published in 1981, the virus was discovered in 1983, and the first official mention from the Reagan administration did not occur until 1986.6 In contrast to the slow recognition of HIV/AIDS as a global concern, worldwide interest in Zika virus soared in early 2016,7 fueled at least in part by information sharing on social media platforms that did not exist 35 years ago.8 But are scientists who are entrusted with the responsibility to inform the public health response to Zika virus taking full advantage of the same technologies? Are we sharing data as quickly as possible when the epidemic is likely most vulnerable to intervention?
Admittedly, we did not think much about this a few months ago. As Zika virus reemerged in 2015, an internal HipChat (https://www.hipchat.com/) group chat was started to share ideas and information. What was first a small chat room with a few infectious disease researchers from our laboratories soon grew to include an interdisciplinary team of reproductive biologists, arbovirologists, clinicians, and other specialists from the United States and Brazil. Soon, the diverse collections of voices were unified under a single moniker: the Zika experimental science team (ZEST). Soon we were confronted with a new question: how could we continue to capture the viewpoints of all the ZEST participants as we began studies with Zika virus? We realized the open-source scientific collaboration platform LabKey Server9 enabled us to publish data in real-time, not just to our team but also to anyone else who was interested. So that is what we did (https://goo.gl/rmNCqf).
We did not expect real-time data to be of much use to anyone outside the close-knit nonhuman primate research community. Many of us in this community have worked together for a long time studying HIV vaccines and pathogenesis. Building on this experience, some initial nonhuman primate experiments with Zika virus were obvious: defining which types of animals could be infected with which virus strains, and via which routes; identifying protective immune responses; and defining whether and how Zika virus infection affects fetal development. By making our plans and results known to our colleagues, we reasoned that we could avoid unnecessary redundancy, use as few animals as necessary, and accelerate the pace of discovery. For example, the viral load data from macaques was used to generate a preliminary estimate of Zika virus doubling time at the request of investigators working to secure the blood supply; this calculation has important clinical implications for determining the duration of the window period when diagnostic tests are inaccurate.
We were surprised to discover that our results portal attracted a broad audience. Users in 98 countries viewed our data (Figure). Even though our experience is anecdotal, it suggests that there is a global interest in real-time data on emerging infectious disease outbreaks that mirrors the general public’s interest in Zika virus. This need not be limited to laboratory data. Clinical information, epidemiologic data, and behavioral studies could also be made available directly and immediately. Indeed, we have worked with LabKey to create a centralized server (http://zika.labkey.com) where other researchers can share any type of Zika virus data in real-time. Already, investigators from Brazil and the United States are preparing data sets for sharing, and we hope that others follow suit.
Color intensity indicates the number of sessions by country. Google Analytics results from February 15, 2016, to March 22, 2016, are shown.
Real-time data sharing is only part of a larger conversation about the best way to communicate results in an emergency. Establishing immediate data sharing as normative behavior within the scientific community, as well as defining formal methods for crediting such contributions, would likely motivate some scientists to be more forthcoming. The Statement on Data Sharing in Public Health Emergencies, issued by companies, research sponsors, academic journals, and others on February 10, 2016, is an encouraging step in the right direction,10 but for its potential to be fully realized, scientists who are generating data need to shoulder the responsibility for making immediate data sharing before publication a de facto standard.
In addition to sharing data, sharing samples also presents an opportunity to maximize scientific progress during an emergency. Anecdotal reports that many qualified scientists have had a difficult time accessing Zika virus samples are worrisome. Legal protections afforded to certain samples in source countries are partly to blame, but some of the blame is due to scientific protectiveness. Samples are not infinite, and there are very reasonable opinions about the extent to which samples should be studied in the places where the samples are collected vs shipped to laboratories in other countries. However, at least some of the blame for the difficulty in sharing samples is likely due to investigators prioritizing the publication of studies in their own laboratories and withholding samples in case new analyses are required during peer review. Relaxing peer reviewer requirements for additional experiments from laboratories that can document sample sharing in public health emergencies could alleviate this tension.
We understand that the data and sample sharing we advocate for Zika virus data does not necessarily apply to other types of studies and may be impractical for some. Nonetheless, our experience shows that there is an appetite for free, Internet-based, widely accessible, real-time scientific information on Zika virus and other novel infectious disease emergencies that will undoubtedly occur in the future.
Corresponding Author: David H. O’Connor, PhD, Wisconsin National Primate Research Center, University of Wisconsin, Madison, 555 Science Dr, Madison, WI 53711 (firstname.lastname@example.org).
Published Online: March 31, 2016. doi:10.1001/jamapediatrics.2016.0857.
Conflict of Interest Disclosures: None reported.
Additional Contributions: We thank Project ZEST team members for constructive comments and helpful suggestions on this Viewpoint. The Wisconsin National Primate Research Center base grant P51 5P51OD011106-54 supported our initial laboratory studies with Zika virus, and grant 3R01AI116382-01A1S1 from the National Institutes of Health supports our ongoing Zika virus projects. Mark Igra, BA, initially encouraged us to use LabKey Server to manage data, and Josh Eckels, BS, LabKey, San Diego, California, facilitates our sharing of Zika virus data. LabKey is a subcontractor on an unrelated National Institutes of Health contract and receives compensation for managing the server that initially hosted the Zika virus data. LabKey did not receive any compensation for assisting with the setup and configuration of the Zika virus data, nor did LabKey receive any compensation for setting up and configuring the zika.labkey.com server that is offering to host projects for other investigators.
Kallas EG, O’Connor DH. Real-Time Sharing of Zika Virus Data in an Interconnected World. JAMA Pediatr. 2016;170(7):633-634. doi:10.1001/jamapediatrics.2016.0857