On February 23, 2017, Alice Panchaud, PharmD, PhD, clinical pharmacist and pharmacoepidemiologist at CHUV, Lausanne University Hospital, Switzerland, presented research calling for the centralization of data through the use of an international registry at the First International Zika Conference.
During the session on epidemiology, Dr. Panchaud opened her presentation by discussing congenital Zika syndrome (CZS), a range of abnormalities caused by Zika virus infection during gestation. She reminded her audience that there is still much that remains unknown, such as: “the magnitude of the risks of maternal-fetal Zika virus (ZIKV) transmission, the full composition of the spectrum of the congenital Zika syndrome,” and the potential associated co-factors “that could modify these risks.”
So, how can researchers go about answering the questions that remain unanswered? According to Dr. Panchaud, “We definitely need a very large sample size. If we face this, CZS, we also face very rare outcomes.” She then presented the captivated audience with a table of rare outcomes of particular interest, such as: microcephaly, anthrogryposis, antepartum stillbirth, optic nerve hypoplasia, and intrauterine growth restriction. She reiterated, “If you want to detect differences in risks, then you need a very large sample size.” One example that she provided was that, in order to detect microcephaly prevalence correctly, 2,000 patients would be needed.
Next, Dr. Panchaud shared with the audience what would be needed to accurately identify potential associated co-factors. She said, “If you want to be able to identify those co-factors that modify the risks that we observe, you will have to break down your population. For example, if you want to identify or analyze co-infections with Dengue, it will have to have equal strata from your population with or without Dengue positive testing.” Another example that she provided had to do with Zika symptomatic infection; in order to identify or analyze Zika symptomatic infection, both symptomatic and asymptomatic strata must be considered.
She continued, “We already have a lot of available datasets and we also know that a lot of people have ongoing data gathering and so, we thought that if we put those together, it could maximize the information we already have and it could help us get answers in a timely manner. So, this means data sharing.”
According to Dr. Panchaud, there has been an increased use of data sharing in clinical research. However, “it is not the norm yet.” She reminded the audience that data sharing has been around for a long time and physicists, astronomers, and geneticists, have been using it for a long time now.
She shared, “There is now a pressing ethical and scientific imperative for data stemming from clinical trials with people suggesting that it should be a condition of publication in major medical journals and doing that means that it’s going to change massively, probably in the next few years.”
Although data sharing is important “if you want to have a large, rich, acceptable sample size to assess risks,” Dr. Panchaud admits that there are some pitfalls. First of all, practical implementation may be difficult; infrastructure is needed in order to “accommodate a data sharing consortium” and regulation compliance “can be very tricky, especially if you do something that is crossing borders.” Another important thing is ensuring that all of the investigators that are generating data receive credit and every patient’s privacy is maintained.
Although it is not as easy as it may have seemed at first sight, Dr. Panchaud feels that data sharing is a tool that could really help provide answers to questions regarding the Zika-associated risks in pregnancy, one that will offer a global perspective that is much needed.
To appropriately address some unanswered questions regarding CZS, in September 2016 Dr. Panchaud and colleagues launched an international web registry
, a registry that would help them “characterize the spectrum of disease, estimate the risks, and identify the factors that could modify observed associations” between Zika infection in pregnancy and “outcomes of interest,” such as “symptomatic ZIKV infection, living in endemic region, trimester at exposure, Dengue positive serology, [and] high maternal viral load.”