Big Data research using electronic health records (EHR) and social media (SM)
available through the internet are an effective and constantly improving means to
generate quick and valuable knowledge about the functioning of human beings and
society. They aid to answer important research questions related to health, well-being,
functioning and behaviour of various populations and institutions. This type of readily
available data is valuable for many disciplines including medicine, psychology, and
sociology. It holds the potential to have a direct impact on quality and cost-efficiency of
health care as validated by real time use of Big Data in smarter healthcare and cities
projects.
As research regulations have been originally conceived for clinical trials, researchers
carrying out studies that use EHR and SM face considerable uncertainty. Recently, the
public has reacted strongly to the perception that research involving data and
biological samples (generating genetic data) is underregulated and therefore data abuse
is becoming a high risk. Big Data research has opened previously unprecedented
possibilities of data research and has increasingly highlighted public fears of abuse:
predictive analysis of EHR is being developed for so-called smarter hospitals to help
identify in real time patients at high risk for health deterrioration. Models based on
algorithms are developed to predict and address healthcare and resource allocation in
smarter hospitals. In medicine and psychology, routine data from EHR offer the
potential to use data from millions of people to answer important research questions at
much lower cost than clinical trials. Combination of information from SM, consumer
data (e.g. data from the Migros Cumulus card) and EHR data could enable in finding
objectivable results that were previously only available through more costly subjective
tools such as patient or consumer questionnaires.
This type of very promising Big Data research raises new technical and regulatory
concerns due to its specific characteristics: the vast amount (variety, volume) of useful
data generated at an unprecedented speed (velocity) encompassing not only medical
information but also sociodemographic and financial data. It thus presents challenges
on veracity at various levels of data flow, ranging from the extraction of information to
data sharing and methods of data management and analysis. Among the most
pressing ethical and regulatory concerns are issues of consent, privacy, confidentiality
and conflicts of interest. Countries are struggeling individually and often differently
with the regulatory void, and the varying domestic approaches lead to concerning
barriers for international research. In Switzerland, it is at present not always clear
which type of Big Data research is considered human subject research and must
receive approval from cantonal research ethics committees (CRECs). Furthermore,
CRECs lack capacities to treat all research from such grey zones, especially if they are
done outside of medical faculties. Some researchers have sought to address this
shortcoming by enacting their own department or faculty commissions to approve
research. These commissions, in contrast to the CRECs, do not have a legal status and their competence and scope remain undefined.
The past has shown that if research remains largely unregulated there is a risk of
public “backfire”. Citizens’ fear of abuse and of research taking place “behind their
backs” can create the feeling that regulations must be particularly strict. There are
several recent historical examples to this effect. The present project fills the existing
research gap of how to deal ethically with dsata while at the same time avoiding
unnecessary barriers to otherwise beneficial Big Data research. Guidance for ethics
committees worldwide, and also locally in Switzerland, is urgently needed. The
objectives and methods of this project are the following:
i.Review existing national and international ethical as well as legal guidance related to
Big Data research involving EHR and SM using classical comparative analysis
methods.
ii.Understand the attitudes and local needs of researchers and barriers they fear: to
this aim we will carry out qualitative interviews with (a) researchers and (b) with those
whose data are used.
iii.Use the results from the preceding parts (i. & ii.) and perform thourough ethical
analysis, after review of the relevant ethico-legal literature, to influence the local and
international regulatory and ethical debate and to propose and further a sensitive and
efficient research ethics framework.