The global COVID-19 pandemic is having an unprecedented impact worldwide. Everyone is affected. Health systems are struggling to keep up, while patients and communities are confronting a once-in-a-generation trauma. There is, however, hope; the pandemic may prove to be a positive turning point in healthcare — disrupting how we think and behave in the future to address threats from infectious diseases and fill gaps in our knowledge of science and human health. We have the opportunity to learn more, and do more, following this period in our history, enabled by Human Data Science.
The coronavirus COVID-19 pandemic has been a major wake-up call for systems and individuals around the globe. Despite our collective experience with pandemics in the recent past (from SARS and MERS to H1N1), we were still ill-prepared to respond to the speed and devastation of COVID-19; we learned quickly that our health systems are deeply vulnerable to surprise attacks. It is clear that we need a better, more interdisciplinary effort to direct research and establish strategies for prevention, early intervention, and action against major disease outbreaks.
This is our moment for action. Not just against COVID-19, but against many of the unsolved issues that plague our health and our systems. The pandemic can – and should – give us new motivation and urgency to look differently at diseases, including the origins of pathogens and their transmission, and to assess what we can do to prepare and better respond to threats in the future. This includes ensuring that supply chains function under stress, that effective public health interventions can be activated, and that treatments and vaccines can be made available at an accelerated pace.
But we cannot achieve new results with the same tools. We need a sea change in how we tap into what we know about health and humans, and what we can do with more robust data. This transformation in understanding and improving both human health and the health system will need Human Data Science. Human Data Science is an emerging discipline — a multidimensional approach that integrates advances in human science, a deep understanding of the proven drivers of human health, and data science and technology.
- Human Science, our understanding of the origin of pathogens, the etiology of infectious diseases and viruses, that impact our ability to deploy interventions — both pharmaceutical (i.e. antivirals and vaccines) and public health (i.e. containment, mitigation and social distancing)
- Human Data, the health, behavioral, and social factors that can be collected outside of a clinical setting that influence virus spread and susceptibility (i.e. animal and human interactions, food and diet practices, urbanization, travel patterns, and communication)
- Data Science, technology (i.e. AI, machine learning, and advances analytics) that can be used to identify community spread, conduct contract tracing, and manage outbreak clusters . AI and ML also provide opportunities for accelerating clinical development by identifying drug repurposing opportunities, enabling in silico screenings, and giving new insight into real-world safety and efficacy of new therapies and vaccines.
A framework for applying Human Data Science in the context of the COVID-19 Pandemic
This framework offers an approach to look holistically at all the aspects regarding COVID-19 from a Human Data Science perspective, encompassing factors pertaining to Human Health, Human Science, and Data Science in an integrated, multidisciplinary manner.
COVID-19 has made it clear we need to rethink our approaches to public health and healthcare. Taking an optimistic view, this pandemic may eventually be a catalyst for positive change across a range of healthcare system issues that have long been recognized but that so far have been inadequately focused on or tackled. Many of these issues fall at the intersection of human science and data science where Human Data Science can provide innovative approaches and solutions to strengthen health systems and improve human health and wellness.
A new approach: 10 areas to rethink with Human Data Science
- Understanding the natural history of disease: We are learning more about the coronavirus COVID-19 every day, such as the genetic profile of the pathogen and the role of comorbidities in the most severe infections, including diabetes, obesity, hypertension, heart disease and immune disorders. But there is still a lot that we do not know, and the disease remains a mystery. Multidisciplinary research is needed to elevate our knowledge and fill the gaps that have prevented timely and effective responses. But the value of understanding the natural history of disease extends far beyond responding to virulent infections. There are approximately 7,000 rare diseases, many of which lack a well-understood natural history.1 Greater focus needs to be placed on developing a basic, foundational understanding of disease by harnessing large registry and real-world evidence datasets which increasingly exist but remain underutilized. As we are learning in real (and uncomfortably rapid) time, the better the understanding of the disease, the faster and more effective our response can be, and the more lives can be saved.
Animals are the origins of the majority of human infectious diseases, and while the origin and the reservoir of the COVID-19 pathogen is still disputed,2 there are many indications that this pathogen has an animal origin. We must do more research on animal to human transmission of zoonotic disease, which will enable earlier identification of reservoirs and help prevent the spread of disease by changing protocols and practices around animal handling, including that of food animals, the sourcing of exotic food animals, production practices, or hygiene at wet markets.
Here is an interesting intersection between science, health, and human behavior that has so far gone under-studied. The sociocultural drivers associated with the transmission of zoonoses demand closer scrutiny and understanding given the known risk factors. Engaging anthropologists and sociologists will be essential to addressing these issues of zoonotic transmission at a societal level.
- Bridging the gaps in understanding of the epidemiology of disease. In the flurry of efforts to understand, analyze, and manage the current global pandemic, the lack of clear methodologies and international standards for tracking understanding and navigating disease outbreaks has been a critical weakness. We are witnessing variations in the quality of reporting of patient cases, and even the lack of a clear definition of deaths from underlying conditions vs. deaths from COVID-19, as demonstrated in the absence in the case reports of excess death rates (EDRs) from the COVID-19 infections versus deaths from baseline conditions. Furthermore, the COVID-19 pandemic has revealed an institutional forgetfulness - an absence of historical memory of great epidemiologic evidence. The vast majority of the learnings about the benefits of nonpharmaceutical interventions, such as social distancing measures, were perfectly described and analyzed many years ago in the masterful study of “Nonpharmaceutical Interventions Implemented by US Cities During the 1918-1919 Influenza Pandemic” published in JAMA in 2007.3 The learnings from the 1918-19 flu pandemic are surprisingly relevant more than 100 hundred year later.
Looking forward, the valuable knowledge culled from previous pandemics can be further augmented with the learnings from the current pandemic by using Human Data Science to combine the most recent knowledge in human science, such as molecular biology and genomics, with the understanding of human behavior and the insights from advanced digital technologies and predictive analytics that were not available 100 years ago.
- Overcoming critical path barriers to accelerate the discovery and clinical development of new diagnostic tools and medicines/vaccines: There are currently no antiviral therapies or vaccines available to treat or prevent COVID-19, and the development of both therapeutic and immunization options will take several months, even though manufacturers have escalated their work to accelerate their discovery and clinical development programs. This exemplifies the general challenge that the biopharmaceutical industry is facing across all therapeutic categories where the median time from patent to patient is 13.7 years and the composite clinical development success rates is estimated to be 7.6%. The steps taken to accelerate development of a COVID-19 vaccine — such as scaling up production capacity in parallel with Phase II trials and other initiatives being taken by the Coalition for Epidemic Preparedness Innovations — may trigger broader use of innovative trial designs that have otherwise been making slow progress in a relatively conservative scientific and regulatory environment. But improving the current process and tools is only half the solution. The other half will come from new, innovative uses of data science which can tap into the growing volume of data and methodologies and mobilize artificial intelligence and machine learning to activities such as drug repurposing and virtual screenings.
- Advancing point-of-care diagnostics: The pandemic has also revealed painful gaps in diagnostic testing capabilities, resulting in the inability to conduct testing of populations with the speed and at the volume required for a pandemic. This has prompted FDA’s Emergency Use Authorization of Abbott’s ID NOW COVID-19 portable molecular point-of-care test that can deliver positive results in as little as 5 minutes and negative results in 13 minutes. This and other rapid tests align with a major need to shift diagnostic medicine towards point-of-need solutions. Of course, adequate reimbursement and investment in diagnostic innovations will be required to ensure a fast ramp-up of large-scale testing. Monitoring the accuracy of these critical tests – levels of specificity and sensitivity – will also be important to ensure public confidence and appropriate health system use and will be mission critical to better/faster public health decision making. Human Data Science can help generate insights on how to build confidence in such diagnostic technologies by understanding human behavior, in particular among people who are non-symptomatic and may hesitate to accept the value of early detection.
- Embracing new models for broad industry and intersectoral collaboration: Collaboration between researchers in academia, public research centers and the private sector is proving critically important in advancing the development of vaccines and therapeutics for COVID-19. No individual company or research group has the capacity to both develop the technologies and therapies and scale-up the required production and global distribution on their own. Across the globe, we have seen major competitors — including therapeutics and diagnostics companies, pharmaceutical manufacturers, technology companies and service providers — come together in an unprecedented way to combine their scientific and technical skills. Such collaborations in the midst of the pandemic may pave the way for future industry and intersectoral efforts to focus on Alzheimer’s disease and other public health priorities.
- Improving commercial viability of vaccines and curative medicines: Vaccines that eradicate pathogens and prevent the occurrence and spread of infectious disease are among the most effective and successful public health interventions. However, vaccines that deliver a “one-and-done” intervention for a patient to eradicate a disease/ infection or prevent an infectious disease through immunization are less commercially attractive than medicines that treat and manage chronic conditions and therefore deliver a sustained stream of revenue to manufacturers and suppliers. This reflects a general challenge for all types of interventions that are curative, whether they be vaccines, gene therapies or other “one-and-done” treatments that disrupt existing payment and reimbursement models. An increased impetus may now be given to innovative financing arrangements for these sorts of treatments that to date have proven vexing to payers.
- Applying digital technologies and AI in real time to detect, track and diagnose disease: The COVID-19 outbreak has provided new learnings for how to manage a severe communicable disease in the digital era. The ubiquitous nature of personal digital tools has fueled a 24/7 flow of information, ranging from official guidance from public health authorities to rumors that have fueled anxiety and panic. But the digital world has also provided early warnings, continuous updates and information that are very useful in activating public health strategies and identifying, controlling, and containing disease outbreaks. Imagine trying to deploy current “shelter in place” guidelines, and getting the response, 15 years ago. Digital health tools, including smartphone apps that individuals use to enter their health conditions or monitor vital signs, can help leapfrog traditional approaches to public health by enabling the application of Artificial Intelligence to data streams to predict disease spread and assess the impact of interventions. This is an enormous opportunity for transforming the way in which communicable diseases are tracked and managed.
- Getting back to basics in personal hygiene and vaccination: The COVID-19 epidemic has elevated the urgency for individuals to improve their personal hygiene habits — washing their hands, avoiding touching their face, etc. — as well as for companies and institutions to take measures to regularly clean and disinfect schools, institutions, offices and public transport. These are measures that are relevant for other infectious diseases and viruses, including the flu and common cold. Moreover, the focus on COVID-19 is a reminder of the critical role that vaccines play in public health. The recent outbreaks of measles across many countries has similarly been a reminder of the importance of maintaining high vaccination rates to prevent the spread of this highly contagious disease. Combined with the COVID-19 pandemic, we should see a renewed focus by public health leaders, healthcare workers and citizens to ensure safe and effective vaccines are used broadly, including those that are well established for flu, pneumococcus, and human papillomavirus.
- Sharing access to data: In the case of COVID-19, the importance of fully sharing data across national borders and between parts of health systems has been reinforced by the World Health Organization and researchers across the globe. Yet it has been hampered by legal, cultural and operational barriers. This has resulted in a multiplicity of metrics, models and interpretations of the spread of the virus and the effectiveness of containment and mitigation measures. Global efforts to remove these barriers, so critical for the fight against infectious disease outbreaks, should now be redoubled. This is particularly the case with new, emerging diseases where there is no existing understanding of the natural history of disease and uncertainty about the underlying cause, which make it difficult to identify and confirm disease, or trace and track the spread of infection. However, the sharing of data, expanding access, and providing transparency are equally important for non-communicable, chronic diseases, such as cardiovascular disease, diabetes and cancer.
- Assessing the global and local nature of the pandemic: A global pandemic is not simultaneously global, or even national. Rather, it emerges and moves in geographic clusters. While the entire world has ultimately been impacted by COVID-19, the epidemic has not hit with the same power simultaneously across the globe. As of April 27, the pandemic has now peaked and is receding in China where the epidemic started, as well as in Korea, but is rebounding in Singapore. It may have peaked in Italy and Spain but is still growing in the U.K. In the United States it is growing at the national level, but state-level dynamics are very different. This means that the same measures are not relevant at the same time across all time-zones and geographies even as countries assess how to prevent the outbreak from reemerging. Furthermore, a one-size-fits-all approach is not meaningful across geographies that have different levels of impact.
Learning from the global pandemic
While the global COVID-19 pandemic continues to challenge the world, health systems and science, constantly yielding new questions, it has also has become a learning journey with the potential to create new insights and knowledge on how we address fundamental unresolved issues in healthcare in the intersection of human health, science and evidence generation. Human Data Science has the potential to be a critical guide on this journey.
This article is the first in a series that will explore the Human Data Science perspective of the COVID-19 learnings.
1IQVIA Institute for Human Data Science. Orphan drugs in the United States (Part One): Growth trends in rare disease treatments. 2018 Oct. Available from: https://www.iqvia.com/insights/the-iqvia-institute/reports/orphan-drugs-in-the-united-states-growth-trends-in-rare-disease-treatments
2Cyranoski D. Mystery deepens over animal source of coronavirus. Nature. 2020 Feb 26. Available from: https://www.nature.com/articles/d41586-020-00548-w
3Markel H, Lipman HB, Navarro JA, Sloan A, Michalsen JR, et al. Nonpharmaceutical interventions implemented by US cities during the 1918-1919 influenza pandemic. JAMA. 2007;298(6):644–654. doi:10.1001/jama.298.6.644 Available from: https://www.ncbi.nlm.nih.gov/pubmed/17684187
4IQVIA Institute for Human Data Science. 2019 R&D achievements: New product launches, clinical trial activity, and investments. Available from: www.iqviainstitute.org/2019RandDAchievements (forthcoming April 30, 2020)
5Coalition for Epidemic Preparedness Innovations. COVID-19. Available at: https://cepi.net/covid-19/
6Abbott. Detect COVID-19 in as little as 5 minutes. 2020 Mar 27. Available from: https://www.abbott.com/corpnewsroom/product-and-innovation/detect-covid-19-in-as-little-as-5-minutes.html