The Promises and Perils of Using Collective Data to Monitor COVID-19

Listen to this story:

This post is a part of our Bioethics in the News seriesBioethics-in-the-News-logo

By Laura Cabrera, PhD

In a state of public health emergency, such as the one brought on by COVID-19, different countries have invoked extra powers to help mitigate the public health threat. These special powers would under normal circumstances be considered infringements on our liberty and privacy. A recent Wired article addressed that big tech companies like Google and Facebook are having discussions with the White House to share collective data on people’s movement during the current pandemic. For example, using phone location data or private social media posts to help track whether people are remaining at home and keeping a safe distance to stem the outbreak, and to measure the effectiveness of calls for social distancing. In the U.S., the government would generally need to obtain a user’s permission or a court order to acquire that type of user data from Google and Facebook. But as mentioned above, the government has broader powers in an emergency.

Obtaining this data could help governments prepare for the coming weeks of this public health emergency. For example, smart phone location data analysis from the New York Times has shed light on the disparities regarding which groups can afford to stay home limiting their exposure to the coronavirus. This is certainly useful to better understand the spread of the disease in different areas and across different socioeconomic groups. Facebook is working with Chapman University and other collaborators to develop maps that show how people are moving between areas that are hotspots of COVID-19 cases and areas that are not, and such maps could be useful in understanding the spread of the disease. Announced in a news release this month, Apple and Google have launched a joint effort to help governments and health agencies reduce the spread of the virus by using application programming interfaces and operating system-level technology to assist in enabling “contact tracing.”

brain-magnifying-glass
Image description: an illustrated image of a pink brain coming out of the top of a person’s head; a magnifying glass with a brown handle, silver rim, and blue lens is above the brain as if looking into it.  The background is light green. Image source: artwork from vecteezy.com.

While this sounds promising, one of the main obstacles has to do with concerns over the privacy of users whose data might be handed over by the companies. It would be unprecedented for the government to openly mine user movement data on this scale. To add to the issue, the current state of affairs where many more people now rely on digital tools to work or attend classes remotely, as well as to stay connected with family and friends, makes the amount and type of data gathered richer. However, as pointed out in a New York Times editorial, we should not sacrifice our privacy as a result of this pandemic.

Another relevant concern related to the use of collective data is government surveillance. For example, the use of mobile data to track the movement of individual coronavirus patients in China or South Korea can be seen as more controversial uses of the collected data.

It is certain that during this challenging time, data sharing and collaboration between academia, governments, civil society and the private sector is key to monitor, understand and help mitigate this pandemic. However, without rules for how companies should anonymize the data, and without clear limits on the type of data they can collect and how the data could be used and kept secure by researchers and governments, the perils might be greater than the promises. Furthermore, we need a clear path for what happens after all of this is over. For example, people should be given the option to delete user profiles they created as part of new work and school arrangements.

Given past scandals around privacy and transparency surrounding these big tech companies (in addition to the several scandals with the current government administration), it is hard to trust that the idea would be to only gather aggregate trends, and that they would not collect any identifying information about users, or track people over long periods beyond the scope of the pandemic.

Civil groups and academics have discussed the need to protect civil liberties and public trust, arguing for the need to identify best practices to maintain responsible data collection, processing, and use at a global scale.

The following are some of the key ideas that have been discussed:

  • In a public health emergency like the one we are living, some privacy intrusions might be warranted, but they need to be proportionate. For example, it would not be proportionate to gather 10 years of travel history of all individuals for the type of two-week incubation disease we are dealing with.
  • This type of government and big tech company partnership needs to have a clear expiration date, as there is a hazard for improper surveillance that could come with continuation of data gathering after the crisis is over. Given the historical precedents on how life-saving programs used in a state of emergency have continued after the state of emergency was resolved, we as a society need to be very cautious with how to ensure that such extraordinary measures do not become permanent fixtures in the landscape of government intrusions into daily life.
  • The collection of data should be based on science, and without bias based on nationality, ethnicity, religion, or race (unlike bias present in other government containment efforts of the past).
  • There is a need to be transparent with the public about any government use of “big tech data” and provide detailed information on items such as the information being gathered, the retention period, tools used, and the ways in which these guide public health decisions.
  • Finally, if the government seeks to limit a person’s rights based on the data gathered, the person should have the opportunity to challenge those conclusions and limits.

A few weeks ago the European Data Protection Board issued a statement on the importance of protecting personal data when used in the fight against COVID-19. The statement highlighted specific articles in the General Data Protection Regulation legislation. For example, Article 9 mentions that processing of personal data “for reasons of public interest in the area of public health, such as protecting against serious cross-border threats to health” is allowed, provided such processing is proportionate to the aims pursued. In the U.S. we are far from having such a framework to start discussing data collection, sharing, and use under the current circumstances.

There is no doubt as to potential public health benefits associated with analysis of such data and surveillance. For example, the utility of identifying individuals who have traveled to hotspot areas, or tracing and isolating contacts of those infected. However, without a clear framework on how digital data collection companies will address privacy and surveillance concerns, the more cautious we should be about access to other areas of our life, access that would also be shared with governments. Without due caution, not only will public trust continue to be undermined, but additionally people will be less likely to follow public health advice or recommendations, leading to even worse public health consequences.

Laura Cabrera photoLaura Cabrera, PhD, is an Assistant Professor in the Center for Ethics and Humanities in the Life Sciences and the Department of Translational Neuroscience at Michigan State University.

Join the discussion! Your comments and responses to this commentary are welcomed. The author will respond to all comments made by Thursday, May 7, 2020. With your participation, we hope to create discussions rich with insights from diverse perspectives.

Article narration by Liz McDaniel, Communications Assistant, Center for Ethics.

You must provide your name and email address to leave a comment. Your email address will not be made public.

More Bioethics in the News from Dr. Cabrera: Should we trust giant tech companies and entrepreneurs with reading our brains?; Should we improve our memory with direct brain stimulation?Can brain scans spot criminal intent?Forgetting about fear: A neuroethics perspective

Click through to view references

Can Big Data and AI Improve End-of-Life Care?

Bioethics in the News logoThis post is a part of our Bioethics in the News series

By Tom Tomlinson, PhD

A recently reported study claims to more accurately predict how much longer patients will live. Researchers at Stanford University assigned a neural network computer the task of training itself to develop an artificial intelligence model that would predict if a patient would die within 3-12 months of any given date. The computer trained on the EMR records of 177,011 Stanford patients, 12,587 of whom had a recorded date of death. The model was validated and tested on another 44,273 patient records. You can find the wonky details here.

The model can predict with 90% accuracy whether a patient will die within the window.

Now this is a lot better than individual physicians typically do. It’s not just that such predictions are fraught with uncertainty, given how many complex, interacting factors are at work that only a computer can handle. If uncertainty were the only factor, one would expect physicians’ prognostic errors to be randomly distributed. But they are not. Clinicians overwhelmingly err on the optimistic side, so the pessimists among them turn out to be right more often.

The study takes accurately predicting death to be a straightforwardly useful thing. It gives patients, families and clinicians more reliable, trustworthy information that is of momentous significance, better informing critical questions. Will I be around for my birthday? Is it time to get palliative or hospice care involved?

The study’s authors are particularly hopeful that the use of this model will prompt more timely use of palliative care services, and discourage overuse of chemotherapy, hospitalization, and admission to intensive care units in the last months of life—all well-documented problems in the care of terminally ill people, especially those dying of cancer. So this is a potentially very significant use of “big data” AI research methods to address major challenges in end-of-life care.

But making real progress toward these goals will take a lot more than this model can deliver.

36302888621_ad3b5fc2e1_h
Image description: A graphic on a blue gradient background shows the silhouette of an individual in the foreground containing colorful computer motherboard graphics. In the background are silhouettes of twelve more individuals standing in a line and containing black and white computer motherboard graphics. Image source: Maziani Sabudin/Flickr Creative Commons.

The first question is how it could inform decisions about what to do next. The limitation here is that the model uses events from my medical history occurring prior to the time it’s asked to predict my survival. Perhaps the decision I’m facing is whether to go for another round of chemotherapy for metastatic cancer; or whether instead to enter a Phase 3 clinical trial for a new therapeutic agent. The question (one might think) is what each option will add to my life expectancy.

Now if the training database had some number of patients who took that particular chemotherapy option, then that factor would have somehow been accounted for when the computer built the model. Assuming the model reliably predicted the mortality of those earlier patients, all we’d need to do is add that factor to my medical record as a hypothetical, run the model again, and see whether the prognosis changed.

But is there something about the chemotherapy being offered that is different than the regimens on which the computer trained? Then the model will not be able to assess the significance of that difference for the patient’s survival. Obviously, this limitation will be even more radical for the experimental treatment option. So in the individual case, the model’s helpfulness in making prospective treatment decisions could be quite limited. It would have to be supplemented, or even supplanted, by old-fashioned clinical judgment, or alternative algorithmic prognostic tools.

This may be one reason the study authors imagine a different use: identify patients with 3-12 months life expectancy and refer them for a palliative care consultation. The idea is to push against the tendency already noted for physicians to wait too long in making palliative care or hospice referrals. Assuming the model is running all the time in the background, it could trigger an alert to the attending physician, or even an automatic palliative care referral for all those the model flagged.

Now, in my ethics consultation experience, getting an appropriate palliative care or hospice referral only one month preceding death would be a stunning accomplishment, let alone three months prior. But the key word here is “appropriate,” since the need for palliative care is not dictated by life-expectancy alone, but more importantly, by symptoms. Not every patient with a projected life expectancy between 3 and 12 months will be suffering from symptoms requiring palliative care expertise to manage. Automatic referrals requiring palliative care evaluations could overwhelm thinly-staffed palliative care services, drawing time and resources away from patients in greater need.

Part of the problem here is the imprecision of the model, and the effects this may have on patient and provider acceptance of the results. A 90% chance of death within 3-12 months sounds ominous, but it leaves plenty of wiggle-room for unrealistic optimism: lots of patients will be confident that they are going to fall at the further end of that range, or that they will be among the 10% of cases the model got wrong altogether. And it’s not just patients who will be so affected. Their treating physicians will also be reluctant to conclude that there is nothing left to do, and that everything they did to the patient before has been in vain. Patients aren’t the only ones prone to denial.

And the nature of the AI-driven prognosis will make it more difficult to respond to patient skepticism with an explanation anyone can understand. As the authors point out, all we really know is that the model can predict within some range of probability. We don’t know why or how it’s done so. The best we can do is remove a feature of interest from the data (e.g., time since diagnosis), rerun the model, and see what effect it has on the probability for the patient’s prognosis. But the model offers no reasons to explain why there was a change, or why it was of any particular magnitude. The workings of Artificial Intelligence, in other words, are not always intelligible. Acceptable explanations will still be left to the clinician and their patient.

Tom Tomlinson photoTom Tomlinson, PhD, is Director and Professor in the Center for Ethics and Humanities in the Life Sciences, College of Human Medicine, and Professor in the Department of Philosophy at Michigan State University.

Join the discussion! Your comments and responses to this commentary are welcomed. The author will respond to all comments made by Thursday, March 8, 2018. With your participation, we hope to create discussions rich with insights from diverse perspectives.

You must provide your name and email address to leave a comment. Your email address will not be made public.

Click through to view references

Trust and the Learning Health System

bbag-icon-decEvent Flyer

Data sharing on a large scale is integral to emerging national initiatives such as learning health systems and precision medicine. Realizing the vision of learning health systems, “in which knowledge generation is so embedded into the core of the practice of medicine that it is a natural outgrowth and product of the healthcare delivery process and leads to continual improvement in care” requires a “trust fabric” to integrate policy and practice in health care, public health, and research. However, as increased data sharing stretches the currently disjointed regulatory and policy environment, the texture and resilience of this trust fabric will be challenged in its capacity to protect the public and its vulnerable populations, and to assure data will be used in ways that reflect societal values. What will it take to trust the health system with all that information? This presentation will examine these tensions and dynamics. Based on preliminary data from the clinic and the community, Dr. Platt will discuss a proposed a framework for trust to guide decision-making for local, state, and national learning health systems.

nov9-bbagJoin us for Dr. Jodyn Platt’s lecture on Wednesday, November 9, 2016 from noon until 1 pm in person or online.

Jodyn Platt, MPH, PhD, is an Assistant Professor trained in medical sociology and health policy. Her research currently focuses on informed consent in cancer and genomic studies, and the Ethical, Legal, and Social Implications (ELSI) of public health genetics, newborn screening, and learning health systems. She is interested in understanding what makes learning health systems trusted and the pathways for achieving and sustaining trust through community engagement using qualitative and survey methods.

In person: This lecture will take place in C102 East Fee Hall on MSU’s East Lansing campus. Feel free to bring your lunch! Beverages and light snacks will be provided.

Online: Here are some instructions for your first time joining the webinar, or if you have attended or viewed them before, go to the meeting!

Can’t make it? All webinars are recorded! Visit our archive of recorded lectures. To receive reminders before each webinar, please subscribe to our mailing list.

Announcing the Fall 2016 Bioethics Brownbag & Webinar Series

bbag-iconThe Center for Ethics and Humanities in the Life Sciences at Michigan State University is proud to announce the 2016-2017 Bioethics Brownbag & Webinar Series, featuring a wide variety of topics from under the bioethics umbrella. The fall series will begin on September 28, 2016, and you can attend the lecture in person or watch live online. Information about the fall series is listed below, and you can visit our website for more details, including the full description and speaker bio for each event.

Fall 2016 Series Flyer

sept28-bbagEthics and Children with Differences in Sex Development and Gender Nonconformity
When should society constrain clinicians from intervening in these contentious arenas?
Wednesday, September 28, 2016
Joel E. Frader, MD, MA, is a Professor of Pediatrics and Professor of Bioethics and Medical Humanities at Northwestern University, and Medical Director of Bridges Pediatric Palliative Care Program at Lurie Children’s Hospital of Chicago.

oct19-bbagChoosing to Test: Dr. A. P. Satterthwaite and the First Birth Control Pill Clinical Trials in Humacao, Puerto Rico
How did Adaline Pendleton Satterthwaite, an obstetrician-gynecologist (OB-GYN) working at a Protestant mission hospital in Puerto Rico, become one of the key architects of the first birth control pill?
Wednesday, October 19, 2016
Kathryn Lankford is a Doctoral Student in the Department of History at Michigan State University.

nov9-bbagTrust and the Learning Health System
What will it take to trust the health system with all that information?
Wednesday, November 9, 2016
Jodyn Platt, MPH, PhD, is an Assistant Professor in the Division of Learning and Knowledge Systems in the Department of Learning Health Sciences at the University of Michigan Medical School.

In person: These lectures will take place in C102 (Patenge Room) East Fee Hall on MSU’s East Lansing campus. Feel free to bring your lunch! Beverages and light snacks will be provided.

Online: Here are some instructions for your first time joining the webinar, or if you have attended or viewed them before, go to the meeting!

Can’t make it? Every lecture is recorded and posted for viewing in our archive. If you’d like to receive a reminder before each lecture, please subscribe to our mailing list.