Friday, March 06, 2020

Data Driven Management of COVID-19: The Case of Taiwan



COVID-19 has produced substantial challenges to states and other entities with responsibility for the well bring of people.  Much of the news about the way that states and other entities have been undertaking that responsibility has focused on medical measures to detect,control and eventually cure or prevent the disease. But states and other entities have also devoted substantial attention to the economic, societal, political, and cultural consequences of the disease. Especially difficult has been consideration of the measures that might be taken to meet the medical challenges posed by the disease without violating institutional or cultural taboos. Most of these measures revolve around the core strategy, implemented in infinite contextually specific variations--containment. 

Containment, however, is a strategy that does not lend itself well to law.  Except for constructing and devolving powers to contain (person, idea, speech, bodies, processes, and the like),  the space within which containment is to be situated (self-containment at home, hospitals, cruise ships, jail, student dorms, or the like),  and the conditions under which the power to determine and enforcement containment within designated spaces (or with respect to speech, suppressed) may be exercised, containment is by its nature an administrative act.  But it is more than that, containment strategies, including, choices about the holding of mass attendance events (like factory work, classroom instruction, cultural or sports events) do not lend themselves to the sort of discretionary decision making that is the hallmark of administrative practice by individuals. 

That insight suggests that a containment based approach to COVID-19 would have to be data driven to be successful.  But ore than that, it would have to be a data consumptive enterprise, the purpose of which would be to feed an analytics that would help development algorithms for judging  (1) disease risks of individuals and (2) likely consequences of such individual ratings given a variety of contexts in which that would be discovered.

A recent study considered the data based comprehensive approach undertaken by authorities in Taiwan  (C. Jason Wang, MD, PhD; Chun Y. Ng, MBA, MPH; Robert H. Brook, Response to COVID-19 in Taiwan: Big Data Analytics, New Technology, and Proactive Testing, JAMA. Published online March 3, 2020). This point considers that research study in the context of data driven governance within a rule of law state.


The Wang-Ng-Brook study explained:
COVID-19 occurred just before the Lunar New Year during which time millions of Chinese and Taiwanese were expected to travel for the holidays. Taiwan quickly mobilized and instituted specific approaches for case identification, containment, and resource allocation to protect the public health. Taiwan leveraged its national health insurance database and integrated it with its immigration and customs database to begin the creation of big data for analytics; it generated real-time alerts during a clinical visit based on travel history and clinical symptoms to aid case identification. It also used new technology, including QR code scanning and online reporting of travel history and health symptoms to classify travelers’ infectious risks based on flight origin and travel history in the past 14 days. Persons with low risk (no travel to level 3 alert areas) were sent a health declaration border pass via SMS (short message service) messaging to their phones for faster immigration clearance; those with higher risk (recent travel to level 3 alert areas) were quarantined at home and tracked through their mobile phone to ensure that they remained at home during the incubation period.

Moreover, Taiwan enhanced COVID-19 case finding by proactively seeking out patients with severe respiratory symptoms (based on information from the National Health Insurance [NHI] database) who had tested negative for influenza and retested them for COVID-19; 1 was found of 113 cases. The toll-free number 1922 served as a hotline for citizens to report suspicious symptoms or cases in themselves or others; as the disease progressed, this hotline has reached full capacity, so each major city was asked to create its own hotline as an alternative. It is not known how often this hotline has been used. The government addressed the issue of disease stigma and compassion for those affected by providing food, frequent health checks, and encouragement for those under quarantine. This rapid response included hundreds of action items (eTable in the Supplement). (C. Jason Wang, MD, PhD; Chun Y. Ng, MBA, MPH; Robert H. Brook, Response to COVID-19 in Taiwan: Big Data Analytics, New Technology, and Proactive Testing, JAMA. Published online March 3, 2020)
The authors also made available the eTable Supplement in which the action items were described.  These effectively provided the basis both for defining the set of data to be harvested and their application through analytics. The eTable is reproduced below.

The article noted challenges.
 First, real-time public communications were mostly in Mandarin Chinese and sign language. Other than the Taiwan CDC website, there was not enough communication in different languages to non-Taiwanese citizens traveling or residing in Taiwan. Second, while its attention was focused on air travel, Taiwan permitted the docking of the Diamond Princess cruise ship and allowed passengers to disembark in Keelung, near New Taipei City, on January 31, before the ship left for Japan. The ship was subsequently found to have numerous confirmed infections onboard. This created a temporary public panic with concern about community spread. The government published the 50 locations where the cruise ship travelers may have visited and asked citizens who may have been in contact with the tour group to conduct symptom monitoring and self-quarantine if necessary. None were confirmed to have COVID-19 after 14 days had passed. Third, whether the intensive nature of these policies can be maintained until the end of the epidemic and continue to be well received by the public is unclear. (C. Jason Wang, MD, PhD; Chun Y. Ng, MBA, MPH; Robert H. Brook, Response to COVID-19 in Taiwan: Big Data Analytics, New Technology, and Proactive Testing, JAMA. Published online March 3, 2020)
The research reports ought to serve as a starting point for discussion.

First, it suggests that in the context of Taiwan, it is possible to implement a data driven system for the coherent and comprehensive approach by public authorities to the threat of disease.

Second, at the same time it is clear that Taiwan's success was due in large part that the system was already in an advanced stage of creation (to fight SARS of course) at the time COVID-9 exploded on the scene.

Third, it required a substantial effort not just to the conceptualization and construction of data harvesting systems, ad a relevant data set, but also of an analytics from data capable of providing useful conclusion and suggestions.

Fourth, these analytics had to provide a basis for developing quick judgment of consequences--especially with respect to identification of individuals, and with respect to their containment.But that speed an efficiency was perhaps purchased at some cost to the developing principles of privacy and individual rights.  Here a balancing and re-calibration might be necessary--but that re-calibration (as evidenced by the operation of the system) dd not appear to have been undertaken with much engagement by the people or relevant political actors.

Fifth, the issue of calibration becomes more acute in the context of information control.  Here it takes two forms.  First is the fight against fake news and rumor mongering.  As was clear in the situation in Wuhan in December, it is sometimes difficult for officials to discern the difference between courageous acts of individual patriotism and anti-social rumor mongering.  The results of mistakes can be disastrous.  Yet there appears to be little attention paid to those issues.  The second involves the management of official information. Here public transparency and accountability requires a more open textured engagement.  Yet most states, the United States included, tend to view its duty to severely manage its public communication in ways that are sometimes possible appear more valuable to the administrative officials than to the public. (for the U.S., e.g., here). That is especially important in the context of accountability.

Sixth, there appears to be little effort either to protect data integrity, or to manage systems systems for removing erroneous data and correcting consequential errors that are produced by bad data. (Automated Law: The Problem of Data Integrity Moves (if only for a Moment) to Center Stage). And yet, protecting the integrity of the system of data driven management of disease based crisis at the cost of damaging individuals who suffer the consequences of erroneous applicaiton due to bad data, may create contradictions between the principle of systemic efficiency and the organizing principles of states founded on the notion of individual human dignity.

Seventh, the issue of public data base integration will prove the most challenging even as it holds the most transformative potential.  What the researchers noted was that Taiwan integrated its national health insurance database with its immigration and customs database.  Both are public sector data bases tied to large public sector operations.  It was put forward for the proposition that data base integration was useful for developing specific focus analytics--in this case targeting the management of strategies of containment against the spread of a contagious disease. Yet that also suggests issues for the future.  The first is the utility of data harvesting fragmentation by public bodies.  In the context of data driven governance, it remains unclear why a taxonomy of data "ownership" among government agencies dictated by a jurisdictional taxonomy with origins in a prior century ought to extend its dead hand to manage the fundamental building block of control in this century.  That, of course, ought to raise the question of political and normative premises built around these emerging systems of control  Instead, the discourse of law, in the face of this emerging data governance modernity, appears increasingly Luddite in form and character.

Eighth, almost entirely ignored are the issues of private data governance of the type utilized by Taiwan authorities.  In most markets driven political structures, an enormous amount of regulatory authority over the daily lives of people are now delegated to the discretionary authority of private actors.  Though sometimes these delegations are more or less closely aligned to public objectives, for the most part the range of discretionary authority is broad.  Those enterprises are also increasingly driven by analytics derived from fractured data bases that they manage (and sell). To manage these managers appears to be an important element for comprehensive strategies, especially in the face of epidemics. Here one encounters the challenges of rethinking conceptual walls between public and private data, data integrity, and the transparency of the analytics and judgments derived from the employment of these emerging regulatory actions.  



__________









No comments: