The fundamental ethical, legal, and policy challenge of public health is balancing public and individual interests, often conceptualized as the conflict between utilitarianism and libertarianism. During the COVID-19 (coronavirus disease 2019) pandemic, this struggle has involved the imposition of extraordinary levels of government-mandated social distancing to protect public health followed by impassioned efforts to lessen these constraints in the interests of individual liberty and economic renewal.
The same type of conflict exists between individual interests in health privacy and public health interests in the collection, use, and disclosure of health information. The COVID-19 pandemic presents these issues in a unique way because it is the most deadly pandemic in more than a century and new technologies permit an unprecedented level of information collection, aggregation, analysis, and dissemination.
In deciding whether to use certain health information technology in a pandemic, policymakers should analyze and apply the following criteria, which have been derived from principles commonly cited in the public health ethics literature related to public health powers generally and applied to privacy: (1) necessity and effectiveness, (2) proportionality and minimal infringement, (3) purpose limitations, and (4) justice.
No public health intervention should be introduced without clear evidence of its necessity and effectiveness. However, faced with a novel, lethal pathogen and with no vaccine or highly effective treatment available, clinicians have felt compelled to adopt measures without persuasive evidence, and public health officials have faced similar pressures to expand information-gathering techniques.
During the severe acute respiratory syndrome (SARS) epidemic of 2003, the World Health Organization recommended exit and entrance screening of international travelers. Unproven thermal screening was used for maritime crews, passengers on cruise ships and ferries, and land border crossings. Canada used thermal screening at the airports in Toronto and Vancouver and 2.4 million passengers were screened, but only 832 passengers were identified for further evaluation and none were determined to have SARS.1
This experience demonstrated that thermal screening for coronavirus, at least in this setting, does not work. Yet, 17 years later, still lacking evidence of its effectiveness,2 thermal screening for SARS-CoV-2 (severe acute respiratory syndrome coronavirus 2), often accompanied by oral questioning about symptoms and exposures, is widely used in the United States, perhaps because it is fast, cheap, and nonintrusive and because it gives the impression that something is being done to protect people. Neither the thermal-screening procedure nor the data it generates pose a significant threat to privacy, but the stakes are too high and the support for public health interventions too tenuous to employ measures whose main value is symbolic.
Public health interventions should be proportional to the risk presented. In a pandemic, public health measures should be granted wide leeway, but not every public health information plan is necessary or can be accomplished without significant intrusion. “Data minimization” seeks to limit the amount of data collected and used to the minimum necessary. Other relevant data protection principles provide that information ought to be collected, maintained, and disclosed in the least identifiable form consistent with the intended use; data should be retained for the minimum amount of time necessary; and security measures should prohibit access by unauthorized individuals.
Because researchers and clinicians are still learning about COVID-19, it is unknown how much individual health information is needed, for how long, and in what form. Although one approach would be to collect and retain all possible health information, it should not take precedence over proportionality and minimal infringement.
Data collected for a specific purpose should be used for only that purpose and should not be “repurposed” for other uses without the consent of the data sources or a compelling public justification. Purpose limitations may conflict with Big Data analytics and other health surveillance technologies based on machine learning and algorithms identifying associations from disparate data sets.
One way of assessing the appropriateness of compiling various data sets is by considering whether the data were collected for a public health purpose or are traditional public health measures.3 For example, public health laws in every state require the reporting of specific conditions to public health agencies. Similarly, contact tracing is an established public health tool used to control sexually transmitted infections and other diseases. The public is likely to view the use of these traditional public health methods as acceptable and ethical to use in a public health emergency.
It is more problematic when data are repurposed for public health from social media posts, geolocation information, sensor data, cell phone records, health check apps, credit card records, and proximity data generated by apps on mobile devices. Some of these nontraditional data sources have been valuable in other countries during the pandemic,4,5 and they should not be prohibited summarily, but they should not be presumed relevant and ethically acceptable. Furthermore, collecting data that can be used for surveillance or law enforcement is likely to be viewed with suspicion by the public.
The pandemic is challenging societal notions of justice. Black Americans are 2.4 times more likely to die of COVID-19 than are White Americans and 2.2 times more likely to die than are Asians or Latinos.6 Discrepancies in risk are even greater in some locations, such as on Indian reservations.6 Justice demands assessing the relative burdens and benefits of all public health policies, including neutral policies with a discriminatory impact. For instance, the burden of home confinement is greater in an urban apartment with several family members than in less dense housing, the burden of discontinuing public transportation falls more harshly on those without cars, and the burden of closing nonessential businesses is more onerous for low-wage service workers who cannot telecommute and who have modest savings.
Other public health interventions may adversely affect various individuals and groups, such as people with disabilities or preexisting health conditions, individuals experiencing homelessness, senior citizens, people with cognitive impairments, and immigrants who lack English fluency.
Health privacy is rarely included in discussions of health equity, but it is essential to fairly allocate the burdens and benefits of health information policies. For example, if contact tracing using mobile apps is deemed effective and acceptable, what about lower-income people, who are most at risk for COVID-19 but less likely to have smartphones? Is the release of public health data more likely to result in discrimination against certain individuals in employment, housing, or access to health care? Finally, will any health benefits derived from collecting and using health information extend to all members of society?
Ethical issues, including health privacy, are especially important during major disease outbreaks.7 Health information policies should be continually evaluated using criteria that support public policy. After a public health emergency ends, measures permitting intrusive surveillance should not be continued without a careful reevaluation.
See also the AJPH COVID-19 section, pp.
CONFLICTS OF INTEREST
The author has no conflicts of interest to declare.
