The Fierce Urgency Of Now: Closing Glaring Gaps In US Surveillance Data On COVID-19
Where are the data on COVID-19 to understand who in the US population is being tested, who is ill, and who is dying? The sole data being reported by the US Centers for Disease Control and Prevention (CDC) is for the total population only, and just for the country as a whole and by state. Worse: case counts at the county level—for just a small subset of the 3242 US counties—are available not from the CDC, but from the websites of a private university (Johns Hopkins) using data credited to the CDC and those of several newspapers and volunteer groups frantically trying to track down data. All in this reporting hodgepodge are grappling with inconsistencies and gaps in state reporting on COVID-19, including some states publicly reporting only the number of confirmed cases without data on the number of persons tested (e.g., CA, NY, WA), or not reporting on the number of persons hospitalized.
Nor are any data available—at the national state, county, city, neighborhood, or health system levels—to enable monitoring and interpretation of testing patterns (including who is being tested), likelihood of positive tests (which will be related to testing strategy), or mortality (which is related to definitions and ascertainment of what is COVID-19 related) – especially in relation to core sociodemographic variables, such as age, race/ethnicity, sex/gender, and socioeconomic position. Health equity is literally off the map, made invisible because data to document inequities are unavailable, even as journalism and social media vividly attest to the risks of those who do not have the luxury to shelter safely at home and who cannot afford the economic disruption.
This is unacceptable and threatens prudent public health action.
The Inadequacy Of Current Data
It is insufficient to ask simply whether the virus is or is not present. Social data about who is infected are crucial for responding to needs now and will allow for better estimation of the likely spread and impact of COVID-19, the toll of which will be measured not only in deaths but also in the second-order, socially disparate spill-over effects on people’s economic well-being and safety. Real-time fast journalistic reporting and advocacy groups in the US and other countries are pointing to the critical importance of racial/ethnic, economic, and gender inequities to shaping COVID-19 risks. In the past week, calls for data on COVID-19 by race/ethnicity have been issued by leading politicians, including Senator Elizabeth Warren and Congresswoman Ayana Pressley, the Congressional Black Caucus, the National Lawyers Committee for Civil Rights Under the Law, and by journalists. Why aren’t the public health data documenting these risks available?
Granted, the CDC has developed a detailed intake form, valid through April 23, 2020, which does include data on a several social variables (race/ethnicity, sex/gender, age, county, and state), along with numerous questions about travel, clinical presentation, and respiratory diagnostic testing. Our understanding, however, is that this form (which the CDC estimated takes 30 minutes to complete – difficult when cases number in the thousands) is not being used and that a short form may be in development to replace it.
The New York City Department of Health and Mental Hygiene laudably has just produced one of the few neighborhood-level maps to show the percent of patients testing positive for COVID-19. This map vividly documents differential rates by neighborhood and makes clear that the distribution of positive tests is far from random, with higher positivity rates in lower-income areas. But without any social data, the map is also completely uninterpretable. Do the results shown reflect that public hospitals are reporting more sick people? That wealthy people are getting screened who are not symptomatic and less likely infected? There is no way to know. The spatial patterns presented cry out for explication and informed intervention. But the available data cannot provide the information needed.
Recognition of the critical importance of societal determinants of health is now commonplace in public health, globally by the World Health Organization and within US health agencies as well. It is not a mystery that social inequalities become embodied as health inequities. Failing to collect and report critical social data necessary to mitigate and prevent COVID-19 will hamper efforts to control the first wave and to handle the uncertain future ahead.
It might perhaps be understandable that data collection in the rapid exponential first phase of the epidemic has been imperfect, although adequately funded preparedness planning could have anticipated and addressed many problems. Going forward, the federal government must provide both leadership and sufficient funding to ensure there is a strong coordinated response so that the relevant social and clinical data are collected and swiftly made nationally available at the national, state, and local levels.
A Call To Action: Improving Collection And Reporting Of COVID-19 Sociodemographic Data
We accordingly urge rapid adoption of a uniform short digital form for COVID-19 testing and surveillance that is up to the task. The federal government should mandate that all testing data are provided to the CDC, in real time, and that data are publicly reported, in real time, in relation to total cases and stratified by race/ethnicity, sex/gender, age, educational level, at the national, state, county, and Zip Code levels. Federal funding is essential for this work, which necessarily will be carried out by a combination of state and local health departments and the CDC.
We emphasize this digital form should be used nationally by every private, academic, and public laboratory doing COVID-19 testing. Additionally, the expedient short list of social variables we have enumerated—already available in death certificates (i.e., age, race/ethnicity, sex/gender, educational level, and Zip Code)—must also be included in COVID-19 hospital intake forms and in population-based seroprevalence surveys when they become feasible. A minimal goal is to have all testing, hospital, and mortality data for COVID-19 publicly reported for both the total population and by these social variables, minimally at the national, state, and county level.
Of course, we have visions of what more comprehensive social data for COVID-19 monitoring would entail, relevant to both modeling the course of the epidemic and policy impacts. But we recognize the fierce urgency of getting core basic data now, so that communities and health professionals can plan and do their best to control and mitigate the community spread now well-underway across the US. Examples of additional social variables, all supported by growing accounts of inequitable risks and burdens, would pertain to employment status, housing status (e.g., private home, incarcerated, other institution, homeless), health insurance, income level, food insecurity, transportation access, safety at home (for oneself and one’s children, given abuse that can happen in a context of remanding people to stay at home), and residential address (for more precise geocoding and linkage to neighborhood social and economic data). In a better world, obtaining such data would be recognized as an intervention, with appropriate resources provided at the time of data collection (e.g., referrals to a social worker if domestic violence or child abuse is reported, especially for people who have to self-quarantine at home).
But we are realistic. Grounded in concerns both for population health overall and health equity, and acutely aware of the perils of this pandemic, we assert the time is now for the COVID-19 public health surveillance system to record and publicly share the critical data needed to protect the people’s health and prevent health inequities. Protecting all communities, especially those most harmed by COVID-19 and its social consequences, is imperative.
