Three Steps to Data Fluency During COVID-19

The longer we live with the COVID-19, the more data there are to navigate. This is a mixed blessing: there has never been more information available, and it can be harder than ever to know which data matter most.

We want teams to be fluent in COVID-19 data because we know this enables them to focus on the small set of data points that can help them answer the questions they are facing.

Data fluency is a required skill during COVID-19.

The new normal is here. It is defined by its fluidity, and our success will be defined by our adaptability. There will be ups and downs as conditions change, and these cycles will be different from one part of the country to another. Where we are able to resume behaviors, the transition back to them will be gradual.

For organizations, adapting and thriving during COVID-19 will require understanding and responding to local data in a timely manner. Knowing the conditions in your area is essential to maintaining a safe and healthy team.

For employees, understanding local data is essential to knowing what questions to ask and what answers to expect to maximize your safety. Ideally, data fluency prepares employees to be active and engaged participants in conversations on workplace safety. Failing that, it empowers them to be educated and critical consumers of the decisions made by their employer.

We coach our clients through three steps of data fluency during COVID-19.

Step 1: Understand key pieces of data and where to reliably source them.

It is important to start from a broad awareness of what data are available and widely used. These nine measures capture most of what is needed to answer questions about safety during COVID-19:

1. Symptoms are measured by the number of reported COVID-19 or influenza-like illness (ILI) symptoms reported for a period or the rate of change in reported symptoms over time. Recommended source: Centers for Disease Control and Prevention.
2. Cases are measured by the total number of COVID-19 cases (confirmed or suspected), the rate of change in cases over time, or the projected number of cases. Recommended sources: The COVID Tracking Project, Johns Hopkins University Coronavirus Resource Center, and The New York Times.
3. Transmission is measured by Rt, which is the rate at which one infected individual results in new cases. Recommended source: Rt.live.
4. Hospitalizations are measured by the total or projected number of hospitalizations or ICU hospitalizations in an area. Recommended source: The COVID Tracking Project.
5. Hospital capacity is measured by the number or percent of ICU beds available. Recommended sources: Centers for Disease Control and Prevention and COVID Care Map.
6. Positive virological tests are measured by the number of tests which confirm active infections or the percentage of these tests relative to a target. Recommended sources: The COVID Tracking Project, Johns Hopkins University Coronavirus Resource Center, The New York Times, and Test and Trace.
7. Positive serological tests, also known as antibody tests, are measured by the number of tests which indicate prior infections or the percentage of these tests relative to a target. Recommended source: Centers for Disease Control and Prevention (in process).
8. Contact tracing is measured by the number of contacts traced or the percentage of contacts traced relative to a target. Recommended source: Test and Trace.
9. Mortality is measured by the total number of deaths attributed to Covid-19 or the rate of change in deaths over time. Recommended sources: The COVID Tracking Project, Johns Hopkins University Coronavirus Resource Center, The New York Times.

Additionally, there are organizations doing a wonderful job of aggregating data from these sources. We recommend and regularly follow The COVID Tracking Project and COVID Exit Strategy for a broad range of state-level data, as well as COVID Act Now and the Harvard Global Health Institute’s Pandemics Explained website for state- and county-level data.

For a deeper dive into the benefits and limitations of some of these measures we highly recommend this recent piece from ProPublica.

Step 2: Know your local resources.

Each state maintains its own website and, in many cases, dashboard for COVID-19 data. All states provide data at the state and county level, with some going further to report local data by zip code or town.

It would be ideal if there was some standardization of state data reporting. Unfortunately, the data reported within these resources can vary widely from state to state.

We recommend that everyone familiarize themselves with the data resources for their state — what data are reported, how frequently, and in what format. We maintain a free index of state data resources for COVID-19.

Step 3: Know what questions you need to answer.

We often see individuals and organizations struggle when they feel like they have to make decisions which account for all of these measures. In most cases, a small set is sufficient.

The key is to start from a well-defined and focused question. Some examples: What does spread look like in my area? How are testing and tracing being used to understand prevalence in my area? What does capacity look like in my local health systems?

Most often, we seek to help clients understand the risk of infection spread in their communities. We use three core measures to understand this risk:

1. New Cases: We track a 7-day rolling average of the increase in new cases of COVID-19. This provides an indicator of recent growth that buffers against day-to-day variation.
2. Positive Tests: We track the percentage of virological tests that are positive for COVID-19. The World Health Organization and Centers for Disease Control and Prevention recommend that this rate be below 5%. Rates exceeding 5% strongly suggest that testing is insufficient and producing an incomplete indication of infections.
3. Transmission: We track Rt, or the estimated average number of infections that result from a single infected individual. This provides an indication of the current level of spread. An Rt value over 1 indicates that continued growth in new cases is likely.

Cases tell us about the recent growth that has been measured. Positive tests tell us how complete that measure is by giving us an indication of what portion of the cases in a community are being detected. Transmission gives us an indication of whether and how that growth may continue. Collectively, these measures provide a strong understanding of the risk of continued spread in a community.

We regularly update these data and make them freely available on this site.

Don’t get lost in or discouraged by the glut of data that are available on COVID-19. Focus on the data and sources that are most relevant to your concerns, and make it a habit to check them regularly.