Photo by Nick Bolton | Unsplash

Is anything getting better? Is this social distancing working? When can we go back outside?

During the COVID-19 pandemic, I hear my family, friends, and colleagues struggling with these kinds of questions. Perfect answers do not exist for any of them. However, with the right data and a method of displaying it, people can get a better understanding of how the pandemic is evolving.

Shewhart Control Charts offer a way for people to use their local data to gain insights into the trajectory of the COVID-19 epidemic in their community, state, region or country. Also known as Statistical Process Control (SPC) charts, IHI has published the results of the SPC analysis of all 55 states and territories as a dashboard, and will update them daily.

As most in the improvement field know, SPC charts provide a way to help distinguish between random variation and non-random variation. This is especially important as measures related to COVID-19 fluctuate daily. We want to know are things getting better or worse. What is noise? What is a signal?

The New York Times GitHub site provides an opportunity to apply SPC methods to look at COVID-19 reported mortality and view the pandemic across the United States. This approach signals when states are starting the early growth phase of COVID-19 reported mortality, when the shocking exponential trajectory takes off, and when the data signals a deceleration and downward trajectory. The SPC method uses data from the local setting, making no assumptions based on trajectories in other settings which may differ in terms of population density, climate, economic status, and local policies. It offers an additional tool for decision-makers keen to understand whether efforts to reduce the spread of COVID-19 are resulting in improvement.

Looking at the data across the United States on the IHI dashboard, as of April 15, 2020, every state and territory has experienced at least one COVID-19 related death. Only 14 states, including Arkansas, West Virginia and South Dakota, remain in the early pre-growth phase, with relatively small numbers of cases (zero, one, or two) per day. The remaining states have already moved through this phase and into the exponential growth phases. Some, such as Rhode Island and Iowa, are only a week into this phase, before the curve bends more alarmingly as it is starting to in Kentucky, Maryland and Texas. Leaders and the public can use SPC methods to visualize where they are on the trajectory, and where, if no changes are made, they are likely to move towards.

There are some signs of hope. The SPC method shows New York has been below the lower limits for over three days, suggesting COVID-19 mortality may have peaked. Leaders and the public in New York need to decide whether to maintain or lift the approaches including social distancing and restricted public movement that occurred before the growth slowed.

From an improvement perspective, maintaining the changes seems most likely to maintain the deceleration in deaths. Lifting the restrictions may risk returning to accelerated growth. Moreover, IHI’s experience working with communities in the Triple Aim, suggests leaders and the public need to include data related to factors including the local economic and social impact of maintaining or lifting restrictions. Social data is notoriously piecemeal or absent and leads us to ask how the impact of COVID-19 differs across, say, race, ethnicity, gender identity or sexual orientation, disability, and economic status.

The aggregate data is clear in our current system: COVID-19 mortality will rise, reaching into every part of our society, and unless that system is disrupted, it will not stop. There are opportunities to learn from the variation in the trajectory of states through the COVID-19 pandemic. Those states still in the earlier phases can collaborate with those in later phases to understand what changes resulted in improvement, then work with their populations to identify and introduce changes appropriate for them.

Gareth Parry, MSc, PhD, is a Senior Scientist at the Institute for Healthcare Improvement

You may also be interested in:

More COVID-19 Guidance and Resources