Government agencies and companies like Apple have developed COVID-19 screening tools to assess individual risk.
Imagine this scenario: as someone concerned about my COVID-19 risk status, I am going through Apple’s screening. As I scroll through these questions, I come across one that I don’t know how to answer. “In the last 14 days, have you been in an area where COVID-19 is widespread?” A follow up question asks whether I have come within 6 feet of a person that tested positive for COVID-19 for more than ten minutes. I cannot confidently answer these questions. Our team has experienced this, and we are not alone in wanting to understand our COVID-19 risk in more detail.
Local Health Departments are overwhelmed by calls where concerned individuals are asking these same questions, facing the same problem as above. How confidently can callers provide answers to those questions without better tools? One option is for widespread testing, but as COVID-19 continues to spread rapidly and Health Departments do not have enough tests to keep up with demand. Now, and even when there are tests available, health departments would benefit from a dynamic method to prioritize testing and better risk summaries to understand severity of COVID-19 spread within a community.
Furthermore, without the ability to evaluate an individual’s risk, local governments have implemented shelter-in-place orders and have closed services and businesses deemed non-essential like business and health care services. These are decision-makers without any other options - so they are forced to make this difficult decision. Though these measures will decrease the spread of COVID-19, there are broader implications - millions have already applied for unemployment and this number is projected to grow. Continuity of care for populations with other health conditions is severely disrupted.
What it does
To address these challenges, we introduce CommonThread as a local knowledge repository that connects individuals, service providers and health departments to not only increase individual awareness about their own risk of transmitting COVID-19, but also to improve how we do contact tracing for airborne infectious diseases. This would offer local government officials a tool for making softer policy decisions to balance trade-offs between the public’s transmission risk and quality of life.
How we built it
We have a diverse team and treated the challenge timeline as a very short project sprint. We applied a design thinking approach. The first few days we did research into our users and the unique opportunity that contact tracing provides for individuals with mobile phones, businesses, and health department staff. We then defined these needs and identified the key assumptions behind our ideas. From there, we went on to rapidly test our assumptions by deploying a survey across 26 channels. We found that while some users did have privacy concerns, a real-time dynamic assessment tool for COVID-19 risk was desired by all our stakeholders. Some interesting stories evolved from our analysis, and we built these into our prototypes, designed using wire-frames. Although we did not build the system, we thought through what functional requirements were needed as well as the technical support elements required. We considered different use cases and those insights crafted the product we are presenting today.
Challenges we ran into
We need to consider addressing data privacy concerns. This was something that primarily arose in our qualitative testing with individual users. Although individuals were keen on using the product, there was privacy concern. Interestingly, older individuals (45 years and older) were less worried about privacy of their location and contact with others than younger individuals. Additionally, while we were able to test a few of our assumptions via surveys, further testing of assumptions are required, especially with business owners and health department staff as these voices remain underrepresented (N=50) in our first survey analysis. For example, we would love to test the effect of normative messaging or game-ification in the app, to see how this would change underlying individual user behavior.
Accomplishments that we're proud of
We assessed the need for CommonThread based on the following assumptions: 1) Individuals want to know their own risk and the risk in their community for contracting and transmitting COVID-19 2) Business owners would want to re-open their businesses if there was a way to determine customer risk status upon entry 3) Health departments would use real-time contact tracing data to prioritize testing and inform policy decisions To test these, we conducted a survey that garnered over 250 responses and as you can see from the graphs shown, our initial assumptions were confirmed.
Additionally, we are proud of our solution design, which has been approved by a Design Thinking expert, Dr. Matthew Trowbridge. Our pilot approach and implementation plan describes how we can add value immediately and how we improve the product over time in phases.
What we learned
From our survey results, three-quarters of our business owners surveyed had to close or greatly reduce hours (N=10/20 closed; N=4/20 reduced hours). Still, 70% of business owners kept all their staff. From this we can infer that our individual survey answers may be biased towards tele-workers, as there must be a significant chunk of the service population that is unemployed or underemployed. Most of the businesses that are currently closed would let in people that are low-risk of having COVID (N=8/10) and they would use a free tool that quickly assess impact (avg 4/5). These people are concerned about their business and person health. Comments from them: “As a massage therapist in a private practice I am concerned I will not be able to return to this career.” and “I closed my facility and paid my employees two weeks pay. I will be applying for government grants. I am 69 so I am keeping a low profile.” and “Ability to know if a person has developed the disease and is already immune would be fundamental in re-opening the business.” The businesses that wouldn’t let anyone in, even if there was a testing tool (N = 4/20) are largely run by older individuals, and half of them haven’t had to change their hours because of COVID. We therefore think that this tool is mostly useful for businesses that had to close due to coronavirus, for people that run e-businesses or other essential stores, this tool could possibly not be worth the risk. This warrants further study in the coming weeks.
What's next for Common Thread
What value do we add today? We have identified a knowledge gap in the existing COVID-19 screening questionnaires. Respondents cannot confidently say how low or high their risk for transmission is, they cannot confidently describe the risk within their community. It is clear that more information is needed for the screen-ees to answer. We propose adding screening questions in current COVID-19 screening tools. These questions would inquire about the observed behavior of people and businesses that individuals have interacted with to facilitate risk estimation.
How do we add value in 2-4 weeks? An agile sprint is typically 2 weeks. To be realistic, based on our team’s combined technical and functional expertise, it will take 2 sprints to implement a pilot system. We will target piloting in a city that expresses interest. We recognize that the COVID-19 crisis is a very fluid situation. For this reason, we plan to make risk estimation a very configurable process in our application.
We also think forward to future crises that cannot be foreseen. We don’t know when or where the next one will strike. But, we can be better prepared. The long-term vision is a high reliability approach that looks at a crisis like this as a process, not a singular event. Through this we could emphasize prevention, anticipation, and resilience at the organization level. If a new natural highly infectious disease emerges, location-based contact tracing will, once again, be critical. Thus, we believe our application has long term value. We believe there is a better alternative to shutting everything down if risk can be properly assessed and communicated. CommonThread is the way forward to bring outdated contact tracing methods into 2020.
Who we look to collaborate with
Local and state government leadership and health departments. Feedback from federal agencies (e.g. CDC, NIH)
University Researchers and Faculty (e.g. medicine, engineering, public health, etc.)
Healthcare professionals Technology Companies (e.g. Amazon, Google, Apple, Microsoft, etc.)