Image by Gerd Altmann from Pixabay 

 What every scientist can do during COVID-19 pandemic

Samuel Wang

PhD Student, College of Nursing

University of Utah 

COVID-19 caused the highest unemployment rate in the past 72 years in the US. The financial foundation of many families was damaged. For many families, they need to face losing commercial healthcare insurance and significant economic and food insecurity. According to the US Department of Agriculture, the current food insecurity rate in the US is 22-38%. The number increased more than double compared with the year of 2018. During this challenging time, we need to utilize our resources more accurately, additionally, due to the fluid situation of the COVID-19 infection rate. The resource distribution should also be dynamically moving with the demands. In this case, the social determinants of health (SDOH) data might be the key to a more flexible resource supply.

 What are Social determinants of health data?

According to the CDC, SDOH is the condition in people’s daily life that affect health risk and outcomes. Examples such as income, education level, food insecurity rate, or employment are SDOH. Understanding the impact of a condition on people’s health and how people experience the situation is fundamental to SDOH. Therefore, it is not surprising that the data source of SDOH not only from the healthcare system; the sources are everywhere around us.

A tremendous amount of data generated every day

            According to Datavant, a healthcare data analysis company, the healthcare system generates approximately a zettabyte (a trillion gigabytes) of data annually. And this amount is doubling every two years. That is just the data from the healthcare system. International Data Corporation suggested that the size of the digital universe in 2005 be 130 exabytes (EB). Dramatically, in 2017, the digital universe expended to 16,000 EB or 16 zettabytes. The same organization is now expecting by the year of 2020, the digital universe would expand to 40,000 EB. The speed is like a rocket taking off from Earth, it just gets faster and faster. The data universe keeps expending. Hence, many data scientists believe if we can interpret data correctly, many unanswered health-related questions can be answered. Moreover, the research perspective can enlarge, healthcare policy implementation can be more productive. Ultimately, social resources can distribute more efficiently, and the public health system can prevent the worse of the pandemic.

Image by Wynn Pointaux from Pixabay 

Social determinants of health data is from everywhere

The healthcare system collected many different forms of data—clinical notes entered by the healthcare providers, standard labs data that healthcare providers ordered, and insurance claim data. I know this is getting a little overwhelming. Still, all these forms of data are only an example of a small subset of healthcare data from one patient and from one clinic. Imagine how much information we can collect from multiple sites and patients. In addition to these data sources, SDOH data from genomic data, wearable devices, social media, credit card transaction history, grocery store member history, and more creates a data pool. Scientists can draw the information from the pool to predict the needs and outcomes of health, create a 360-degree view of a population, and personalize healthcare for people. Many studies recommended utilizing the SDOH data for positive impact, especially in crises.

The barriers

            We are in an era that computing power is strong enough to analyze this behemoth size of SDOH data. However, two primary barriers are stopping us from utilizing the full potential of SDOH data. The ambiguous of causal conjunction between each SDOH data, and disconnection of the data pipeline.

Drowning in a data pool

We can use data to create information, knowledge, and wisdom. Companies like Google, Facebook are trying to collect more data from their users so that they can increase their profit. However, knowing what data to use for what purpose is the key. Only by collecting data will not solve any problem. Instead, people might drown in the data pool and never know why. Therefore, we need to understand the data. To know how to use the SDOH data. The SDOH data is from people’s everyday living environment; it is full of chaos and confusing information. For example, no one really knew the connection between social media and depression before scientists found the pattern. In many cases, we need to use research methods to determine the relationship.

Broken data pipeline

Human beings are not entirely lost in the realm of SDOH data. We sometimes know what we want and need to study SDOH data. However, we are just like a dog looking at a butcher’s shop drooling outside the window. In other words, we cannot always get what we want. Take the healthcare system as an example. Traditionally, healthcare organizations are loosely connected. The same kind of data can be created under different coding. On top of that, the definition of medical terminologies can be various cross facilities. Therefore, even when government agencies are asking for data, it is rarely a case that they can use the data directly without cleaning. If the data size is too big to clean, the data might become useless. Additionally, if you are not a government agency, many policies will stop you from getting the data. You either know the data is there but cannot use it, or the data is too messy that you cannot analyze it. 

Image by Omni Matryx from Pixabay 

Right time, right data, right hand

            During this worldwide public health crisis, we need the SDOH data to solve the problem together. And we need to do it with the right time, right data, and right hands fashion. Right time means we should get the data exchange without delay. Right data means we know the data we need for the analysis. Right hands means only authorized entities can access the data. No matter what background you have. I believe this is what every scientist can do during this crisis. Pay more attention to SDOH data. Try to use some of your cognitive load to consider what is the relationship between the data you are working and the benefit of public health. At the same time, consider how you can make these data more accessible to the public in a secure way. If we can break through the barriers, the crisis caused by COVID-19 may end sooner or even prevent for next wave.

Image by Gerd Altmann from Pixabay

References:

Dash, S., Shakyawar, S. K., Sharma, M., & Kaushik, S. (2019). Big data in healthcare: management, analysis and future prospects. Journal of Big Data, 6(1), 54.

May, T. (2018), The Fragmentation of Health Data. Retrieved from https://medium.com/datavant/the-fragmentation-of-health-data-8fa708109e13#:~:text=The%20healthcare%20system%20generates%20approximately,seeking%20to%20understand%20health%20data.

Trading Economics. United States Unemployment Rate. Retrieved from 

https://tradingeconomics.com/united-states/unemployment-rate