WHAT DATA MATTER?
There are lots of moving parts when it comes to tracking and predicting a virus. In this section we discuss the data elements that we track.
READ MOREWHERE DOES IT COME FROM?
Most of the data we use is publicly available. In some cases, we access a more computer-friendly source. In this section we discuss the data we use and where we get it.
READ MOREWHAT ARE INFECTION PARAMETERS
From Doubling Time to Beta and R0, you’ve likely heard a lot of words thrown around. In this section we discuss what these key measures mean and how they’re used in policymaking.
READ MOREWHAT DATA MATTER?
There’s a lot of data out there, and it’s important to identify it and present it to you.
Data Drives Decisions.
In a fast-moving crisis like Coronavirus/COVID-19, data play a vital role in shaping public policy. After all, you can’t fight what you can’t see. But, having the right data, right now, is crucial for shaping effective public policy.
NUMBER OF INFECTIONS
The number of new infections represents the number of persons testing positive for Coronavirus. These are called the known positives. The nature and method of testing vary by jurisdiction (each county) and setting (drive-through testings, physician’s office, emergency department, etc.). This number represents, more than anything else, testing activities. When testing slows down, the number goes down. We must be careful to distinguish the known positives from the true positivies, which represent the total number of persons infected, whether tested or not.
NUMBER OF HOSPITALIZATIONS
The number of hospitalizations represent the daily number of admits for confirmed COVID-19 diagnoses and suspected COVID-19 diagnoses. These data are reported by each hospital. Standards for reporting are typically set at the state level, and may vary over time.
NUMBER OF DEATHS
The number of deaths represents the number of persons dying from COVID-related conditions. Individual counties report the number of deaths, and standards for attributing deaths to COVID-19 may vary across counties and states.
ILLNESS SEVERITY
Illness severity data are drawn from the hospital data. In the hospital data, illness severity is represented by the level of care. The least severe among the hospitalized are admitted to medical/surgical rooms (general patient population). More severe cases receive critical care, which is pulled from Intensive Care Unit census data. Among the most severe are those patients who require ventilator assistance to breathe. Not reflected in this data are those whose illness is not severe enough to warrant hospitalization. This is the vast majority of patients.
WHERE DO YOU GET YOUR DATA?
DATA SOURCES
The data in our dashboard come from various sources. These sources are, for the most part, updated automatically each day in our Power BI dashboard.
INFECTION AND DEATH DATA
This data is updated automatically from a feed provided by the Johns Hopkins Whiting School of Engineering Center for Systems Engineering and Computer Science, who update it daily. You can find the data posted on GitHub here.
HOSPITALIZATION DATA
Hospitalization data are collected in North Texas by the North Central Texas Trauma Regional Advisory Council (NCTTRAC). Each day, hospitals report their data to the RAC, who compiles it and present the regional summary for the 19-county Trauma Service Area E.
POPULATION DATA
We compute the death and infection rates using the most recent population data from the American Community Survey 5-Year Estimates. We use the 5-Year estimates rather than 1-Year estimates because smaller counties may not be represented in the 1-Year estimate data. While we retrieve the population data via an API call, you can browse the data yourself by visiting the Census.
GEOGRAPHIC CORRESPONDENCE
We present county data in state and Core Base Statistical Areas (CBSAs) using a crosswalk provided by the MABLE/GeoCorr engine at the Missouri Census Data Center. While many of us think in terms of cities and counties, infections are a regional issue, and many of the healthcare resources we need are regional assets. These regional perspectives help to inform regional policy.
TESTING DATA
in his emergency declaration of March 12, 2020, Dallas Mayor Eric Johnson required all labs performing Coronavirus tests in the city of Dallas to report their daily summaries to the City of Dallas Office of Emergency Management. These results reflect only those tests run by labs located in the City of Dallas, and may represent patients from places on than Dallas.
So, What Are These Infection Parameters?
It’s not just the infection parameters, but also a series of other measures that we calculate. These measures are meant to inform policy decisions. They vary in both context and construction, but it’s helpful to know just a bit about how they’re constructed and how they might be used.
GROWTH RATE
SMALLER GROWTH RATES MEAN SLOWER SPREAD
The growth rate represents the average daily increase in cumulative cases over the time period analyzed. A 14-day growth rate of 2.5% would mean that, on average, the cumulative number of cases grew by 2.5% per day. We find the growth rate between time 2 and time 1 with the following formula: 
DOUBLING TIME
LARGER DOUBLING TIMES MEAN SLOWER SPREAD
Doubling time here is reported in days. It represents the number of days it would take the case volume to double at the current growth rate (which depends on the period analyzed). For instance, a doubling time of 11 days, computed from a 14-day period, means that based on the last 14 days of growth, the number of infections will double in 11 days. We find the doubling time from the growth rate with the following formula: .
