The following summary and its attached report are the second of an ongoing series by the W. Capra Data Science team on the impact of the COVID-19 outbreak on the industries we service over time. The first report can be found here.
Using data current as of April 21, new trends and patterns are emerging as more data is collected giving a more complete view of the situation the country finds itself in. New additions in this second report are updated results with testing, cases, and deaths, updated conclusions, and categorizations of each state’s current circumstances. Furthermore, an initial view of county level data is available in the attached report, which looks at the top 25 counties in the country by cases and deaths.
A note before we begin: This outbreak and the data surrounding it changes daily. This report was created when looking at the outbreak as a data problem that might benefit from data driven solutions and insights. It is not intended to be a substitute for medical or safety advice, nor is it a recommendation on outbreak response currently in place in various locations around the country. Individual assessment of local laws and current official government and health guidance should be reviewed before making any decisions.
COVID-19 Trend Analysis
Testing the population is the first step to containing the outbreak in each state across the United States (“US”). Inconsistent testing data reinforces that testing is still not being performed at the paces desired by most states. The ability to test wide populations quickly will be crucial to not only achieving containment but also maintaining it. Initial decelerations of cases and deaths is good sign; however, until states have the capacity to test people at the appropriate scales, states will continue to be at risk of resurgences in COVID-19 cases/deaths. Currently, 1.27% of the entire US population has been tested. This is very small from a sampling perspective, but not too far off from the rest of the world. The distribution of these tests is also very unevenly distributed – states such as LA and NY are leading the country with testing rates around 3% whereas states such as KS and VA remain at or below 1%.
Case rates around the country also vary widely, with some states reaching greater than a 40% case rate. One large reason for the trend is self-quarantine directives: patients with symptoms are told to stay home if the symptoms are manageable. Once severe enough, patients are told to see a doctor and get a test. The true rate is probably much lower, but because every individual cannot be tested at this time, the rate is an unknown. NY, NJ lead the country in case rates – likely due to the transmissibility of the virus along with the population density of those states – with MI following closely, likely due to its testing for only the most severe cases.
Death rates, or the rate of positive cases that result in morbidity, is currently 2.23% for the US. It tends to be higher for populations with comorbidities and advanced ages. WA is higher, but that is mostly due to how COVID-19 had a big impact on older populations. KY, OK, and NY are also rather high, while MI is currently experiencing the highest rate at 8.2%.
Tracking the Growth Curve
Although it is still early in the growth curve of the disease (i.e. all states are either in the “Exponential” or “Linear” growth phase), a framework is in continual development for evaluating the severity of the disease and success of treatment among the states. To be able to compare states against each other, they must be viewed by looking at rates or taking a per-capita approach, although looking at total numbers overtime can reveal trends as well. Time among all states is normalized by using a baseline “infection day”, which equals the first day when the state had reported more than 100 positive cases cumulatively. Because of this, the entire US total line is sometimes below other states.
The growth of positive cases is directly tied to testing availability, but trends are emerging to see how transmission rates might differ between different states. Since infection day, CA and WA have not experienced nearly as much growth in cases as NY. A notable factor in these differences has to do with NY administering more tests per capita (3,176 per 100,000 people) than CA (703) or WA (1,740). The slow growth may also point to the successes CA and WA have had with early implementations of lockdowns and social distancing measures. In fact, WA appears to be reaching its initial “peak” in Case Velocity. NY and NJ appear to very close to approaching their “peak” in Case Velocity. Not only are new cases in NY and NJ leveling off, but so are state-wide deaths.
Similarly, the growth in deaths is in viewed against “infection day”. Deaths in CA and WA are growing at a much slower rate. This can be attributed to early lockdowns, proper social distancing, and lower population density. Deaths in LA were starting to outpace NY and the US at large due to a spike in the data, but LA deaths have since started to level off. In terms of death rate, KY has performed the worst over the past six days with MI following closely behind. Please see the attached report for more detailed information on the current accelerations of each state.
The rate at which new observations are growing each day represents the acceleration metrics which are used to measure states’ progressions along the disease growth curve. These metrics will help flag states that are starting to slow in growth and reverse course towards containment. The US appears to be peaking in terms of test and case growth, while deaths are still accelerating in many states except for the New England region, which are entering the “Improving” phase. This is only an initial snapshot at the situation right now. A healthier setup would be if testing were increased while cases were decelerating like they are. If testing capacity is opening, hopefully testing for patients with less severe symptoms can increase, and eventually reach asymptomatic individuals as well. Another concern is that higher levels of new deaths are still occurring. Deaths are expected to follow a similar pattern to cases, but 5-7 days out. To compare specific states’ velocities and accelerations and see where each state is currently placed with regards to the four stages, please reference the State Comparisons section in the attached report.
Overall, the United States has reached an initial “peak” in new Case Velocity and new Case Acceleration has turned negative. Despite concern in Testing Velocity reaching a plateau, new Death Acceleration has been dropping across the US for the past week – signaling a positive trend across the country. In terms of new cases, most states are currently either in Stage 1 (Exponential Growth) or Stage 2 (Linear Growth). Several starts have entered or are soon to enter Stage 3 (Improving), as their new Case Accelerations have fallen below zero. Although some states appear to be entering Stage 3 (Improving) according to new Case Accelerations, falling Testing Velocities in these states may be falsely indicating improvement in new Cases. An example of such a state is Florida.
Early indications from new Case Velocity and Acceleration following the initial US “peak” are that the deceleration of new cases will take place over a much longer period of time than the rapid acceleration of new cases that led to the “peak.” In simpler terms, it appears that the recovery following the initial “peak” will take significantly longer than the duration of the Exponential Growth and Linear Growth stages. This trend is in line with what other countries across the world experienced with COVID-19. Louisiana represents an outlier in that it encountered both a rapid acceleration AND a rapid deceleration of new cases within a short period of time. We hypothesize that this may have been caused by a short, high-impact infection event, such as Mardi Gras celebrations.
Assessing the Business Impact of COVID-19
Very little has changed with regards to the situation various verticals find themselves in. Most of the country remains under full stay-at-home orders and those dollars remain at home as well. Almost every single consumer industry is seeing reduced traffic and purchasing, apart from some e-commerce merchants. However, moving forward, some interesting developments may change the situation for people – and therefore businesses – around the country. First, the increasing amount of people receiving their stimulus checks may boost spending in the immediate future. While this will not completely reverse the spending downturn being experience across the board, increasing the amount of spending money in people’s hands may help in the short term. Second, some states are beginning to lift their full stay-at-home orders and are reopening some of their places of business. These states will be monitored for both the change in their circumstances regarding the outbreak as well as the effect on business in the various states.