By now, the coronavirus pandemic has affected nearly every American in some fashion. The U.S. has become the epicenter of the COVID-19 crisis with greater than 217,000 cases as of April 2, 2020. Across the country many states have issued mandatory closures of schools, restaurants, and other businesses and mandated cancellation of public events and other mass gatherings. On March 31st, White House officials released a grim projection that the U.S. could see between 100,000 and 240,000 deaths by the end of this crisis. How we and many other countries have chosen to combat this virus is through social distancing. As self-quarantine becomes an all too familiar aspect of life, it’s difficult not to wonder how this became our reality.

Identifying COVID-19 cases early is paramount to flattening the curve.

The simplest answer is testing—or lack thereof. Early, widespread testing for the COVID-19 virus would have significantly changed the trajectory of the U.S. outbreak. Testing informs us where the disease is and how to most effectively allocate resources. Outside of China, South Korea had the earliest cases of COVID-19. However, South Korea was able to slow the spread of the disease through widespread testing, without resorting to the strict lockdown strategies that China and many other countries have adopted.

AI has been able to distinguish COVID-19 from community-acquired pneumonia and other respiratory diseases.

Among other factors, inefficient testing has hampered our response to the crisis. Backlogs at labs continue to delay results and testing inequalities further divide America along socio-economic lines. As important as early screening is in the U.S., the accuracy of current COVID-19 tests vary significantly, sensitivity have been reported to range from 42% to 71%. Therefore, a person may test negative but still shed the virus and infect others. To make matters worse, it may take days before a retest may indicate they are positive. In those cases, mildly symptomatic individuals may unwittingly infect close high-risk relatives. Identifying COVID-19 cases early is paramount to flattening the curve, which the U.S. is woefully ill-equipped to do. However, there have recently been interesting developments in early screening and detection.

Artificial intelligence (AI) has shown promise through a chest CT to detect COVID-19 by distinguishing the virus from community-acquired pneumonia and other respiratory diseases. Although CT screening to identify COVID-19 is not currently recommended by most radiology societies, this may change as COVID-19 detection and management parameters evolve. Based on an expert consensus statement from the Radiological Society of North America, its estimated that a chest CT may predict early COVID-19 infection 92% of the time versus current polymerase chain reaction (PCR) tests which have a 42%- 71% accuracy rate. This may be an avenue worth exploring, especially if it will lead to flattening the COVID-19 curve.