In the span of just a few short months, the Covid-19 pandemic has drastically altered life as we once knew it. Nearly 5 million cases of the novel coronavirus have been recorded in the United States since March, and more than 150,000 Americans have died as a result of complications after contracting the virus. The immediate consequences of the outbreak are tangible and clear in regards to the calamitous loss of life and the devastating economic impact, but the long-term effects of this pandemic are still difficult to assess and nearly impossible to predict. Furthermore, Covid-19 has placed tremendous additional stress on a national health care system that has long been insufficient and outdated even prior to the pandemic. 

Hospitals in areas across America that have been hit the hardest – such as San Benito County and San Joaquin County in California – are rapidly nearing ICU bed capacity. Doctors and nurses are being forced to improvise in the wake of personal protective equipment (PPE) shortages by reusing N-95 masks and repurposing snorkeling gear and garbage bags into face shields and scrubs. Then there’s the crucial issue of patient health data, much of which gets trapped in health record system archives that communicate inefficiently with one another. This can potentially create substantive issues with regards to the coronavirus, as a newly suspected case in one health system might be imperceptible to that of another provider within the same community. An oversight of this magnitude can effectively blind public health officials to potential Covid-19 hotspots emerging right under their noses. 

Many of these problems are due to the fact that our healthcare system – much like the economy – is designed to manage incremental, linear demand, while the virus continues to grow at an exponential rate. Fortunately, modern developments in artificial intelligence (AI) are proving to be quite promising, and the acute adoption of digital operating systems may just turn the tides of battle in our favor in the long and arduous war against Covid-19. The current healthcare model in the United States is simply not equipped to handle such an unprecedented surge in demand, and there has never been a greater need for an immediate and large-scale implementation of machine learning algorithms.

As we work to slow the spread of Covid-19, healthcare providers must begin looking towards optimizing response mechanisms by digitizing as many steps of the process as possible. This is due largely to the fact that traditional processes rely on individual employees to operate in the crucial path of signal processing. In other words, the conventional methods are limited by the rate at which workers can be trained, organized and deployed. Furthermore, traditional processes yield smaller profits over time as they continue to scale. Digital systems, on the other hand, are more than capable of keeping up with exponential growth, as they operate with virtually no constraints and can be scaled up nearly infinitely. 

This is not to say that AI will render healthcare professionals obsolete, as nothing can serve as a truly functional replacement for human empathy and decision making. Despite the overall improvements to efficiency and enhanced patient outcomes that artificial intelligence can deliver, human clinical expertise is still necessary within the healthcare industry. Data evaluation and quantitative analytics undoubtedly account for a significant portion of a physician’s work, but interpreting a diagnosis and establishing a patient treatment plan are not linear processes. They require a degree of creativity and problem-solving capability that algorithms and robots simply do not possess. The application of AI should be supplemental to human rationality and reasoning in order to ensure the highest quality patient care possible. 

A number of hospitals in the United States have already started to implement machine learning algorithms into various systems and processes. Partners Healthcare in Boston, for example, has recently begun experimenting with chatbots and interactive voice response systems in order to promote patient engagement. Several health centers throughout Boston have taken a similar approach by employing AI to perform straightforward tasks which were previously done by trained clinicians and by offering a broader range of telehealth applications where possible. One such hospital – the Betsy Lehman Center – is currently utilizing a series of new protocols developed by the Center for Medical Simulation which allows doctors and nurses to check in with co-workers throughout the duration of their Covid-19 care shifts. These innovations began with a single hospital, and now health centers across the city of Boston are adopting similar measures to help combat the coronavirus. 

Healthcare providers and professionals aren’t the only ones who stand to benefit from the widespread application of AI during the pandemic. Artificial intelligence systems that are equipped with machine learning capabilities are able to offer clinical decision support that could play a vital role in keeping Covid-19 patients alive. A team at Stanford University led by physician Ron Li is currently evaluating an AI system for possible integration into clinical workflows. The team is utilizing a model developed by Epic – Stanford’s electronic health record provider – that can identify patients who are more likely to go into cardiac arrest or need ICU admission. This “deterioration index” is now being evaluated to see whether or not it can accurately predict which Covid-19 patients will require more intensive care as a result of deteriorating conditions. “Currently, a dedicated triage team closely monitors patients with COVID in our hospital. It’s a highly manual and resource intensive process of chart review and in person evaluations. But if you have hundreds of patients in the hospital, that won’t work,” Doctor Li explained. In addition, an automated system would provide an objective standard for all healthcare professionals within a given facility, which could help prevent certain patients in need from falling through the cracks. 

The Covid-19 pandemic has wrought havoc throughout the world and caused unimaginable pain and suffering, but there may be a silver lining. By exposing shortcomings in the industry, the coronavirus may have indirectly been the necessary catalyst for widespread healthcare reform to take place. In other words, the virus has evolved into a vehicle for AI implementation within the healthcare sphere. The problem isn’t that the current system is failing, but rather that it’s working exactly as it was designed to. Therefore, the solution is not to wait for things to return to normal, but to institute a new ‘normal’ that adequately addresses the countless failures of the current healthcare system. By automating data collection processes and working with medical professionals to improve patient outcomes, machine learning algorithms can begin addressing problems at every level, from doctors to patients to entire communities. Now is the time to comprehensively reconstruct our anachronistic healthcare system in a manner that has never been done before, so that we might be better equipped to handle the next national health emergency.