Healthcare systems are society’s bastion against Covid-19. But two years of working long shifts to cover quarantined colleagues and witnessing patient after patient die of Covid-19 complications has led to massive burnout among frontline healthcare staff. Many highly experienced healthcare professionals, who tend to be older, are choosing to retire rather than face the Covid-19 related risks of working in a hospital. Healthcare, like much of the rest of the economy, is facing a labor shortage.
This shortage is biting as health systems are accumulating big pandemic backlogs. During the past two years, routine checkups, elective surgeries, cancer screenings, CT scans and other important procedures have been canceled or delayed due to the multiple Covid-19 waves. Many of those delayed scans and empty hospital beds will result in early interventions being missed, leading to later, more acute and highly urgent admissions that could cost lives.
A report by Mercer suggests that, within five years, as increased healthcare demand and retirements continue to outstrip new recruits, healthcare in the United States will be 3.2 million workers short. Healthcare organizations must consider other options, forcing multipliers to keep the system working during ongoing Covid-19 waves and all year round. Additionally, a survey from the American College of Healthcare Executives found that personnel shortages ranked as the No. 1 concern in 2021.
Technological advances in medicine, surgery and healthcare have revolutionized patient outcomes and made the impossible commonplace. Some advances in research and technology can shrink the healthcare load: Vaccines reduce disease burden and keep people out of hospitals, early cancer screenings enable less invasive therapies and laparoscopic surgeries cut recovery times. Artificial intelligence (AI) has proven to help in a number of areas with efficiency.
Patient management automation
One major class of labor-saving innovations fall under the label of “automation.” Healthcare of all fields was slow to adopt automation. In the 20th century, healthcare automation tended to focus on the manufacturing floors of pharmaceutical producers and simple computer-assisted diagnosis (CADe) tools.
A Deep-dive Into Specialty Pharma
A specialty drug is a class of prescription medications used to treat complex, chronic or rare medical conditions. Although this classification was originally intended to define the treatment of rare, also termed “orphan” diseases, affecting fewer than 200,000 people in the US, more recently, specialty drugs have emerged as the cornerstone of treatment for chronic and complex diseases such as cancer, autoimmune conditions, diabetes, hepatitis C, and HIV/AIDS.
In the last decade, though, AI has begun to enable real automation in hospitals. New healthcare AI solutions are emerging every few months that automate processes in healthcare facilities, taking certain tasks in medical processes off of the physician or nurse’s plate. To name just a few applications, AI is proving to be a valuable tool in medical record automation, such as AI-aided documentation for regulatory and reimbursement, and chatbots for communication with patients.
Resource planning
Beyond automation, the pandemic has presented both an urgent need and an opportunity to use AI to make healthcare more efficient through predictive resource planning.
Many hospitals deployed AI models to predict which Covid-19 patients were most likely to deteriorate. In the UK, for example, radiology departments were overwhelmed and understaffed even before Covid-19. Teleradiology firm Hexarad is developing software to help British radiology departments understand their staffing needs and redeploy to meet capacity. Renown Health, headquartered in Reno, Nevada, created a command center to monitor patients so that fewer nurses could monitor more patients as Covid-19 admissions kept rising.
These tools help to deploy existing human resources in the most efficient way possible, but AI has the potential to reach beyond EMR automations and planning practices in healthcare.
Triage and expediting treatment
Another area where AI is demonstrating high value in healthcare is in triaging patients. Hospitals struggling with human resources must learn to quickly provide the same quality of care to the same, or even a rising, number of patients. Prioritization is one such way to tackle the problem.
By highlighting critical or urgent cases from medical records, AI can help get the right patients in front of the right healthcare workers at the right time. Radiology is one of the first disciplines to harness the capabilities of AI and currently has the most mature AI-powered triage tools. For example, radiology AI tools have been able to demonstrate results in improving emergency department (ED) throughput by 20% for ruling out brain bleeds.
But many other disciplines are well on the way to their own solutions that could provide similar results. While not yet deployed as a solution for clinical use, one promising AI model predicts a Covid patient’s oxygen needs from a chest X-ray.
As the crisis environment of the pandemic and staffing shortages may improve efficiency within departments, there’s always a risk that these local efficiencies could be limited by the organizational boundaries of healthcare organizations. Let’s imagine an AI solution that helps an emergency department triage much more efficiently: Although prioritization creates a positive effect within the department, its real impact depends on how quickly the response teams can treat the patient.
For example, AI triaging a patient with a suspected pulmonary embolism is a critical starting point, but a bottleneck may occur in coordinating the subsequent necessary care. Relaying the positive findings to the ED physician can sometimes be impeded by poor communication and an inability to share relevant patient information rapidly—often carried out manually.
Some health systems have deployed rapid response teams for stroke and pulmonary emboli, which are composed of multidisciplinary specialists like neurologists or interventional radiologists. In these scenarios, AI could facilitate notification of these response team members of a detection as soon as possible, and allow them to confer with the relevant patient data. Instead of allowing only radiologists to read faster and better, AI can empower cross-specialty collaboration, connecting disparate departments facing data silos and shortening or eliminating the manual process of patient management.
In the longer term, AI will serve a valuable role at a level beyond the individual patient or single department, working to monitor changes between departments, flag troubling anomalies in treatment processes, and help coordinate to ensure that nothing is missed. Whole-hospital enterprise-grade AI solutions like this might be the most complicated, but they’ll ultimately help clinicians and healthcare organizations as a whole. We have no doubt that AI will become the backbone of hospital operations in the future—automating the healthcare operating system and taking a load off of the hands of heroic, and exhausted, medical professionals.
Photo: Natali Mis, Getty Images
