Can AI help doctors fight burnout?

Even before COVID-19, 40% of physicians said they felt burnt out. But the pandemic was a turning point. Working in jerry-rigged PPE in overcrowded and understaffed intensive care units, more than 3,600 American healthcare workers died in the first year of the pandemic alone. After witnessing the solitary death of about 1 million patients, holding the phone as they shared their last minutes with family members via FaceTime, more doctors are deciding to retire early, adding to a looming shortage. A report last year by the Association of American Medical Colleges predicts a shortage of up to 124,000 physicians by 2034. This includes a gap of up to 48,000 primary care physicians, who report higher levels of burnout than other specialties. And it’s not just doctors: In a January 2022 survey by Prosper Insights & Analytics, only 50% of all healthcare workers said they were “happy” at work.

Happiness cannot be bought overnight. Staffing gaps will take time to fill. But in the meantime, say proponents, artificial intelligence (AI) could be used to help ease the burden on doctors as much as possible. “We need to turn every doctor into a super-doctor,” says Farzad Soleimani, assistant professor of emergency medicine at Baylor College of Medicine and partner at Houston VC 1984 Ventures. “Ultimately what clinicians do is learn to recognize patterns. That’s the power of AI.

Of course, there are skeptics. An April 2019 Medscape survey of 1,500 doctors in Europe, Latin America and the United States found that a majority were anxious or uncomfortable with AI, with American doctors expressing the most skepticism (49%). Relying on algorithms for patient care also presents ethical, clinical, and legal issues. AI can bring considerable threats to privacy issues, ethical concerns and medical errors. Developers can unknowingly introduce biases into AI algorithms or train them using faulty or incomplete datasets. Data used to train AI systems could be vulnerable to hacking. By entrusting certain aspects of decision-making to machines, physicians could lose their traditional autonomy and authority, and notions of accountability will be tested if AI-guided recommendations result in patient harm.

Nevertheless, hAccording to CB Insights, healthcare AI companies, including nearly 500 early-stage startups, raised a record $12 billion in funding last year. Here are some ways tech companies are using deep learning algorithms and natural language processing to automate routine tasks in hospitals, reduce the hours medical providers spend on paperwork, and reduce errors caused by fatigue.

Accelerate pre-visit assessments

Managing patients and preventing provider burnout begins before care recipients even show up at the office or hospital. San Francisco-based Health Note streamlines patient admissions with a text-based AI chatbot that collects patient information before the visit and automatically writes notes for their doctor, reducing admission time and documentation up to 90%, depending on the company. Decoded Health – a spin-off from SRI International, the nonprofit research organization that developed the technology behind the computer mouse, ultrasound and Siri – offers what it calls a “virtual medical resident” which pre-screens patients using natural language processing, creating a summary of their medical complaints with actionable care recommendations. Keona Health is focused on helping nurses and non-medical staff with triage over the phone, guiding them through symptom checking, offering care recommendations, and automating appointment scheduling.

Help with triage

When ERs come under fire, AI triage tools are designed to help flag patients who need intensive care and who might otherwise be missed, flagging the most severe cases and prioritizing them for care. The first major clinical application of AI triage tools was in radiology; companies such as RapidAI,, and Arterys all have FDA approval for algorithms that detect signs of strokes, brain bleeds, and pulmonary embolisms from CT scans. Imagen’s FDA-cleared OsteoDetect analyzes wrist x-rays to detect distal radius fractures, one of the most common injuries to the joint. Mednition’s real-time triage-guidance tool, KATE, analyzes EHR data and vital patient data collected on admission to help emergency nurses spot the warning signs of sepsis, which accounts for more than 50% of hospital deaths. It is used throughout the Adventist health system and others to avoid emergency room admissions through earlier treatment. Johns Hopkins University-run ERs use Stocastic’s TriageGO, which analyzes vital signs and other admission data, as well as patient demographics and medical histories to make timely care recommendations, reducing the “Door to decision” time up to 30 minutes.

Transcribe doctors’ notes

A recent study found that doctors spend an average of about 16 minutes on electronic health records for each patient visit. DeepScribe is a digital voice assistant that allows a doctor to have a normal conversation with his patient, transcribe it, extract key information and automatically integrate it into the appropriate sections of medical records. In January 2021, the San Francisco-based startup raised $30 million. Competitors include Nuance, Suki and Corti.

There’s also Rad AI’s Omni software, a virtual assistant designed specifically for radiologists that helps write a formal “clinical impression” based on dictated notes, automatically inserting guideline recommendations and flagging potential errors.

Management of the invoicing process

“When people talk about health care staffing shortages, they often think of nurses, doctors and front-line caregivers, but the issue is across the entire organization,” says Cat Afarian, vice- president of communications at South San Francisco-based Akasa, an AI provider. services for health operations. According to recent surveys by the Healthcare Financial Management Association, more than 57% of healthcare systems and hospitals have more than 100 open back-office positions (billing, recording and scheduling) to fill. A recent survey by Change Healthcare found that 65% of healthcare leaders are already applying AI in their “revenue cycle management” and that by 2023, 98% plan to do so.

Akashabased in San Francisco, provides services to more than 475 hospitals and health systems and more than 8,000 ambulatory care facilities in all 50 states, using a constantly learning artificial intelligence system to help them automate verification , billing and collection of insurance claims status. Privia Health provides scheduling and billing tools to some 3,300 independent physicians, using robotic process automation, in which an intelligent system learns a scripted process to handle repetitive billing tasks as a human would.

Help with testing

Laboratory tests shape about two-thirds of the decisions doctors make. Prior to COVID-19, medical laboratory professionals performed some 13 billion lab tests per year. In a February 2020 survey by the American Society for Clinical Pathology, more than 85% of medical laboratory workers reported burnout; 36.5% complained of a lack of staff. That was before the added burden of performing more than 900 million COVID-19 tests since the start of the pandemic. Many hospital laboratories operate with 10% to 35% vacancies.

Automating repetitive work could allow fewer people to do more, and possibly improve results as well. In a 13-month pilot project, the University of Texas Medical Branch Hospital at Galveston used Biocogniv’s ‘lab intelligence platform’ to help process more than 325,000 COVID-19 tests and make personalized interpretations based on PCR and antibody testing, patient vital signs, and medical history. The result: almost doubled efficiency, lower escalation rates to intensive care and reduced mortality rates. “COVID-19 has been a time of immense change both clinically and operationally,” says Peter McCaffrey, assistant professor of pathology at the hospital and director of its pathology informatics and information systems. laboratory. “With Biocogniv’s platform, we were able to adapt interpretation and guidance for COVID and coordinate everyone during this time of unprecedented uncertainty.” In the company’s pipeline: laboratory-based predictive tools for sepsis, respiratory failure and acute heart failure.

Whether or not AI proves itself in each of these areas, there is no turning back. “By minimizing or offloading repetitive diagnostic tasks, [AI can help] physicians are spending more time on sophisticated clinical reasoning and judgment, and inherently human work, such as engaging with multidisciplinary care teams to support patient care,” says Mark Schuster, pediatrician and founding dean and CEO of the Kaiser Permanente Bernard J. Tyson School of Medicine in Pasadena, California. In addition to addressing shortages of specialist physicians in fields like radiology, where AI has proven to be highly accurate, Schuster predicts that clinical care algorithms will become more powerful, with “a gradual increase in accuracy and personalization of diagnosis and treatment. “Yet he recognizes the potential danger that AI could reinforce biases that already exist in the healthcare system. “We recognize,” he says, “that there remains a substantial risk of unmeasured bias being introduced by machine learning in AI.”

Comments are closed.