AI Nurses: A Staffing Solution or a Threat to Quality Care?
The next time you schedule a medical appointment, the voice on the other end of the line might belong to an AI assistant. These programs, designed to handle tasks traditionally performed by nurses and medical assistants, are becoming increasingly prevalent in healthcare settings. While some see AI as a vital solution to staffing shortages and a way to improve efficiency, others express concerns about the potential impact on the quality of patient care.
One example of this technology is Hippocratic AI, which has developed an AI program called Ana. Ana is designed to assist patients by answering questions and guiding them through preparation for appointments. She is available around the clock and in multiple languages.
This trend signals AI’s growing presence in healthcare. Hospitals are using sophisticated AI programs to monitor patients’ vital signs, identify emergencies, and create care plans. Proponents suggest that AI can help nurses work more effectively, addressing burnout and understaffing.
However, nursing unions are raising alarms about the technology, arguing that it undermines nurses’ expertise and potentially degrades patient care. Michelle Mahon of National Nurses United, the largest nursing union in the U.S., stated, “Hospitals have been waiting for the moment when they have something that appears to have enough legitimacy to replace nurses.” There are concerns that the technology may automate, deskill, and ultimately replace human caregivers.
The National Nurses United has organized demonstrations and is advocating for the right to influence how AI is used and for protection from disciplinary action if nurses choose to disregard automated advice. Recent developments that include suggestions from Robert F. Kennedy Jr. and Dr. Mehmet Oz about the role of AI in healthcare have further fueled the debate.
One of the initial promotions for AI assistants included a cost of $9 per hour compared with about $40 per hour for a registered nurse. The company, Hippocratic AI, has since shifted its marketing strategy, now focusing on the thorough testing and superior services it can offer, but declined requests for interviews.
Hospitals have been experimenting with technology, including sensors and cameras, for years. Now, this data is being integrated with electronic medical records. The goal is to anticipate medical issues and direct nurses’ care, sometimes even before they’ve assessed the patient.
Adam Hart, a nurse who worked in an emergency room, recounted an instance where the hospital’s computer system flagged a patient for sepsis. Hart’s assessment, however, revealed that the patient was undergoing dialysis and required a different approach to IV fluids than suggested by the system. Despite his concerns, he was initially instructed to follow the standard protocol until a physician intervened.
“You need to keep your thinking cap on— that’s why you’re being paid as a nurse,” Hart said. “Turning over our thought processes to these devices is reckless and dangerous.”
Nurses like Hart express that the goal of AI—to assist nurses in monitoring multiple patients and responding quickly to problems—is a worthwhile objective. However, they also point out the frequency of false alarms, potentially caused by misinterpreting routine bodily functions as emergencies.
Melissa Beebe, a cancer nurse, described how constant alerts, accurate or not, can be distracting. Michelle Collins, the dean of Loyola University’s College of Nursing, noted that even the most advanced technology can miss nuanced cues like facial expressions and odors that nurses are trained to identify.
Despite the concerns, AI is also being seen as a vital tool for hospitals facing a significant nursing shortage. According to an estimate, over 100,000 nurses left the workforce during the COVID-19 pandemic. The U.S. government anticipates more than 190,000 new openings for nurses each year through 2032. Thus, hospitals are looking to AI to help in ways that don’t replace nurses, such as gathering information and communicating with patients.
For instance, the University of Arkansas Medical Sciences uses an AI assistant from Qventus to make hundreds of calls weekly to prepare patients for surgery. The AI confirms information about medications, medical conditions, and issues like sleep apnea before anesthesia. The hospital can then use this tool to contact several patients within a limited timeframe.
“So what we need to do is find a way to call several hundred people in a 120-minute window — but I really don’t want to pay my staff overtime to do so,” according to Dr. Joseph Sanford.
Qventus reports that 115 hospitals use its services. The company claims that its technology improves hospital earnings by quickening surgical turnarounds, reducing cancellations, and mitigating burnout. Each call begins with the program identifying itself as an AI assistant, and Dr. Sanford says it’s important to be transparent with patients about when they’re talking to a human and when they’re not.
Other AI developers are creating more advanced applications. Israeli startup Xoltar, for example, is developing humanlike avatars. They are being tested with the Mayo Clinic for patients with chronic pain. The company is also creating an avatar to help smokers quit. Testing shows patients spend about 14 minutes talking to the program, which can recognize facial expressions and body language.
Nursing experts suggest that, while such programs may be effective for relatively healthy individuals, the sickest patients are the ones who utilize healthcare the most, raising questions about AI’s ability to address their needs. Some experts advocate for a cautious approach. Roschelle Fritz of the University of California Davis School of Nursing said, “It’s the very sick who are taking up the bulk of health care in the U.S. and whether or not chatbots are positioned for those folks is something we really have to consider.”
