Artificial intelligence (AI) is not the first topic nurses consider on their way to work, but it is integrated into many aspects of workflow. As a young Intensive Care nurse, I remember being fascinated by the rhythms of electrocardiograms, learning how electrical patterns represented the flow of blood being pumped through chambers of the heart. I was even more thrilled to learn the skill of reading monitors, entering information into a bank of oscilloscopes that would interpret my patients’ rhythms as I prepared medications and worked at the bedside. The “machine intelligence”, or the intelligence of machines (AI), could save lives, sounding an alarm if a deadly ventricular rhythm appeared, warning nurses an immediate intervention was required.
We learned the algorithms would only function if nurses entered data accurately. Did the patient have a pacemaker? Was the device single chamber or dual? Any aberrancies to the basic algorithm needed to be entered into the system for the reading capabilities to be optimal. If the equipment read a rhythm in error, the nurse could declare a rhythm interpretation as false, and ask for the system to relearn. As such, nurses and artificial (machine) intelligence learned together.
AI has always worked this way, building upon basics and then replicating systems faster and faster. One of the earlier systems worked on identifying cat faces. Yes, cats.
AI is hard to ignore as a global race to become the world’s leader in the field has become more frantic, but the field didn’t begin that way. Although AI is described as the “new electricity” with global spending predicted to be around $36.1 billion by the year 2025, the roots of this burgeoning field began in 1956, when the term artificial intelligence was first coined during a summer research project. Did anyone know how big the field would become? That is doubtful, especially since interest in robotics has seen many ups and downs since that time.2
However, by 2011, the field began to take off when an individual by the name of Andrew Ng demonstrated what could be done with chips and microprocessors and algorithms by proving that computers could learn at an amazing rate without being told what to do. His research proved that a computer, by viewing over 10 million online videos of cats, could start to recognize what a cat was, without being told what it represented. His technology formed the basis for what speech recognition is today.2
Additionally, his research brought new processing power to many of the computers and applications that are now used in healthcare!
What is possible? A trip around the globe could show us what the future might hold for nurses. Japan currently has the oldest population of any OCED (Organization for Co-operation and Development) country in the world, but unfortunately, many of their elders live alone. With many centenarians, their population requires assistance that isn’t readily available. Enter the field of robotics, or AI. They have several comfort care (friendly) robots, as well as a robotic bear that can assist with lifting and turning patients, carry a patient from point A to point B, or even turn a patient in bed. One of the friendly robots is named Paro, who can engage in simple conversations to provide the elderly with social interaction.3
Taking AI one step further for patient needs, robot Pepper assists in a more clinical way. Pepper can assist in scheduling medical appointments, providing education, and interpreting vitals and labs. Another friendly robot named Dinsow has been designed to assist with mood, improving activity, plus reminders to take medication.3
In Canada, AI experts predict nurses will rise to the occasion, with a few taking the lead in integrating AI into nurse-patient education and wellness. Nurse leaders will understand patients expect more from computer-driven interactions than a frozen face on a screen. It is exactly this recognition from nurses that has driven the success of programs such as TAVIE, an AI model that is utilized for patient teaching in the field of HIV.1
TAVIE has been particularly helpful with medication compliance, not just getting patients to “take medications as instructed” but recognizing instances when patients might not be willing to risk taking medication, such as parties or social activities. TAVIE interacts with patients and assists them in problem-solving real life. Without input from nursing, a different model might prove less successful, less compassionate, less attuned to the types of issues patients with chronic conditions could encounter.1
But what about the United States? How is AI affecting nursing process within America? Whether we discuss the process of linking inpatient and outpatient encounters via electronic charting systems, or the ability to rescue inpatients through real time recognition of changes in vital signs when sepsis begins, machine intelligence is everywhere nurses care for patients.
Computers and robotics have also begun to provide care where providers may be scarce, through video and teleconferencing. Seniors may be provided with monitoring equipment in their home, so they can provide frequent updates to providers on data such as oxygen saturation, BP, body weight, heart rate and rhythm, as well as perceived exertion levels. These data points prove extremely helpful for prompt medical/pharmaceutical intervention, thus keeping the elderly in their homes as opposed to resurfacing in hospital Emergency Rooms in advanced crisis states.
And robots? Yes, we have those in the United States as well. Although a few people may believe AI is frightening and a bit analogous to science-fiction, it might be easier to think of it as augmented intelligence, like algorithms that assist us in scanning patients for lung cancer or in detecting breast cancers in their earliest forms.4 As for robots, nurses need to see the addition of robots as an extension of nursing, rather than a replacement. Like the products utilized for the elderly in Japan, robotics is designed to augment the work of nursing, which is why nursing needs to be heavily involved in the strategy and design.
In Toronto, for example, a robot has been designed to interact with residents at a retirement home and monitor residents for signs of dementia. At MIT, researchers are testing robotic decision support that might “schedule nursing tasks and assign rooms to patients.” Additionally, throughout the United States, novice nurses are being trained in simulation labs, utilizing SimMan automated training to replicate patient neurologic or physiologic situations they might encounter at the bedside. These high-fidelity mannequins utilize advanced training algorithms that simulate trauma including cardiopulmonary arrest, with variations in heart and lung sounds for trainees to detect.
So…what might nurses expect from AI in the future? Highly promising aspects of care integration, whether it be smartwatches that communicate to the EHR and document in real time, or robots that document as nurses “dictate” care, comparable to what is seen currently with scribes; exciting possibilities!
But whatever is around the corner with augmented intelligence, nurses need to remain engaged, to become the true leaders in this electrifying field.
- Canadian-nurse.com “Artificial intelligence, automation and the future of nursing.” Glauser, W., May 01, 2017.
- Journals.lww.com “How artificial intelligence is changing nursing.” Nursing Management, Robert, N., September 2019.
- Nurse.org “Will these nurse robots take your job? Don’t freak out just yet.” Hamstra, B., February 27, 2018.
- Nursingtimes.net “What is artificial intelligence and what does it mean for nurses?” Jones, L., August 20, 2018.