Remember the opening scene in Back to the Future, where Doc Brown’s chain reaction machine is programmed to make breakfast but fails miserably? As a nurse, well-intended new technology can sometimes feel this way—interesting, but not useful and sometimes more trouble than it is helpful.
But with all the recent talk around AI, will this time be different?
Overworked, and over-teched
Nursing is driven by deep compassion, relentless advocacy for the patient, and the fulfillment that comes from giving comfort during a life-altering situation. As a nurse for the last two decades, I can vouch that my own reason for joining the profession was similar; I’m inherently drawn to supporting patients and their families in their time of need.
But increasingly, nurses are being pulled away from their defining purpose. They’re required to document numerous aspects of patient care in order to meet a wide variety of regulations at the federal and state levels, not to mention the expectations of accrediting entities. The result is that nurses report spending half their time on any given shift documenting or searching for data in the electronic health record (EHR) and also report that they have reduced patient contact.
Adding to this complexity are the current nursing shortages, and future predictions that nearly one third of nurses plan to leave the profession in the next few years. Nurses are also feeling these new pressures with half reporting burnout symptoms a few times a week to every day.
To alleviate these woes and increase productivity, hospitals are increasingly investing in new technologies. Like, a lot of new technologies. So many in fact, it’s hard to keep up. And many of these “solutions” actually add to our workload rather than allowing us to focus on our core purpose: caring for patients.
More bandwidth, more human interactions
So, it’s not a matter of more tech. It’s about ensuring we have the right tech. Tech that allows us to focus on the people we serve. Tech that gives us the bandwidth to engage in deep, meaningful human interactions. And while it may seem counterintuitive, I firmly believe artificial intelligence (AI) is the way we achieve this.
Take documentation, for example. Companies like Amazon Web Services are using AI to capture conversations between patient and clinician, extract relevant information, and create summaries for EHR systems. AI applications like these allow nurses to allocate more time and energy to direct patient care.
Artificial intelligence, real outcomes
But can AI actually help with health outcomes? Yes! And this is what excites me the most about AI’s potential, especially for emergency departments and inpatient care.
AI can be a powerful ally in helping patients and family members stay engaged throughout their care experience and beyond. For example, mobile apps powered by AI not only give patients a real-time view into their progress, but can also crunch data from various systems to predict when they might be ready for their next step in their health journey, including discharge.
When it comes to health literacy—a factor the American Medical Association says is a stronger predictor of health status than age, income, employment status, education level, and race—AI can make a huge difference. For example, the same algorithms Netflix uses to recommend content can be used to automatically prescribe video education based on information contained in the EHR and clinician notes. While this isn’t necessarily a big time saver—I’m still going to explain things to my patients—it allows us to have more productive conversations. When they’re better informed, patients ask better questions and feel more empowered to take charge of their health outside my hospital’s walls.
I’m also quite excited about the potential that generative AI—the kind used in ChatGPT—has for improving health outcomes. Large language models (LLMs) are really good at simplifying complex medical jargon that seems like a foreign language to most patients. “Edema” becomes “swelling.” “Cerebral infarction” becomes a “stroke.” And “NPO at 00:00.,” means the patient “shouldn’t eat or drink after midnight.” When LLMs translate medical terms into plain language, patients have much more actionable information on hand and that leads to better health outcomes.
Given the right training, validation and testing, generative AI can parse things like 15-page discharge summaries to pull forward the most relevant details for each patient. Nearly 80% of the content is boilerplate information (e.g., COVID policies) or things the patient already knows (e.g., their body mass index). In mere seconds, AI can sift through all that stuff and present the patient with, for example, the top three things they need to do after leaving the hospital: pick up medications, avoid certain foods, set up an appointment with a specialist, etc.
One nurse, so many questions
These advancements in AI have created an important discussion within health care about the path forward to address the challenges facing more than 5.2 million nurses in the U.S. today. I myself was once a skeptic. But now I find myself engaging in more critical thinking on the matter, asking myself and my colleagues questions like:
- How can AI automate, even incrementally, the rote tasks and administrative duties expected of nurses?
- Where can AI make the biggest impact on my patient’s health literacy?
- How can technology help me have more productive conversations with my patients?
- How can LLMs cut through the crap and help patients focus on what really matters?
- Who within my hospital should be consulted when considering new AI applications?
Clinical input, critical implementation
When integrating AI into nursing workflows, it’s crucial to ensure that the technology aligns seamlessly with clinical practices. In my experience, when a new technology is implemented in the hospital, the decision makers aren’t the ones at the bedside.
Unfortunately, the nurses are expected to learn and implement any new technologies that are introduced into their workflows. And, in many cases, this pulls them away from patient care. For many nurses, learning and implementing new technology that doesn’t align with clinical workflows creates frustration and mistrust of technology. Technology is supposed to help augment and improve patient care, but we know that is not always the case.
Any technologies, including those powered by AI, should be designed in a way that doesn’t require extensive training or a drastic change in clinical workflows. It should be intuitive and work seamlessly. That’s why it’s so important for nurses to have a seat at the table, from decision making through implementation.
Nurses’ expertise and insights are invaluable in ensuring that AI addresses real-world challenges and actually contributes to positive patient experiences. By actively involving nurses in the development and deployment of AI solutions, healthcare organizations can make informed decisions that prioritize patient care and nurse well-being.
Back to the future
So, will AI be more like Doc Brown’s breakfast machine, or will it actually help nursing? I think it’s all about perspective.
AI doesn’t replace nurses. It augments what we do best: care for patients. So for me, getting “back to the future” means having deeper, more meaningful interactions with my patients. After all, that’s why I got into this profession in the first place.
Stephanie Frisch, PhD, MSN, RN, CEN, CCRN-K is a healthcare technology nurse leader using artificial intelligence to keep patients informed. She is Director of Nursing at Vital.
Stephanie Frisch
Stephanie Frisch, PhD, MSN, RN, CEN, CCRN-K is a healthcare technology nurse leader using artificial intelligence to keep patients informed. She is Director of Nursing at Vital.
