Navigating the future: AI, and its impact on healthcare
The landscape of technology is evolving at an unprecedented pace. When Facebook arrived on the scene, it reached 100 million users in 4.5 years. Instagram hit that same milestone in half the time. TikTok, nine months. And OpenAI’s ChatGPT rocketed to 100 million users in just 60 days—proving that AI isn't just the stuff of science fiction anymore. It's everywhere.
Never, in my years as a CIO for health systems, have I seen this speed of innovation. The tools and computing power available will enable our industry to do remarkable things that simply were not possible a year—or even six months—ago. Artificial intelligence will have a profound effect on our collective mission: the delivery of safe, high-quality, and compassionate care.
With any seismic shift, we must remain vigilant. As AI is rewriting the rules of our work, we need to hold onto healthcare’s core values, taking lessons from previous periods of transformation. How do we navigate this new frontier in a way that is ethically sound, and benefits both our industry and humankind?
Human Experience in the age of AI
The power of AI has begun to accelerate progress in healthcare across several key areas. First, it is revolutionizing how scientists develop new therapies, speeding up the entire process from drug discovery to clinical trials. Second, clinicians will be able to diagnose and understand diseases faster and more accurately, while being better equipped to treat each person, at reduced cost.
Third, on the day-to-day operational level, AI can assume the administrative burden that occupies so much time and energy. Less clinical documentation and coding. More automation for things like prior authorizations. Doctors and nurses can spend more time at the bedside, focusing on what they do best: caring for patients.
But perhaps one of the most immediate and wide-ranging applications will be in shaping and improving the Human Experience—helping us more deeply understand the challenges, needs, and preferences of our communities, meaning our patients, members, and employees. AI will enable our workforce to personalize care, build trust and nurture loyalty, and boost efficiency at the systems level. This translates to a competitive edge: better employee experience and health outcomes, higher reimbursements, and, ultimately, greater financial sustainability.
"First, do no harm”: Responsible AI in healthcare
While the use of AI has entered the mainstream, questions around safety and ethics have not gone away. Data integrity. Data privacy. Algorithmic biases. And other unforeseen hazards that may materialize. We must continue to let healthcare's fundamental principle—"first, do no harm”—be our North Star.
Hallucinations in AI are very real. When the computer perceives a pattern that doesn't exist—or worse, provides incorrect or misleading results—bad things can happen. Information about symptoms for diseases might be invented, citing nonexistent sources. Models built on real-world data can reinforce or exacerbate inequity. Instead of being a force for good, it can cause harm if left unchecked, undermining trust in technology—and healthcare as a whole. So how do we mitigate the risk?
If AI is the engine, then data is the fuel. Larger, more diverse datasets, oversight by humans with healthcare expertise, regularly auditing and validating for biases—all of these strategies can keep AI moving in the right direction.
Through our partnership with 41,000 healthcare facilities, Press Ganey has the industry’s largest integrated dataset for experience, clinical, and safety measures. While our database of 5.4 billion encounters, representing the unique voices of 312 million people, is unrivaled in size, what’s just as important is what it represents. People from all walks of life. Not seen through the prism of siloed interactions, but as individuals and the sum of experiences throughout the care journey. By connecting disparate feedback into one place, we can zoom out to identify broad trends, or zoom in to understand unique experiences at the individual level.
Healthcare-trained language models activate unheard voices
Organizations already collect mountains of data that can improve how care is delivered. Patients and members are phoning up call centers, providing feedback online (on review sites and social media), and interacting with providers in the patient portal. Yet much of this data sits silent—untapped and unused at the organizational level. Being able to distill this vast quantity of feedback—from multiple data streams—into clear and actionable insights helps paint a more detailed picture of behavior, needs, and emotions.
The ability to perform passive listening is powered by a type of artificial intelligence called natural language processing. While many language models exist, not all of them are created equal. Understanding the sentiment of patients and members requires fluency with healthcare concepts. Clinical abbreviations, like “SOB” for shortness of breath, are commonly seen in clinician notes. A patient description of a physician being “cool” is likely to be negative.
Models trained on healthcare-specific language and concepts can reduce misinterpretation and help improve the time from insight to action by getting the right information to the right person at the right time.
Ripples of change create tidal waves of innovation
As technology sprints forward, healthcare organizations need to take a strategic long-term view, with a unified platform that scales to realize AI’s full potential. The temptation might be to pursue lots of new ideas, but that could be a recipe to get a bunch of things started and nothing finished. Healthcare organizations that succeed in transformation prioritize focus over fragmentation.
As we embark on the AI frontier, collaboration will light the way. How to implement AI properly, measure success, establish guardrails, and ensure ethical applications across the industry. To navigate these challenges and more, we, at Press Ganey, are launching an AI collaborative that’s open for healthcare providers and health plans. Together, we can chart the course and champion the future of healthcare AI—a future that's not just technologically groundbreaking, but also ethical, safe, and beneficial for all.