Two years ago, every health conference I attended had multiple clinician burnout groups. This problem is well known, and actual system-level contributors have been summoned.
Last year, burnout solutions were splashed on every conference app login screen: Generating artificial intelligence (AI).
Labor productivity statistics have led some to declare the United States “at the forefront of productivity prosperity.” Extensive investment in artificial intelligence has strengthened economists’ optimism about the “roaring 20s” of worker productivity on the horizon. But without the accompanying systematic and organizational action, this will not be done in healthcare, rethinking what we are financially motivating, how new technologies are integrated, how tasks are transferred, and how labor is prepared.
AI copying revolution
In healthcare, environmental documentation tools have become the star of the show. These “AI scribes” listen to patient-physics conversations, transcribe the discussion, and then use Generative AI to create a clinical note for the first draft. These solutions eliminate the laborious work of fully capturing patient stories or doctors’ ideas about plans. Some people see this technology as a miracle. Before the popularization of large language model chat applications, many doctors thought that such a solution was impossible in our careers.
Professional doctors like me started our careers, setting up templates in Electronic Health Records (EHRs) and typing wildly, while our patients told us intimate stories about what led to their scheduled visits. So having draft notes that are simply edited and marked for us feels like the core burden is lifted. This may be why demand for these products has increased dramatically over the past few years. The nationwide chief medical information officer is the requirement (not requirement!) for clinicians to provide such solutions. For many, these tools will come home faster at dinner, reducing the time spent on EHR and reducing the cognitive burden of trying to remember the time patients are hearing to when they can enter or indicate to the right place in the EHR.
This environmental documentation technology is one of the first solutions to blister from grassroots end users as a key tool to support current clinical practice. This comes with the last 15 years of EHR and other digital add-ons launched from organization leaders or suppliers to find the next unicorn.
The long-awaited unicorn is here, it’s real.
Contradictions on the Impact of Artificial Intelligence
Organization leaders are shocked by the future that AI may create. There is endless administrative inefficiency that can affect patient care or inflated costs. According to the American Hospital Association, Labor accounts for nearly 60% of the average cost of hospitals in the United States. If there is any way to make clinicians more productive – but measurable, then the hospital leaders are doing it.
However, not everyone is ready to embrace the entire unicorn without having to know where the unicorn comes from first.
Frontline clinicians are still very alert. Their concerns stem from the technology itself, and from what organizations or system leaders do with the new efficiencies gained from deploying such technologies. These next steps – following policy changes and operational measures implemented by a broad AI Scribe – will be a key part of determining its success.
Is the organization just going to continue adding more patients and more tasks to the doctor’s plate? Will we insert generated AI solutions into existing clinics that may not serve the ideal workflow for clinicians? Will we integrate these generated AI tools into an EHR system that does not support the doctor’s mindset or the desired storytelling purpose?
Or, now that we are confident in the performance of AI documentation solutions, will we take a step back and rethink the current paradigm of how clinicians can spend their time? Now that I know that the initial draft notes will be very accurate, I can spend more time focusing on the patient’s difficult stories and digging into the root causes of health, rather than clicking on the keyboard to complete the basic data entry task while talking? Can I browse the patient’s schedule effortlessly so that I can stay on time with later patients, keep a reasonable “lunch time” and have chart time to really focus on communicating the complex reasoning behind my diagnostic and therapeutic decisions?
Systematic solutions to maximize the potential of AI
These are system-level decisions and interventions, where AI magic will be implemented. Policies that transfer payment models to value-added and improve health outcomes can further reward patients with increased visits. Redesigning efforts in organizational care ensures that multidisciplinary teams use the best AI to operate in their license priority to serve their populations most effectively. In higher education, baking AI literacy into basic education programs for all health workers can encourage clinicians to use AI to enhance their work. AI can handle low-value tasks – checkboxes, data capture, management steps – and a trained workforce does its best: connect and communicate with patients.
A well-doing organization or manager will better recruit and retain trained health workers, improve responsiveness to patients, and potentially reduce costs for all stakeholders. Health systems that set the right payment incentives will leverage the benefits generated by AI will ultimately make the economy more productive and safer when better population health ends.
We must take advantage of the current opportunity to ensure this enormous expansion of AI-driven efficiency and gain access to information to access underserved communities in a way that bridges equity gaps rather than worsens the differences.
Finally, research efforts should rigorously examine the impact of these solutions on clinician well-being and patient outcomes. Without close assessment and reliable evidence, we risk implementing solutions that seem promising but fail to provide meaningful improvements or have unintended consequences.
An overall implementation strategy needs to be formulated
Yes, burnout among doctors has improved. But less than half of the doctors felt burned was not the reason for the celebration. The growth of the global health workforce is not enough to meet the rising demand for health care driven by demographic and epidemiological shifts. Family Medicine – a core primary healthcare major that can provide a large return on public health investment – remains unpopular as a career path.
For a “productive boom” in healthcare, timing couldn’t be better. AI tools are here. Marc Benioff predicts the AI-driven wave of productivity; we have a great option to work in healthcare. By pulling new policies and operational levers, systemic actions will ultimately determine whether we really take advantage of this game-changing ability or whether we spread the drivers of clinician burnout and poor population health
Photo: Pixelembargo, Getty Images
Travis Bias, DO, MPH, FAAFP is a family medicine physician and deputy chief medical officer for Solventum Health Information Systems. He is co-director of the Comparative Health Systems Program at the University of California, San Francisco Institute for Global Health Sciences. He has 15 years of experience in multiple clinical settings and is currently practicing telemedicine.
This article passed Mixed Influencer Programs. Anyone can post opinions on MedCity News’ healthcare business and innovation through MedCity Remacence. Click here to learn how.
