Dr. Corey Scurlock MD, MBA is the CEO & founder of Equum Medical.
Seemingly overnight, artificial intelligence has caught fire in healthcare, with feverish interest from boards, the C-suite, medical staff, information technology and revenue cycle management — you name it, and there’s an algorithm in the works. Fueled by the flashy arrival of OpenAI’s ChatGPT and its newest large language model, GPT-4, the race is on among IT vendors and care providers to use AI to improve everything from triage to patient communications to treatment to discharge. The dream is that AI can reduce costs, make work so productive that fewer staff are needed, triage patients faster and ensure they are cared for in the right setting.
At least for now, much of the promise of AI in healthcare is a dream deferred. AI systems require stupendous amounts of data and must be trained to recognize patterns in medical data, understand the relationships between different diagnoses and treatments, and provide accurate recommendations that are tailored to each patient. This learning, either guided by a clinician (machine learning) or working on its own (deep learning), takes time.
Integration challenges loom large, including “maximizing AI’s capabilities in EHR systems and providing opportunities to grow their potential and improve overall global healthcare.”
As a physician entrepreneur, I am actively exploring ways to incorporate AI into my company’s telehealth-enabled clinical services. AI-powered algorithms can analyze vast amounts of patient data, enabling physicians to make more informed decisions in real time. Our goal is to empower physicians with the tools they need to deliver efficient, personalized care while optimizing patient flow and outcomes. By embracing AI, organizations can drive positive change in healthcare and shape a future where technology and human expertise work hand in hand to revolutionize the patient experience.
Early intervention is crucial in healthcare, as it can significantly impact patient outcomes. AI algorithms can continuously monitor patient data, including vital signs, lab results and patient-reported symptoms, to identify potential warning signs. By leveraging predictive analytics and machine learning, I believe AI could help physicians identify high-risk patients who may require immediate attention, enabling timely intervention and potentially preventing complications or hospitalizations.
I have written previously about the promise of telehealth for easing staffing shortages and overcoming “hot spots” in the patient’s journey through the health system—care transitions that cost every stakeholder through delayed care and worse outcomes. Already, some experts say AI can “analyze patient flow, wait times and time spent doing certain tasks” to ensure staffing levels are appropriate.
Telehealth physicians can access electronic records and clinical decision support systems and have the technological chops to work with AI systems as they mature. As healthcare professionals, we face mounting administrative burdens that often impede our ability to dedicate sufficient time to direct patient care. AI tools such as natural language processing and intelligent automation could assist with tasks such as documentation, coding and scheduling, freeing up valuable time for physicians to focus on what they do best: providing personalized care to patients. This could not only enhance physician productivity but also contributes to reducing burnout and improving satisfaction for patients and caregivers.
Even more promise may be found in AI’s analytical prowess. An article in British Dental Journal explains that AI can “analyze large amounts of patient data, such as medical records, imaging studies and laboratory results, to support clinical decision making and improve patient outcomes.” Clinical decision support in current electronic records systems has not always been warmly embraced by care providers.
Recently Microsoft, which invested billions in OpenAI, said it plans to bring GPT-4 to Epic’s electronic medical record. The initial use cases will involve patient communication and data visualization.
In medical imaging, one paper explained that “in many applications, the performance of machine learning-based automatic detection and diagnosis systems has shown to be comparable to that of a well-trained and experienced radiologist.” Medical diagnostic facilities could see fewer errors and misdiagnoses, reduced operational costs and faster lab results.
There remain significant concerns about AI’s projected impact on organizations and people. Just as driver-assistance tools don’t mean people should take their hands off the steering wheel and take a nap, algorithms should be used to augment clinicians’ decision making rather than replace it entirely. Doctors and their patients should always have the final say in care.
At the same time, data being fed into AI programs can retain or further inherent biases from past treatment decisions. Many systems may not have enough “Ns” on particular minority groups to take into account these populations’ unique healthcare needs.
There are also governance issues. Most health systems have formal IT structures to reduce duplicative systems, standardize care and maintain patient privacy. Anyone who has followed AI knows it makes mistakes and can even make stuff up. We need to be able to answer questions such as: Will AI make IT more or less governable? Who decides where and how it is used and for what purpose? Will we choose the best technology and just roll it out, or should we wait for evidence-based practices in the literature?
Properly regulated, I believe AI has nearly limitless potential for helping solve some of the biggest challenges facing healthcare. New efficiencies should lead to greater productivity. For example, if an AI algorithm could analyze a patient’s vital signs and medical history to predict the likelihood of a cardiac event, healthcare professionals, whether onsite or remote, could instantly get that patient into critical care.
I believe AI in healthcare will be transformational, improving patient care from triage to treatment to discharge to final billing. We must also recognize that these advanced tools come with new challenges. When OpenAI’s chief executive asks Congress to regulate AI, it’s probably time to install and maintain the guardrails needed to keep us, humans, in charge.
Forbes Business Council is the foremost growth and networking organization for business owners and leaders. Do I qualify?
Read the full article here
