Building AI to Treat Every Patient as an Individual

In this three-part series with Robert Laumeyer, we’ll be delving into the topic of artificial intelligence (AI) in healthcare and how payers and providers are applying the technology to help streamline heavily manual processes such as prior authorization. Robert is an entrepreneur and software inventor with more than 30 patents, who has dedicated his career to bringing innovative technologies to fields in need of improvement.

At Availity, Robert uses his vast knowledge, experience, and passion to build transformative AI systems that help provide personalized care to patients. Below is part one of our series, detailing Robert’s unique perspective on the challenges and opportunities presented by AI in healthcare, as well as his journey in creating new solutions to tackle important industry issues.

Q: What inspired you to enter the healthcare sector as a technologist?

A: My journey into healthcare was sparked by the stark contrast between the advanced technology I encountered daily in fields like finance and the Internet, and the sheer lack of transformative technology available in healthcare. I found it alarming that an industry as critical as healthcare lagged so far behind in utilizing cutting-edge solutions, so I wanted to help make a meaningful impact.

I lack the expertise of a healthcare provider in providing bedside care, but I excel in managing data and information flow. I have done so in various technology fields, including navigation, geoinformatics, space, and embedded software development. I wanted to bring that level of personalization to healthcare.

 Q: You are focused on prior authorizations. Why?

A: The problem of prior authorization and utilization management are my current primary technology focus areas. Although prior authorizations are critical to ensure that patients receive necessary care, current processes are fraught with frustration and inefficiency for healthcare providers and payers alike. The primary source of this frustration lies in the manual processes and analog technologies that underpin it. These elements currently render prior authorizations among the most time-consuming and burdensome transactions in healthcare, ultimately risking adverse effects on patient care. It is a perfect area to address with cutting-edge AI technology.

The prospect of improving this process was particularly appealing because I saw it as an opportunity to help people when they’re most in need of help—when their health or that of a loved one is at stake.

Q: How should AI work in healthcare?

A: My approach to AI in healthcare revolves around treating patients as individuals rather than statistics or members of a cluster. The emphasis is on recognizing discontinuities or critical shifts in health data that indicate a change in a patient’s condition—those crucial points where health data takes a significant turn. Many artificial intelligence techniques ignore those subtle variations, and some will even try to smooth them out to fill them in with other data to fit better with the cluster or curve, but unlike industries where smooth trends suffice, healthcare requires pinpointing these inflection points accurately, capturing those pivotal moments when someone transitions from healthy to unhealthy. This level of precision is crucial for the best patient outcomes.

Consider the scenario of a cancer patient who visits the ER due to a broken arm. Are they a cancer patient or a patient with a broken arm? The answer is that while this patient has an extensive history of cancer treatment, their immediate concern is the broken arm. Healthcare providers differentiate that easily. They may do different things in treating that patient to ensure infection risk is low, for example. AI must be capable of discerning this as well, just as a healthcare provider would, while considering the patient’s complete medical history.

Our goal should be to treat patients as unique individuals, like doctors and nurses do, rather than grouping them into generalized clusters or values in a regression model. The key is ultimately creating supporting AI technologies that recognize the extensive diversity among human patients. This extends beyond cultural differences to include individualized aspects such as medical histories, lifestyles, genetics, diet, location, and more. The goal is to enable personalized treatment tailored to each patient’s unique circumstances.

Navigating patient data diversity requires a multi-faceted approach. I tell my engineers to treat the data like it’s their mother’s. Would you want your mother’s data going into the process that you’re doing right now? It’s always vital to maintain a high ethical standard when handling sensitive information.

Furthermore, ensuring data diversity involves more than cultural considerations; it’s about capturing the intricate nuances of patients’ backgrounds, habits, and health journeys. The diversity of patients makes healthcare AI very difficult but also very important. This understanding is pivotal to building AI models that offer accurate and individualized insights.

Q: What is the ultimate goal of your work in AI for healthcare?

A: The ultimate goal of my work is to empower healthcare practitioners with tools that treat each patient as an individual to facilitate personalized, effective, and timely care—resulting in improved patient outcomes and experiences.

Accomplishing this requires a deep understanding of the field’s unique challenges, backed by transparent and trustworthy AI systems that ensure healthcare providers can trace the reasoning behind AI-generated recommendations. By embracing these patient data complexities and upholding the highest accountability, traceability, and transparency standards in our AI tools, we can pave the way for a more advanced, streamlined, and patient-centric healthcare system.

Visit our website to learn more about Availity AuthAI or to schedule a meeting with our team!