Healthcare professionals and solution providers gathered last week in Las Vegas for the sixth annual HLTH Conference (HLTH23), a venue where over 10,000 attendees met to discuss and foster innovation in healthcare. Among the most significant issues facing the healthcare industry are chronic personnel shortages and rising costs, long-standing challenges for which many are looking for innovative solutions.

Customer experience, or let’s call it patient experience, should remain at the forefront of any solution aiming to mitigate the problems of personnel shortages and provider burnout. Recognizing these challenges, Google Cloud unveiled a suite of features within Vertex AI Search that hold promise for the healthcare and life sciences sectors. These features leverage generative AI (Gen AI) capabilities and medically-tuned search functionalities, making them well-suited to address some of the pressing issues facing healthcare delivery today.

The features and functions are exemplary of the “domain-specific” resources that will enable businesses and individuals in other verticals to provide better care and improve employee engagement. Note that the resources are not designed to replace healthcare professionals; rather they are designed to improve their ability to provide better care and to reduce stress. Note that it is presented as a search resource, rather than a co-pilot or automated assistant. It is there to respond to human-initiated queries.

Relief from Administrative Burdens

The integration of a Gen AI-powered search capability like Vertex AI Search, which is paired with Google Cloud’s Healthcare API and Healthcare Data Engine, empowers healthcare professionals to access a comprehensive range of clinical information with unprecedented efficiency. This, in turn, supports clinical decision-making, enhances the quality of care, and improves the patient experience.

One of the core applications of these new features is the reduction of administrative burdens. At its core, the technology within Vertex AI Search is designed to streamline the process of retrieving, analyzing, and summarizing clinical data. It can sift through vast troves of structured and unstructured information, including patient records, clinical notes, and data sources such as Fast Healthcare Interoperability Resources (FHIR). The result is a comprehensive and real-time view of patient information, available at the fingertips of healthcare professionals.

Patient-Centric Approach

The adoption of a solution like Vertex AI Search aligns with the broader objective of fostering patient-centric healthcare experiences. By simplifying data access and enhancing information delivery, healthcare organizations can personalize the patient journey, providing seamless, frictionless, and more customer-centric healthcare experiences akin to those expected from top-tier retailers.

Gen AI-powered search technology can add a new dimension in understanding patient data. These technologies have the ability to discern nuances within structured and unstructured clinical information, ensuring that healthcare professionals have a comprehensive view of each patient’s medical history, needs, and preferences. This personalized data access means that every interaction with a patient can be informed by a deep understanding of their unique healthcare journey.

HIPAA Compliance and Data Security

Crucially, Google Cloud upholds HIPAA compliance standards to protect patient data in healthcare settings. Robust security measures, coupled with each customer’s privacy controls and processes, ensure the safe handling of sensitive healthcare information.

Furthermore, Google’s medically-tuned large language model (LLM), Med-PaLM 2, adds a significant layer of expertise. While Med-PaLM 2 is designed to handle complex medical topics, Vertex AI Search caters to medically-tuned search capabilities grounded in the patient’s record. Together, these tools have the potential to empower healthcare organizations to find relevant answers to intricate medical questions, thereby enabling faster and more informed decision-making, thus benefiting both healthcare providers and patients.

Google Has Competition And Challenges

Google is not alone in developing GenAI-infused approaches to improve healthcare administration and patient care. IBM Watson, infamously, made the healthcare vertical an early focus for its branded cognitive services, investing something on the order of $62 million dollars to have IBM Watson act as a physicians assistant to take on the fight against cancer. In January, 2022, the “healthcare data and analytics assets” of IBM were spun out of Watson Health into a new company called Merative.

Microsoft also comes to mind, considering that the $19.7 billion it paid to acquire Nuance was thought to focus on clinical speech processing technologies capable of capturing and transcribing conversations between patients and healthcare providers. These included Dragon Ambient eXperience (DAX) for capturing conversations between patients and professionals, Dragon Medical One to replace note taking, and PowerScribe One for radiology reporting. OpenAI, Microsoft’s horse in the LLM race, does not have a medically-tuned LLM like Med-PaLM2; but it is in very good position to integrate a tailored generative AI resource into the workflows and talkflows of medical professionals, administrators and patients.

Another potential rival is Hippocratic AI, which came out of stealth in May 2023, placing emphasis on “safety” and relying on healthcare professionals both define the training material for foundation models and to provide reinforcement learning and human feedback of the model’s output.

The healthcare landscape challenged by workforce shortages and mounting costs, is ripe for Gen AI solutions like Google Vertex AI Search and Med-PaLM 2. Focusing on efficiency, privacy (in this case, HIPAA compliance), and patient-centric care, is relevant to all customer-facing verticals and we should take note of emerging service offerings, their successes and failures.

(Dan Miller, Lead Analyst with Opus Research, contributed to this post.)

Categories: Conversational Intelligence, Intelligent Assistants, Articles