Karen Conway provides an interesting analysis on SupplyChainBrain.com about the potential and application opportunities of AI in the healthcare supply chain. The below provides a summary of her writings and in particular provides the list of promising applications she sees in the supply chain.
Predictive analytics. One of the more exciting applications of AI is the use of genomics, combined with other patient clinical, social and behavioral factors, to predict future disease states and healthcare treatments, such as whether a patient is likely to experience a cardiovascular event or need a knee replacement. At the individual patient level, there is relatively little downstream relevancy for supply chain. But consider what could happen if we had data on entire patient populations — say those served by a healthcare system or accountable care organization. Could that help predict the types and volume of products that will be needed, including when and where, while providing valuable demand signals to manufacturers and distributors?
Demand matching. With more data on how products perform in routine clinical practice and the drive to redesign care pathways based on the needs of specific patient populations, there is an increasing need to match the right product to the right patient. AI can play an important role in understanding what works best on what kinds of patients, and leverage this data for value analysis and sourcing, as well as making sure the right products are in the right place.”
The above is an area where MUUTAA has developed a specific interest and is applying skills through a number of projects we have developed with key actors in the healthcare (Rx) supply chain. Our mission is to bring the right medication to the right patient with the least disruption.
Logistics optimization. AI-enabled companies focused on patient flow are utilizing tools commonly deployed by third-party logistics companies, such as UPS, to chart the fastest ambulance routes to transport patients to the hospital or other care delivery sites. Why not deploy these same technologies to assist healthcare supply-chain professionals as they grapple with the migration of care outside the acute care settings? AI can help determine the best transportation methods, frequency and routes to move both products and caregivers to the rapidly expanding number of locations where they will be needed, from home and retail clinics to urgent care and ambulatory surgery centers.
Supply continuity. Recent events — from natural disasters and infectious disease outbreaks to product recalls and sterilization facility closures — have increased attention to the challenges incurred by disruptions in supply continuity. Unlike the retail industry, where a backorder is often only an inconvenience, supply disruptions in healthcare can have grave consequences.
AI dependencies. As with many new technologies, there is considerable excitement over what AI can do to improve clinical, operational and financial performance, along with patient and clinician experience. At the same time, there is relatively little discussion about what needs to be in place to ensure that AI delivers on its promise.
One of the most underestimated areas is data governance. The beauty of AI is that it can analyze large amounts of data to identify patterns and hidden correlations that would otherwise take humans considerably longer to decipher, if at all. It also allows users to feed the AI engine with a wide range of variables, even those you only suspect might have some bearing on the problem you’re trying to solve. But despite the sophistication of the tool, the old adage — garbage in, garbage out — still applies. Before starting an AI initiative, make sure you have enough data (likely drawn from different sources), and that the data adheres to well-defined data policies, standards, definitions and processes.
Finally, consider to what extent you want to utilize AI to augment decision making, i.e., whether to allow the system to provide insights and recommendations while a human still makes the final choice, or to fully automate decision making. The magic and mystery of AI is the lack of transparency in how the system makes decisions, because it’s continually learning and changing how it selects, weighs and relates different variables to reach conclusions. Only once you have trust in the system — especially when dealing with patient care decisions — should you move to applications of AI in which the system makes decisions and takes action without human intervention.
The potential for AI and machine learning in healthcare is awe-inspiring, especially as we consider how to harness the rapidly expanding wealth of knowledge that is generated every day. On the other hand, there is still much to learn about how best to apply AI across the various aspects of healthcare. As we aspire to new heights, guided by AI, it’s important to remember the foundation upon which AI is built. Are your AI initiatives based on accurate, complete, standardized and normalized data? If so, then the sky, seemingly, is the limit.”
Thank you, Karen, for this great write up outlining the areas of consideration in order to embark on a successful AI journey in the healthcare supply chain. MUUTAA can help.
Image: dogtownmedia.com