The Potential for Artificial Intelligence to Improve Healthcare Supply Chain Management

By Matt Stewart, CEO, RiseNow
Twitter: @WeAreRiseNow
Twitter: @RiseNowLLC

As artificial intelligence (AI) continues to evolve into a must-have technology in almost every industry, healthcare organizations have continued to expand and accelerate their AI strategies. A recent Optum survey found that more than half of healthcare executives expect AI will begin delivering tangible cost savings within three years at about a 90% increase – which is certainly a lofty goal.

But what’s missing from this projection is a more substantive discussion about what is actually required to ensure that AI can deliver on its cost-saving promise – most notably, where AI is being applied within healthcare to ensure money is saved and investments are returned. When you stop and think about how AI is being applied outside of healthcare, there emerge implications and opportunities for the supply chain, from demand planning, forecasting, and predictive cost analysis. While that seems like a futuristic topic for healthcare at the moment, there are several real-world, practical strategies you can incorporate into your organization right now to ensure a thriving supply chain.

Supplier Management and Discovery

Managing suppliers through differentiated portals in an effort to capture clean, reliable supplier data is an incredible challenge. However, through implementing AI technology, health systems can bring supplier data together to ensure the health system maintains a diverse group of suppliers to ward against disruptions.

Using AI to track and manage this data helps organizations of all sizes gain the insights into their suppliers, assess risk, and ensure supply chain diversity – three things that, prior to the advent of AI, would require significant time, money, and manpower. Now, AI tools can be employed to clean and present data in digestible ways, enabling health systems to, at a glance, see exact numbers on where their suppliers are, and what kind of shifts they may need to make as a result.

Better Invoice Details

Capturing at or near 100% of invoices electronically at the line level has been a real challenge for health systems for decades now, and while solutions to these issues continue to be released on a regular basis, they consistently overpromise and underdeliver. The best example is the often-used Optical Character Recognition (OCR), where a health system has to map very specific prescribed zones for each individual supplier invoice template and instruct the software to input information according to what data was in each zone. The next iteration of OCR had to do with keywords and pattern matching – a bit more efficient, but still clunky and prone to error.

As you can imagine, the maintenance and upkeep of this model was incredibly time-consuming – suppliers changed invoice layouts, layouts were inconsistent, and onboarding new suppliers into that system was a headache. This demonstrates the fact that our supply base is a living and breathing organism in a continual state of evolution and change.

What we’re starting to see, coming into the invoice capturing space, is the application of AI combined with natural language processing and a machine feedback loop. Invoice capture rates are higher through AI, and the feedback loop created through the machine learning process enables the machine to learn which invoice values humans should process and validate so the AI learning mechanism can continue to build its knowledge base and get faster at capturing those invoices, and even routing non-PO invoices to the appropriate parties for approval.

By introducing AI into your invoicing process, your organization can begin training these new softwares to standardize invoice processing – leaving more time and space for other, complex tasks that only humans can do.

Virtual Item Master

AI is harnessed in cleansing, enriching, and creating a single source of truth of global item data. By connecting this powerful virtual item master to supply chain and clinical systems, accurate information is available to the right people at the right time. Health systems’ newfound clean data will provide efficiencies and cost-saving benefits.

These include scanning barcodes for utilization, identifying recalled items, quickly resolving backorders, and leveraging an automated substitute or clinical equivalent list to prevent future backorders from occurring. In addition, AI identifies standardization opportunities, purchase quantity optimization, and proper reimbursement coding. By utilizing AI, health systems spend less time reviewing data and more time rationalizing their supply base. Deeper insights into purchases and suppliers will allow health systems to prepare and remain agile during the next crisis.

Transparent Risk Reviews

The passive mechanisms that AI can apply to a myriad of data sources can also aid in assessing a health system’s financial health without requiring direct interrogation or input from suppliers, which is certainly a benefit for those seeking to understand where every penny of their supply spend is going.

It’s all about being able to trust your data; but if you’re relying solely on the information that’s being shared with you by suppliers, it’s typically too late to avoid disruptions to your supply chain. After all, those suppliers have a vested interest in dusting anything negative under the rug and taking a “there’s nothing to see here” kind of approach. AI can provide insights into your supplier base that will prompt you to ask the right kinds of questions to gain an understanding of where your supply chain may be falling short and what you can do to set things right before it becomes a crisis.

With the ever-evolving challenges healthcare organizations are facing today, it’s no wonder many have turned to AI to automate problem-solving across key manual, labor-intensive areas such as supplier information, invoicing, data cleansing, and supplier risk reviews. As AI continues to play a critical role in these areas, it’s worth asking yourself: are you incorporating AI in the right places in your supply chain?