As AI use cases in healthcare continue to expand, the priority list for most IT organizations is dominated by areas where they want to leverage automation to improve existing workflows. But AI and automation initiatives shouldn’t just be at the top of the priority list – they need to be a fundamental part of everything on the list, from top to bottom. In other words, healthcare organizations need to adopt an automation-first mindset where they question how and where to use automation as part of every initiative.
It’s not as easy as it sounds. You can’t simply decide to prioritize automation and expect it to naturally integrate into every decision. There are both technical and cultural barriers to overcome in the evolution to an AI-enabled enterprise, and in healthcare, especially, it’s a long-term shift that requires thoughtful strategy and planning. That’s why it’s imperative to start now. Every other organization is, and you don’t want to be last.
By adopting and embracing an automation-first mindset, healthcare organizations can become more agile, optimize operations and uncover opportunities for innovation.
Where to Start Implementing an Automation-First Mindset
Many healthcare organizations have already started the shift towards automation for administrative tasks, deploying AI for productivity across scheduling, patient engagement, clinical documentation, electronic health record (EHR) management, prior authorization, and more. There is a plethora of data being collected, and since it is generally well defined, it can be used by both payers and providers alike in their robotic process automation (RPA) tools. For example, digital intake forms are standardized and allow health systems to automatically create and update EHRs, eliminating administrative waste and allowing physicians to spend more time with patients. Automatically aggregating and integrating additional data shared through chats, emails and other interactions into the EHR can add exponential value that gives payers, providers and patients a big-picture view of their healthcare journey. And it can all be done without any manual data entry.
The next phase of automation is a bit more difficult. Organizations should look to implement automation in systems that can complete more abstract tasks and reduce human intervention where the data isn’t as well defined. For example, if a provider’s data set is structured differently than a payer’s, a human has to intervene to move the process forward. This is a perfect scenario where agentic-led systems can bring immense value. Agentic systems have the ability to automate more complex administrative tasks, such as retrieving a file, querying a system or navigating unstructured data. This gives valuable time back to employees who were once dedicated to such manual tasks.
But healthcare is governed tightly, and for good reason, so deciding which tasks or projects can be automated should be a careful process. In healthcare, we can’t just throw AI at the problem; decisions around automation and the risks involved must be carefully considered.
One major benefit to using agentic-led systems is that they don’t have to expose regulated data or PHI to any outside entity or large language model (LLM), like a generative AI solution might be doing. Looking for areas like this where agentic AI can be put to work while keeping sensitive data contained will help advance automation projects more quickly with fewer roadblocks. However, to get moving even more quickly and show progress with this next phase of AI and automation, it’s still best to look for applications in non-clinical areas where rules and regulations aren’t as tight.
How to Start Building an Automation-First Culture
For IT professionals, the technical barriers to achieving success in any project are naturally top of mind. But in most cases, addressing cultural concerns will be just as important. A shift in mindset, or any major organizational change initiative, regardless of size and scope, is always going to feel like a daunting task. But it doesn’t have to be.
The first step to achieving an automation-first mindset is to get buy-in from the leadership team. It’s likely that senior leadership may need some convincing on exactly how automation and AI will solve their biggest challenges. With that said, it’s best to start with an attainable timeline for a phased rollout – clearly articulating how the new tools will reduce costs, save time, and/or alleviate clinician burnout while improving patient outcomes and member satisfaction.
With senior leadership fully invested, you can then consider which processes should be automated first. Keeping in mind some of the technical hurdles around PHI and regulatory requirements, start consulting with key stakeholders to identify workflows that are most hindered by tedious manual tasks and what changes will have the greatest benefit on cost and care. You’ll also want to think about which teams are most open to adjusting operations and adopting new workflows.
Taking all these elements into consideration, set that path forward with attainable goals for applying AI and automation to low-risk activities that can offer high-reward by reducing human intervention and boosting productivity. Starting small to ensure successful execution will ease the transition and lay the foundation to establish an automation-first mindset across teams, whether clinical, business or technical.
Keys to Success
I’ve led or been involved in many transformation initiatives at leading tech companies, and in my experience, the focus must be on the people leading the change just as much as the new technology itself. At this point, AI is weaving itself into the bedrock of nearly every industry, so naturally it has a prominent place on the ever-expanding to-do lists of executives who want to keep up. But again, it can’t be viewed as points on a list; it should be viewed as a change management strategy and a fundamental part of everything on the list.
The ultimate goal is creating an AI “muscle memory” across all departments and functions where users continuously ask themselves how every task in their day could be reimagined with AI. It won’t happen overnight, but with education, encouragement and a thoughtful roadmap for implementation, it will create more agile and efficient operations across the healthcare industry. With AI as part of the solution, making substantial progress towards the long-stated goal of bending the healthcare cost curve while improving care and experiences for patients and members is more attainable than it’s ever been before.
About Robert Duffy
Robert Duffy is Chief Technology Officer at HealthEdge, where he is applying his extensive background in product development and engineering to drive digital transformation internally and for the 100+ health plans that rely on the company’s technology to power their most important value streams. Rob previously served as the Chief Product and Technology Officer at Drizly, an Uber Company, where he played a pivotal role in scaling the company’s product and engineering teams post-acquisition. Prior Drizly, he held technology leadership roles at Salesforce.com, Amazon, and Time Inc.
