For one, leadership wants to know the solution will garner a positive return on investment — be that financial savings or freeing up time for employees to devote to higher-value tasks. As another example, back and front office staff management will want to know how the new tool will affect their teams and if they’re teams will require training to learn new skills as a result of the AI implementation.
In this multipart series, Becker’s Hospital Review will examine AI in the hospital setting, including AI education, best use cases, seamless implementation and the benefits and opportunities machine learning holds.
For this installment, Becker’s caught up with the Customer Success team at Olive, an AI and robotic process automation company, to learn tips on how healthcare organizations can successfully implement an AI solution.
Here are eight considerations for healthcare organizations to weigh when implementing a new AI solution:
Before you begin working with vendors
1. Start with the end in mind by defining goals and setting key performance indicators. Without a heavy focus on KPIs from the get-go, there is risk that providers won’t recognize the immediate impact to their business or ROI.
2. Decide which process, or processes, to automate. The best tasks for automation are high-volume processes that tie up a lot of resources. To identify the right tasks to automate, organizations should ask themselves these questions:
- What processes require the most from your current employees?
- What areas of your business do you continuously fall behind on?
- Where do errors commonly occur?
- What processes are causing you to lose revenue?
- What work requires temporary employees to catch up on volume?
- Which tasks require the use of multiple systems, but the systems can’t talk or connect?
- Which processes can be automated from end-to-end with no human intervention?
3. Decide who from the hospital will lead the project. Olive recommends organizations take a hybrid approach to project management and bring all the key cross-functional players to the table. For example, the CFO knows the ROI the solution should achieve. The CIO knows the technology already in use at the organization and where it should look to invest next. Lastly, revenue cycle leaders understand the pain points in their processes, as well as which tasks are most burdensome and, subsequently, ripe for automation. Additionally, forward-thinking organizations have established AI steering committees to identify essential KPIs and goals for AI, as well as the best candidates in the business for AI assistance.
During the sales process
4. Work with the vendor to ensure the solution is compatible with all existing software and applications. This ensures no hiccups in the implementation and onboarding process for quick go-live.
After contract sign
5. Kick off the security review early. A security review should evaluate the controls that both the vendor and the healthcare organization have in place, or the controls that should be added, to monitor the environments in which the AI lives and protect sensitive data, such as protected health information.
6. Plan to execute. Olive laid out five of the most important steps for organizations:
- Enable the AI vendor to gain access to your systems as quickly as possible after signing the contract. They will need login credentials, such as usernames and passwords for all tools used in the workflows being automated. Any delay in assigning the vendor login credentials can cause big delays in implementation.
- Gather the data. Know how many days are spent in accounts receivable and other key metrics of success so both your organization and your AI partner are on the same page and can ensure quick ROI recognition after the go-live.
- Document your current state. All workflows and processes should be properly documented so the AI can accurately learn how to complete them.
- Set clear expectations for each member of your team so they understand implementation timelines and what is expected of them during and after implementation.
- Know your systems and when they update. Establish a notification plan so your AI partner can be notified and address system changes immediately when they occur post implementation.
7. Consider the possibilities. The biggest change for staff is the adjustment to no longer performing the same repetitive tasks they’ve always done. Instead, they will shift their focus to more complex work. With potential for new priorities, organizations should identify the tasks they’ve always wanted their team to complete but their staff never had time to execute, as well as what staff can add to the patient experience now that they aren’t focused on repetitive, yet important, tasks. Having a plan for how to refocus employees’ time will ensure a smooth and happy transition once the AI solution goes live.
Reap the rewards
8. Automation can help organizations prevent revenue losses and reallocate their workforces to tasks that require the human touch. Olive has already realized success at healthcare organizations across the U.S. through its product by the same name. For example, should organizations automate claims verification with Olive, they could spend one sixth of the time and capture nearly 3,000 percent more information.
Picking a partner for optimal implementation results
When considering or embarking on an AI implementation, there are many things to keep in mind, several of which were described in this article. Planning and preparing are key to ensure a smooth and successful go-live. The right partner will help your organization outline, define and work through these items, laying the path for a speedy and successful launch.
If you’re interested in connecting with Olive to understand its implementation approach, contact the company here.
This article serves as the third installment in a five-part series on AI in healthcare. For part one, “AI pushes into healthcare: 5 terms to add to your AI playbook,” click here. For part two, “Deploying AI for optimal impact: It’s more than a ‘brain in a jar,'” click here.
