“It isn’t working” is the easiest statement to make when there is an issue, but the statement itself may just be the biggest obstacle to solving the one it’s referring to. Imagine calling 911 and only telling the operator, “There’s a fire.” Yes, that is a real problem, but if anyone is going to solve it, they need much more information. Where is the fire? Is there anyone hurt or in danger? How long has it been burning? If those questions don’t get answers, that fire is going to keep burning. As obvious as this seems, the same (albeit less life threatening) impediment holds back practices every single day. They’ve gone through the effort of connecting with a new AI scribe service, and now “it isn’t working”.
Identifying a problem effectively requires specificity. With AI scribes, that means understanding the expected workflow for each clinician, not just the AI scribe itself. The AI scribe fits into the clinician’s workflow, not the other way around. Practice managers who know each step of the process will be much more successful at troubleshooting. So, when “the AI scribe isn’t working,” specify the issue using the workflow for reference. Is the error with the service itself or the EHR? Is the note inaccurate, loading in the wrong place, or not loading at all? Identify the expected outcome and document the error relative to that expectation. This will establish a baseline for consistent error tracking in the future.
Well documented issues lend themselves to solutions by indicating possible sources. A recurring problem that is localized to a single provider is completely different from an issue that all of the providers encounter. Pattern recognition is the key here. Using the error tracking from before, recurring problems should be grouped wherever possible. For example, “the physical exam section contained inaccurate information” can be easily grouped with other cases where note information was inaccurate. Cases where the note did not load at all should have their own group, and so on. These groupings will of course depend on the most common problems within the practice. Practice managers are particularly well positioned to identify these patterns by leveraging their access to centralized documentation. They are not constrained to navigating isolated issues because they are able to evaluate problems in their practice-level context.
The practice manager’s ability to not only know what the problem is, but what kind of problem it is, is an incredible advantage because it makes the resolution process significantly easier. When problems are identified but poorly routed there is precious time lost in the transfer between responsible parties. Service errors should be assigned to the appropriate service liaison for technical resolution. EHR issues that get sent to the AI scribe support team don’t just go unsolved, they waste more time. Some issues may not be appropriate for support tickets either. If service utilization and integration training can be managed efficiently within the practice, it may be less optimal to depend on vendor support. Establishing a workflow for problem solving streamlines this process and helps prevent backlogs of “I wasn’t sure how to fix this or who to talk to, so I stopped using it.” The only thing worse than an inefficient service is a service with zero utilization. Clarifying the support process makes it more efficient and more accessible to the clinicians.
A good 911 operator knows that each call warrants a specific response. Fire and EMS are needed here; animal control is needed there. Those decisions are made in real time, using plans determined before the call comes in. Like before, that operator depends on the information provided by the caller. A good practice manager navigates their issues in the same way, but they have the advantage of being able to train those ‘callers’. They can coordinate with their clinicians to standardize the support process and its documentation. The information that they collect over time enables them to determine the best plan of action is for each issue. Knowing the needs of the practice at the practice level, documenting those needs with intentional specificity, and using that information to map the support process ahead of time is a roadmap to “it’s working”.
Author Information:
Cooper Rock is an Executive Account Manager at ScribeBrain, an ambient scribe company focused on merging RCM with efficient clinical documentation. He coordinates with providers, practice management, and health organizations to communicate their needs to the engineering team, facilitating troubleshooting and product improvement. His involvement with active users and product development places him at the intersection of real-world provider needs and emerging technological opportunities.