Successfully piloting ambient AI for clinical documentation

Implementing new technology in organizations is like navigating a ship through uncharted waters – it requires a steady hand and a good map. When the hype is intense, and the stakes are high, understanding what’s worked for others can exponentially increase your chances of success.

This year, the topic of clinical documentation ambient AI was arguably the most talked about in healthcare AI. It is not only understandable but also advisable for healthcare leaders to test a solution before considering an enterprise-wide implementation.

Over the last 18 months, “Nabla” has been tested in over 80 organizations. We have had the privilege of working with pioneers in organizations such as Children's Hospital of Los Angeles, Kaiser Permanente, Stratum Med Clinics, Carle Health, and many others, all of whom have been exceptional partners. These partnerships have given us valuable insights into specific requirements and challenges.

During this time, we have worked side-by-side with CMOs, CMIOs, CIOs, and their teams. For each of these pilots, we made it a priority to:

1. Define shared success criteria: Collaborating on defining the success criteria for each organization.
2. Understand clinical workflows: Understanding our users' unique clinical workflows and how nAbla could seamlessly fit into those workflows.
3. Provide tailored support: Provide support through customized training, onboarding, and feedback loops with each of the pilot participants.
4. Track success and adoption: Track the impact and adoption of our product so that our partners can make informed decisions about the next steps for implementation.

Success factors for pilot projects

From the outset, it is crucial to focus clearly on the metrics that will determine the success of the project. This includes working closely with the provider to develop a shared vision of the desired outcomes and to define strategies for achieving these metrics. This includes both qualitative and quantitative metrics that are continuously evaluated during the pilot project.

You should also clarify with your provider which data can be used to measure success. It is recommended that you integrate existing data sources such as the EMR or other relevant systems to enable comprehensive and accurate evaluation.

Evaluation phases should be scheduled at regular intervals to allow you to analyze progress and make adjustments to the project plan. Clearly communicating the key performance indicators to all parties involved is another key factor. Only when all project participants know which metrics are crucial and how they can contribute to them can you ensure that they receive the necessary answers.  

In summary, planning, continuous monitoring and regular adjustment are key elements to ensure the success of AI solutions. The early involvement of the relevant stakeholders and a clear focus on measurable results create the basis for a successful implementation, both in the pilot project and in the entire company.

‍

★ Pro-tip: We usually recommend that at least 50 people take part in a pilot for organizations over 500 providers, for several weeks.