Early Wins in Integrating AI Into a Healthcare System

The number of potential applications for artificial intelligence (AI) in healthcare is growing by the day—but where should we concentrate our attention?

At Montage Health, a nonprofit healthcare system in California’s Monterey Peninsula area, our roots date back 100 years. We are reaching forward to technologies that support us in providing high-quality care into the future.

Like other healthcare systems, we are trying to take care of our workforce and stay competitive, and AI can help us do both. We are also focused on patient access, knowing that some barriers to care, like long phone hold times, can be reduced with AI tools. Finally, we consistently strive to improve financial performance.

Today, we are using AI in areas that include referral automation, revenue-cycle optimization, patient engagement (website chatbot), and clinical decision support (AI-assisted radiology). We are also incorporating two additional AI-facilitated improvements for our practitioners and patients:

• Clinical documentation: We are experiencing positive early results with the DAX ambient scribe.
• EHR search and extraction: We are closing both apparent and actual patient care gaps.

We have learned to establish metrics early and review them often—but not to believe that metrics will tell the whole story. As we plan, implement, and improve the performance of novel technologies in our healthcare system, we are gathering with our practitioners to hear their experiences.

Streamlining Clinical Documentation With an Ambient Scribe Tool

Our preliminary data on the DAX ambient scribe for clinical documentation looks promising. A pilot study with 25 practitioners in primary and specialty care found that DAX allows clinicians to focus on patients, knowing that the tool will capture relevant clinical information and draft a note for the practitioner to edit. While some patients express concerns about being recorded, many are happy to have their practitioners focus on them instead of typing notes.

DAX users are also saving time: One urologist is saving 50 minutes a day on crafting notes. An endocrinologist’s after-hours documentation work has decreased by 33 minutes per night. Although some practitioners maintain a preference for working with a human scribe or drafting notes themselves, many say things like, “Don’t take it away from me. Everyone should have this.”

Overall, I would say that our DAX pilot is a success. In fact, we have decided to expand the program to additional practitioners across our system.

Closing Care Gaps With EHR Search and Extraction

Closing care gaps is important to quality care and to earning star ratings from Medicare Part D, among other quality incentives. Our automation tool, Notable, scrubs the charts of patients who have an apparent open care gap—for instance, patients who appear to be missing a cervical cancer screening. Automation helps us close these care gaps in three ways:

1. Finding information about offsite screenings. Perhaps a lab report is filed under the media tab. If the automation finds that the apparent gap is already closed, it reflects that closure in the patient’s record. If not:
2. Asking the patient about offsite screenings. The automation reaches out to ask the patient, “Have you had your cervical cancer screening elsewhere?” If the patient says yes, we may be able to retrieve the lab report. If not:
3. Inviting the patient to book a screening. The automation asks the patient via text to self-schedule a screening via the online tool.

This system is saving us time while allowing us to reach more patients. We have applied this approach to screenings for colorectal cancer, breast cancer, cervical cancer, hypertension, and more. We also went live with annual wellness exam screenings.

Our colorectal cancer screening automation also produced good outcomes. Over a six-month period, we scrubbed 10,000 patient charts with open colorectal screening care gaps. Our outcomes for those patients revealed the following:

• The automation could close the care gap for 5 percent of patients, because information about patient screening was found elsewhere in the EHR.
• Another 5 percent of patients reported completing their screening at another location. We followed up, collected the labs, and closed the gaps.
• A final 5 percent of patients agreed to the screening and either came in for a colonoscopy or requested Cologuard.

In total, we closed 15 percent of the 10,000 gaps investigated. Closing those gaps has impacted 1,500 lives—especially the 500 patients who might have been lost to follow-up because they did, in fact, have an open care gap.

Our practitioners are happy with this system. Their gap closures are increasing, and they are feeling peace of mind and meeting quality incentives—without more work on their plates.

Secrets to Success

We have learned to start small and develop our metrics for success before we start a pilot. Other lessons learned include:

Go to Gemba: The Japanese expression “Genchi genbutsu” may be translated as “go and see,” meaning go to the place where the work is being done. “Go to gemba” is an English adaptation of the phrase. This can mean sitting elbow-to-elbow with practitioners, because analytics may not provide the entire picture.

Tell the Story: I talk with patients and community members about the work that we are doing—and not just the outcomes we are achieving, but the patients and practitioners who are affected. Visibility helps our executives and community members sponsor and support programs.

Keep the Horse Before the Cart: At Montage, we have solidified a governance strategy for AI as part of our existing structures for digital health—AI brings its own set of needs for policies and safeguards. As AI applications and use cases continue to multiply, we must stay ahead on governance now to avoid potential issues later.

Benchmark Against People, Not Perfection: AI systems may never be perfect—but neither will we. We need to benchmark AI-powered tools against humans and obtain results that are better than what human practitioners can achieve alone, rather than insisting that we cannot implement an AI tool until it is perfect.

Our thanks to Meg Burke Dingae, MHSA. Ms. Dingae leads Enterprise Human Experience (Hx) and Digital Health at Montage Health in Monterey, California. In her current role, Ms. Dingae is responsible for delivering on the expectations of all stakeholders—including patients, practitioners, employees, health plan members, and community members—to ensure an exceptional experience for all who engage with the Montage Health system. By leveraging innovative technologies like AI, she and her team focus on simplifying system navigation, creating more self-service options, and automating internal standard workflows so that practitioners and staff can spend more time with the patients in front of them, and less time on their computers. The Hx and Digital Health programs aim to create more equitable access to care while improving clinical outcomes for patients and the community as a whole.

The opinions expressed here do not necessarily reflect the views of The Doctors Company. We provide a platform for diverse perspectives and healthcare information, and the opinions expressed are solely those of the author.

---

The Doctor’s Advocate is published by The Doctors Company to advise and inform its members about loss prevention and insurance issues.

The guidelines suggested in this newsletter are not rules, do not constitute legal advice, and do not ensure a successful outcome. They attempt to define principles of practice for providing appropriate care. The principles are not inclusive of all proper methods of care nor exclusive of other methods reasonably directed at obtaining the same results.

The ultimate decision regarding the appropriateness of any treatment must be made by each healthcare provider considering the circumstances of the individual situation and in accordance with the laws of the jurisdiction in which the care is rendered.

The Doctor’s Advocate is published quarterly by Corporate Communications, The Doctors Company. Letters and articles, to be edited and published at the editor’s discretion, are welcome. The views expressed are those of the letter writer and do not necessarily reflect the opinion or official policy of The Doctors Company. Please sign your letters, and address them to the editor.