Behind the Scenes of a Bon Secours Mercy Health Co-Development Partnership

Accrete Health Partners, the venture arm of Bon Secours Mercy Health (BSMH), is working with a startup company called Atalan on accelerating the development of a new platform module designed to bridge clinician retention and patient experience data, helping health system leaders improve both workforce stability and patient outcomes.

Atalan has created what it calls a Clinician Retention Intelligence platform, helping health systems predict and prevent surprise clinician resignations up to 12 months in advance. Using real-time, objective data that health systems already have, Atalan says its machine learning models pinpoint at-risk clinicians, uncover the causes, and tailor proactive interventions.

Healthcare Innovation recently sat down with Cyril Philip, vice president of digital ventures at BSMH and Accrete Health Partners, and Tiffany Chan, CEO and founder of Atalan, to talk about their partnership. 

Healthcare Innovation: We are always interested when health systems have venture arms that can make an investment in a company but also pilot it internally. 

Philip: At Accrete, we leverage what we call acceleration funding to help drive co-development. Acceleration funding is a separate pool of capital from our direct investing, and it's specifically focused on projects where we have an operational sponsor. In this case, our medical group physician recruitment team, as well as our market leadership, were really excited about the work that Atalan had done for them, and are looking to expand the relationship further. They recognize the problem that physician turnover has been and what Atalan could bring to a table. We launched this initiative in 2025 so it's relatively new. There are a couple of other companies that we've done this with. But frankly Atalan has been our first and fastest-moving partner.

HCI: Tiffany, can you describe how your platform works? 

Chan: What we find is that a lot of health systems rely on surveys to understand the pain points of their clinician population. A survey comes with a lot of its strengths, but it also has its weaknesses. You can only deploy the survey once a year, and the response rate is very low, like in general, it's 35% nationally. And then there are survey non-response biases. So there are a lot of silent sufferers among these clinicians, people who are struggling a lot, but they’re not responding to surveys, so you're not really uncovering their needs. Instead of relying on survey responses, we look at the operational data. How many patients did you see in a day? How much time are you spending on the electronic health record, typing up notes? All of these factors around workload and work environment actually impact and can predict where the struggles might be in the clinician population. We have a platform that taps into these operational data that health systems already have, and we use that to parse out where the signals of silence sufferers are, and identify them for the operational leaders at health system so that they can proactively support them before they get really burned out and leave the organization. 

HCI: There’s a lot of work that goes on around EHR optimization, and efforts like the KLAS Arch Collaborative, working on trying to make the EHR less onerous for clinicians. Is that where a big chunk of the problem is — too much data entry and feeling kind of overwhelmed by note taking? Or are there a lot of other elements too?

Chan: The beauty of Atalan is we look at very comprehensive data, and we actually find that for our health system partners the risk is very distributed. The root cause is not just EHR usage, but it includes workload, the kind of work environment, the number of breaks you take. We pick up more than 100 different departure triggers that can lead to people burning out and leaving. The EHR usage and optimization is part of the picture, but only a part. The picture is a lot more complicated and comprehensive, which is why it's important to look at data holistically the way we do.

HCI: So is AI involved in this in some way? 

Chan: I tend to shy away from the term AI, because I feel like it tends to reference something that we're not. We are using machine learning, which is connected. We learn from historical data, past behavioral patterns from clinicians who have left the organization. What were they struggling with? We are looking at those operational data 12 months before they leave, and then using machine learning to pick up on the patterns for existing clinicians, and that's how we can predict with 70% accuracy up to 12 months ahead of time.

HCI: Does the company also offer solutions for how to help the health systems retain those clinicians?

Chan: This is where a lot of the collaboration with the operational leaders of Bon Secours Mercy Health comes in. When it comes to identifying and defining action, there are two pieces to it. The first piece is understanding what are the departure triggers for this specific clinician or department. The second piece is understanding what can be done about that trigger. We not only surface the why, we also identify triggers that are actionable, right? Certain things are outside of their control. There is no point in surfacing that because they can't act on it. So we focus on the triggers that are within their control. They tend to be more operational in nature. Is this person spending a much higher time at work than is healthy? Is it because you're not efficient in scheduling your patients? That's where we are collaborating with operational leaders, and Bon Secours has really been a fantastic partner in this.

Philip: In terms of the ability for Atalan to use such large data sets from different data sources — that has been a really valuable tool for us, because from a health system standpoint, it's impossible to catch all of these potential stressors that are out there on our own or manually or even through dashboards. It's a challenge. Atalan essentially provides our market and on-the-ground leaders with proactive, actionable data that they can use. And this goes back to your AI point as well, in that the most valuable AI tools are the ones that actually generate value and actionable insights. It's one thing just to have a ton of data, but if you can't actually bring useful insights, it's not very helpful.

HCI: I understand that you're working on accelerating the development of Atalan PX, which is a module designed to bridge clinician retention and patient experience. Could you talk about why it's important to bridge those two things and how you'll do it?

Philip: When we look at the Atalan PX module, it's really essential to bridge the gap between workforce stability and  patient experience. Historically, health systems have looked at those things separately. This is our patient experience data and this is our workforce data. Actually, there’s a lot of value in connecting those, because they're essentially deeply linked. A well-supported physician is a foundation for good patient interactions, right? If you have a physician who's suffering or challenged, the patient experience is going to suffer and and perhaps the other way around as well. With Atalan's expertise in pulling all these data sets together, we can make sense of some of that noise that's out there. It's still early on in the development, but we're very excited about the potential.

Chan: For health systems, their No. 1 goal is to make sure their patients are healthy and have access. But if your clinicians are not healthy, or if you don't have enough clinicians because of high turnover, it becomes very hard to fulfill that goal of serving patients. This is why we're so excited to partner with Bon Secours to start pulling these very disparate siloed data sets together for the first time and look at them holistically and say OK, what are some common things that trigger both clinician struggles as well as patient issues, so that operational leaders have a very clear action plan and a targeted way to help both populations.

Philip: One of the exciting things about the PX module is that we collect a lot of patient feedback and experience data, but it's often pretty siloed. If  you're able to essentially correlate all of these patient signals and then bridge it to the workforce data that Tiffany had mentioned, it creates a much clearer picture of the total clinical environment. This will enable our market leaders to see exactly where the friction is that's affecting both the physician and the patient. We can then make those targeted interventions that can improve care delivery. 

HCI: Cyril, can you talk about some positive early results you saw? I understand you are expanding it to additional markets. 

Philip: We don't comment on individual market results, but I will say the pilot had a meaningful enough impact to make our leadership move forward in expanding it across the system. I would say the pilot really did validate that Atalan provides a level of essentially proactive, real-time insights into clinician wellness that we just didn't have before. We knew we were getting something that we couldn't do on our own. And the feedback from our frontline leaders is that these actionable signals allowed them to intervene early and really support our physicians, which is our goal. We're trying to make our physicians’ lives better.

HCI: Anything else you would like to add?

Philip: Success for Accrete means helping Atalan build a product that other health systems use as well. We're essentially a testing ground. That’s our co-development model. Eventually this will expand to Atalan’s other customers. We see that as the true metric of success, from the Accrete standpoint, is that we've co-developed a product that works well within BSMH, and that other health systems would find valuable as well.