Changing Patient and Provider Behavior Through Tech with Thomas Maddox of BJC HealthCare

This is part of our weekly executive insights series where Elion CEO Bobby Guelich speaks with healthcare leaders about their tech priorities and learnings. For more, become a member and sign up for our email here.

Organization: BJC HealthCare

Can you give a quick overview of your role at BJC?

My primary job is leading the Healthcare Innovation Lab, a collaboration between BJC Healthcare and our associated School of Medicine at Washington University. We focus on care delivery innovation, often involving digital health solutions, to improve care for patients in the St. Louis area.

How do you approach identifying and pursuing innovations?

We’ve developed a framework to identify care delivery problems. We consult physicians, nurses, administrators, back-office teams, patients, and families to uncover issues. Then we evaluate those issues against four key questions:

1. Does solving this problem align with our system’s strategic priorities?

2. Is the problem verified by data, not just anecdotal?

3. Does it require a new solution or simply process improvement?

4. Are the right stakeholders willing to drive the change?

Using this process, we’ve assessed over 500 ideas, leading to about 100 pilot projects.

How do you measure success? And looking at those 100 innovation projects you’ve taken on, are your solutions generally internally developed or externally sourced?

Usually we run these pilots for 1 to 3 months, and we evaluate them across four metrics: adoption, satisfaction, clinical benefit, and financial or operational ROI.

Looking at whether those pilots have been internal or external, almost all of our innovations rely on external technology. I would estimate we customize existing platforms like Epic about half the time, and the other half come from new outside vendors.

What categories have you explored?

The three primary areas we’ve looked at are:

• Analytics and AI: Predictive modeling and generative AI for care delivery support.

• Digital point-of-care tools: Clinical decision support and voice assistance for providers.

• Care in the home: Remote patient monitoring and tools to support patient-care team communication.

Can you share examples of successful innovations you’ve worked on over the last year or so?

There are a lot of cool things we’ve been able to do. A few examples include:

• Predictive analytics for palliative care

• Improving the sleep experience of hospitalized patients

• Using virtual nursing to supplement bedside nursing.

• Remote patient monitoring at home for heart failure and pulmonary conditions like COPD.

• Generative AI for documentation

Let’s focus on virtual nursing and remote monitoring at home. Can you elaborate?

Remote monitoring is a little more technology enabled, whereas virtual nursing is a little bit more about reimagining workflows, although by definition you have to have the AV connection to do it.

One remote patient monitoring pilot we did involved ambient sensing for heart failure patients using a pressure sensor under their mattress. It collected data like weight, heart rate, and respiratory patterns without requiring active patient involvement to improve patient adherence. The passive data collection helped us detect issues earlier, such as slight respiratory rate increases indicating potential future problems.

In our virtual nursing pilot, we used AV systems to allow virtual nurses to handle admission interviews of patients being admitted to the hospital. Admission interviews typically take a bedside nurse anywhere from 45 minutes to an hour, and are often interrupted by other patient needs. This pilot freed bedside nurses to focus on tasks that only they could do (e.g. starting an IV line or administering medications), reduced interruptions during the admission interview, and improved patient satisfaction.

Patients appreciated the undivided attention, and virtual nurses could input information from the interview directly into the EHR in real time so the full care team could access it immediately. The pilot saved about 20 minutes per patient per shift, providing significant time savings on med-surg floors. We did that pilot in 3 different hospitals, and now we’re expanding that to additional units within those hospitals, as well as additional hospitals that were not part of the initial test.

What are you excited about as you look toward 2025?

I think AI will continue to transform documentation and administrative tasks. I also think there is a big opportunity to reimagine clinical workflows using behavioral nudges. For example, gamification or lottery-style incentives could improve patient adherence to medications or lifestyle changes. Similar principles could drive care teams to adopt evidence-based practices more consistently. By aligning workflows with behavioral insights, we can make better outcomes the default.