How M Health Fairview transformed its discharge processes with predictive analytics

Angie Stewart - Print  | 

Improving operations and efficiency is the No. 1 strategic priority for hospital executives, with many focusing on length of stay, according to 4sight Health's 2019 hospital executive survey.

During a Jan. 30 webinar sponsored by Qventus and hosted by Becker's Hospital Review, M Health Fairview Vice President and Chief Nursing Officer Julie Sethney, MSN, RN, discussed how the hospital addressed length of stay and discharge processes without creating additional burden for providers. Imaad Rashied, a customer success expert at Qventus, explained how the company's solutions played a role.

Identifying process variation across the health system

About two years ago, Minneapolis-based M Health Fairview found that the problem of excess lengths of stay was only compounded by a merger, according to Ms. Sethney. Various hospitals were carrying out their own performance improvement initiatives, creating silos without effecting sustainable change. Processes for discharge planning and bed planning ranged widely across hospitals as individual nursing departments adopted their own variations.

All these disparities culminated in utilization issues. For instance, one hospital with an overburdened ICU was transferring incoming patients out of the M Health network — when another of the system's hospitals 20 minutes down the road had the capacity to take them.

"We were acting like several independent hospitals versus a single system," Ms. Sethney said. "We were putting the burden on our patients to navigate the complexity of our system instead of making it easy to seek care at M Health."

Out of these problems arose two opportunities for the inpatient setting. The first was to standardize discharge processes. The second — what Mr. Rashied called the partnership's "capstone" — was to create a centralized system operations center that would enable the entire organization to communicate with the same language and reliable, real-time data.

Starting with each site: Improving discharge process management

M Health Fairview began using Qventus' Pathfinder discharge management application to run multidisciplinary discharge rounds and manage length of stay. The system's charge nurses and care transition teams could rely on the tool to proactively identify which barriers to prioritize during their shifts and ensure consistent communication through hand-offs.

"It also allows our staff to work with families to explain what's remaining in the discharge process and how they may be able to help, really allowing them to help own the care of their loved ones alongside us," Ms. Sethney said.

As the M Health team generated a mass of data from the rounding process, Qventus used it to train machine learning models, Mr. Rashied said. This, in turn, will help clinicians predict discharge dates and barriers to discharge.

Qventus supplied reports breaking down the barriers to discharge, the frequency at which they were occurring and the associated avoidable inpatient days. The M Health team homed in on physical and occupational therapy, seeing that those areas were linked to more barriers.

Whereas before, the therapy manager would see a laundry list of patients facing therapy as a barrier to discharge, now the Qventus platform uses machine learning to make it clear which patients' therapy orders must be prioritized to avoid delaying discharge.

"The goal here is to automate some of this work so folks can get back to delivering the high-value care they want to," Ms. Sethney said.

The change was a relief for front-line inpatient staff, who previously spent much of their day tracking down updates on patient care and discharge planning via manual, back-and-forth calls. Plus, daily meetings ensure that leaders are held accountable for sustaining these changes and meeting their goals.

Centralizing system operations

As for the second goal, setting up a centralized operations center, M Health started with co-locating various bed placement teams. The three large teams that were serving M Health's 10 hospitals were consolidated at one location: the system operations center.

The organization then hired a full-time flow analyst to proactively track and manage bottlenecks throughout the system.

"That role is important for us because in our previous state, it was everyone's job to pay attention to operations and the potential capacity constraints," Ms. Sethney said. "Now that we have a full-time dedicated resource for this, we're hoping to unlock a lot more operational efficiencies as we mature our system operations center."

Now, through its partnership with Qventus, M Health can gain visibility into real time hotspots across the health system and use machine learning to predict census to identify anticipated bed demand. With these capabilities, M Health can proactively solve problems and propel efficiencies across the system.

For organizations that want to effect similar meaningful change, Ms. Sethney shared a key piece of advice.

"Reducing length of stay is important. It's also important to ensure our patients stay within the system and … ensure our staff and physicians are brought into that mission," she said. "As we are asking our staff to buy into our mission, we also have to work, as leaders, to help them do the work differently and to provide them the tools to do so efficiently and without creating additional burden.

"If we do all the above well, it'll have the net effect of reducing complexity for our patients. Then, our patients will be able to trust that when they're going to an M Health hospital or clinic, they're getting the same high-value care, regardless of where they are in the system."

To listen to the webinar, click here.

More articles on patient flow:
Wake Forest Baptist Health to close Lexington hospital's labor and delivery unit 
Advocate Medical Group to close 7 Chicago-area clinics
Pennsylvania hospital banned from reopening after racking up 40 citations

© Copyright ASC COMMUNICATIONS 2020. Interested in LINKING to or REPRINTING this content? View our policies by clicking here.