Efficiency in hospital operations can tackle length of stay — how AI can help

As hospitals move toward value-based care, reducing length of stay has become a top priority. The national average length of stay for hospitals significantly decreased from 9 days in 1990 to 5 days in 2014, representing a 44 percent drop, according to CMS. However, to stay competitive and to lower costs, experts say there's more work to be done.

In a webinar sponsored by software company Qventus and presented by Becker's Hospital Review, Diane Karagory, managing director of Huron Consulting Group, and Mudit Garg, founder and CEO of Qventus, discussed how approaches that leverage artificial intelligence can help hospitals tackle length-of-stay issues while also reducing staff burnout.

"With value-based reimbursement tied to quality, obviously providers have to focus on ensuring quality of care and improving outcomes," Ms. Karagory explained. Huron Consulting Group, a professional services firm, works with clients such as hospitals to develop goal-oriented strategic plans.

"Also, in order to effectively manage financial margins, we need to effectively use our resources and manage capacity," she added.

The struggle to streamline patient flow

The key to reducing length of stay is proactively detecting problems related to patient care and hospital operations, according to Mr. Garg. However, many frontline staff are already overworked.

"We have smart, hard working, incredible frontline managers and staff," Mr. Garg said, commenting on processes he's seen in emergency departments during his career.

"When managers had the time and the capacity to proactively catch problems before they became big, the department ran well, it ran smoothly, it ran with high quality," he added. "But these managers are being asked to do so much, with [fewer resources] over time."

A key resource care teams lack is sufficient time to focus on each individual patient. "You have a lot of patients in the unit, have 15 minutes to 30 minutes to do the rounds, and you're spending less than 30 seconds to a minute on the patient," he said.

This lack of time proves a major challenge for hospital teams, since care coordination — which often starts with hospital rounds — is a foundational element of length-of-stay management.

"We recommend daily interdisciplinary rounds that focus the care team on the plan of care," Ms. Karagory said. "They can proactively set discharge dates, they can break down any barriers in a timely fashion and also, importantly, [engage] the patient and their family."

How AI addresses patient discharges & reduces excess days

Mr. Garg described one medical-surgical unit he saw with growing lengths of stay. Upon investigating the issue, hospital officials learned a hospitalist and a nurse manager at the facility had been spending an additional hour after each shift reviewing patient charts and addressing operational holdups, such as calling radiology to check on orders.

However, after six months, the two-person team was burning out, according to Mr. Garg. "We can't continue to rely on the heroics of frontline teams and rely on them going above and beyond every time," he explained. "Our teams are burning out and they don't have the time and energy to look through data."

For Mr. Garg, the answer to staff burnout is AI. If technology systems are tasked with data-driven activities, frontline staff are "free[d] from the minutia to really think about more challenging, more difficult problems that are coming up," he said. In particular, he said hospitals can leverage AI to improve patient flow by identifying operational challenges.

With Qventus, an AI-based software platform focused on healthcare operations, hospital staff are able to use AI to predict and prevent bottlenecks that delay discharge. One aspect of the Qventus system is a chart that sorts patients by diagnosis, length of stay and estimated discharge date, among other criteria, to help staff coordinate patient care plans during rounds. A particular category of note is "discharge barriers," which encompasses challenges like access delays or pending lab orders.

"When they select a test as being critical [for] this patient getting discharged, and if action is not taking place, it fires off a coordination, an escalation, in the background," Mr. Garg explained. After flagging a particular test as a barrier, the AI system delivers a text or email notification, called a "nudge," to the selected care team.

Through this early warning system, Qventus helps frontline staff prioritize specific processes to resolve discharge barriers before the next day's rounds, according to Mr. Garg. "As you look to AI [systems] … don't look to them to replace what your staff's doing or to eliminate what your staff's doing, but to really augment them," he added.

To request a copy of the webinar email info@qventus.com.

© Copyright ASC COMMUNICATIONS 2017. Interested in LINKING to or REPRINTING this content? View our policies by clicking here.

 

Top 40 Articles from the Past 6 Months