The 2019 Innovator Awards Co-Second-Place Winning Team: Bon Secours Mercy Health

March 26, 2019
The pairing of an analytics platform with changes to EHR tools has been part of a broader data-driven approach to reshape providers’ opioid prescribing behaviors

Every year, millions of Americans use opioids to manage pain, and while sometimes this is the appropriate medication, problems arise when patients begin to rely on opioids. This overreliance has helped contribute to the worst drug crisis in American history.

In 2017, the Department of Health & Human Services (HHS) declared the opioid epidemic a public health emergency, and the federal health agency has estimated that more than 130 Americans die every day from opioid-related drug overdoses. What’s more, in 2016—the last year numbers were available—11.4 million people misused prescription opioids, according to HHS.

Across the healthcare ecosystem, clinical and IT leaders are quickly learning that making health-system-wide changes to address the opioid crisis can be quite challenging if they lack actionable data about prescribing patterns. And for those healthcare professionals in states where the epidemic has struck even worse, the harsh reality has hit even more quickly. That’s why for the 2019 Innovator Awards Program, Healthcare Innovation recognized two health systems that are truly on the leading edge of using technology to help combat the crisis. Both of these enterprises—Bon Secours Mercy Health and Allina Health—earned co-second-place recognition in this year’s program.

The new Bon Secours Mercy Health system is the result of a recent merger between the Maryland-based Bon Secours and Cincinnati-based Mercy Health systems—which now operates 43 hospitals across seven states. It is the Mercy Health system, whose work in serving patients who are battling opioid addiction, that has received special attention; Ohio and Kentucky, two states in the Mercy network, are both among the top five states with the worst overdose rates in the nation.

How it all began

It was not long ago when the health commissioner of Cincinnati’s Hamilton County received a call from the White House to spark a discussion on what could be done to reduce opioid-related deaths in the county, as the statistics there were some of the worst in the U.S. Stephen Feagins, M.D., vice president, medical affairs, Mercy Health East Market, was looped into the conversation, and soon after, an initiative blossomed—a driving force between the Mercy Health group and the community providers who are providing services to individuals with opioid dependencies.

What quickly ensued was a process of bringing the right people to the table to talk about providers’ prescribing behaviors and treatment, which then led the health system’s leaders down a path to look at early identification of at-risk patients in its emergency departments (EDs) and acute care hospitals, as well as what tools were available in the EHR (electronic health record) to help facilitate things, explains Mark Binstock, M.D., CMIO (chief medical information officer) at Bon Secours Mercy Health.

Speaking from the 30,000-foot level, Binstock says the initial intel was brought back to the collective group, and efforts to address the dependencies then commenced. Tools were created that aimed to improve the identification of at-risk patients, encourage early treatment referrals, establish acute withdrawal protocols, and provide more efficient access to Ohio’s Prescription Drug Monitoring Program (PDMP) database. As officials at Bon Secours Mercy Health point out, the core ambition of their work would be to reduce the opioid prescribing rates across the health system. By doing this, “we reduce the future risk of opioid dependency, overdose and death,” they attest.

The goals of the project, according to its senior leaders, were to first accurately measure and report on outpatient opioid prescribing behavior at the patient, provider, office, specialty, regional, state and enterprise levels. Second, there was a need to implement informatics changes to reduce inappropriate and excessive opioid prescribing in terms of potency, dose, frequency, duration and quantity of opioids used in the outpatient setting. And third, the project set out to reduce two key metrics by at least 10 percent: the percent of opioid orders that exceed 30 morphine equivalent daily dose (MEDD, a conversion method used to translate the dose and route of each of the opioids the patient has received over the last 24 hours to a parenteral morphine equivalent); and the total morphine equivalents prescribed normalized to patient volume.

The motivation behind these goals was clear: experts, including the CDC (the Centers for Disease Control and Prevention), have come to the conclusion that it’s highly unusual to develop heroin addiction, opioid overdose and opioid overdose death de novo. Rather, there is an antecedent event. “These patients are generally being given opioids for some legitimate purpose in the healthcare system, be it a dental extraction or orthopedic injury,” explains Binstock, who adds that some people—particularly adolescents—are very susceptible to opioid misuse. “They continue to experience pain, and they seek longer durations and higher doses of opioid therapy, and some of them get that through the healthcare system. But some encounter roadblocks, and they turn to illicit markets; oftentimes, it’s cheaper that way.”

As such, he attests, “There is a recognition that prescribing behavior [patterns] in the healthcare system need to be altered in terms of the strength of opioids and the duration of opioid therapy, and trying to substitute non-opioid therapy for pain control. That’s our underlying academic and guideline premise. We are trying to urge providers to be conservative in their opioid prescribing patterns, for both acute and chronic pain,” Binstock says.

Technology at the forefront

Various IT tools were involved in the health system’s ambitious endeavor, perhaps none more important than an opioid analytics platform built on a database that permits evaluation of opioid prescribing behavior at the order, provider, department, specialty, market and enterprise levels. Frequent monitoring of this “data cube” enables the organization to track progress for two key opioid performance metrics, including the MEDD limit for acute pain prescriptions and opiate burden, officials note.

The platform specifically takes the EHR data for all the prescriptions that were put into the system from the EDs, and other Mercy Health facilities that use Epic, such as its ambulatory care practices, and from that data, a forward-facing tool for the end user emerged, explains Brian Latham, director of pharmacy at Mercy Health St. Rita’s Medical Center. As such, this enabled an array of folks throughout the system—such as administrative staff, quality directors, and physician leadership—to pull that data and look at the prescribing practices based on specialty, regions, or hospitals, and then compare and contrast how providers were prescribing differently, Latham says.

Another critical IT element in the project was integrating the PDMP data into the provider workspace, as it enabled clinicians to get a patient-level view, not only of what Mercy Health providers were prescribing, but also what the patient was picking up from essentially any provider in the U.S. Although PDMP use varies by state, with different mandates and technology vendors complicating matters, Mercy Health was able to pull about 40,000 daily reports that come from U.S. pharmacies submitting data to the various state PDMP databases, says Anna Lendl-Hancock, application coordinator at Mercy Health.

“Before we implemented this one-click view of all the [state] PDMPs, a provider or staff would have to log into different state or government [databases] to pull all those reports,” recalls Lendl-Hancock. “Our team integrated with a [vendor] that pulls data not just from Ohio, but multiple states, so with one click a provider could [now] view all the opioid prescribing history for that patient, as it’s required at the end of every day for the pharmacies to report all of the opioids prescribed to these programs,” she says.

Meanwhile, one-click ordering functionalities have also been implemented in the EHR, with low starting points that give prescribing clinicians the ability to type in the drug, and immediately all the defaults are filled out for them to make an appropriate prescription decision, explains Nick Waggamon, application coordinator, Bon Secours Mercy Health. “With that, there are a number of alerts we put in place to help drive them to appropriate durations and dosages,” he adds.

These changes to provider tools and workflows did not come without complexities, however, notes Kelley Recker, EHR vice president operations and training, Bon Secours Mercy Health. Interdisciplinary informatics committees—both in ambulatory and inpatient operations that are representative from each of the health system’s markets—were created to understand what changes were being made and why. Sometimes the committees would even vote on the proposals, Recker adds.

The results of all this work have led not only to a changed culture in prescribing behaviors, but also in profound reductions in inappropriate and excessive opioid prescribing. Some of the tangible evidence of these improvements from 2017 to 2018 include: a 17 percent reduction in total opioid orders as well as in the rate of opioid orders to all medication orders; a 21 percent reduction in the rate of opioid orders to patients; a 14 percent reduction in the rate of MEDD greater than 80 to all opioid orders; a 43 percent reduction in the day supply of opioids greater than seven days; a 25 percent reduction in the rate of MEDD greater than 30 for acute opioid orders; and a 30 percent reduction in the total morphine equivalents prescribed per patient, adjusted by patient volume.

As noted earlier, the initial goal was to reduce these latter two metrics by 10 percent; those targets were blown out of the water. Also importantly, says Jedediah Tuten, director of pharmacy operations at Bon Secours Mercy Health, “We further went and looked at lot of our inpatient-based order sets and determined that oftentimes, we didn’t need to put significant pain management [on the order set].” So for example, for a pneumonia order set, clinicians began to pull off opioid-containing drugs from those order sets, and that has contributed to about a 20 percent reduction in the amount of opioids used within the inpatient stay, he notes. “We feel that since those drugs are not on the patient’s profile [anymore], they are [more likely] not being prescribed on discharge, and that’s especially important.”

In the end, the project’s leaders attest that the unique pairing of an analytics platform with changes to the EHR tools made available to the provider ordering opioids is all part of a broader data-driven approach to reshape provider ordering behavior.

The Mercy Health team accentuates their intense motivation to combat the opioid epidemic, pointing out that overdoses have supplanted car crashes and gun violence to become the leading cause of death for Americans younger than 55 years old. The epidemic has killed more people than HIV at its zenith, and its death toll surmounts those of the Vietnam and Iraq wars combined. Funerals for the young have become ubiquitous. Every 11 minutes, another life is lost, they reference. 

“We are in the midst of the largest drug epidemic in U.S. history, with 11 [opioid-related] deaths per day in Ohio and typically about two per day in Southwest Ohio,” says Tuten. Waggamon adds, “Most of what we tried to do was simply to make it easier for [providers] to do the right thing.”