A payer-funded initiative hopes to use its scale to drive quality improvement in the emergency setting across an array of diverse EDs in Michigan. Aptly called the Michigan Emergency Department Improvement Collaborative (MEDIC), the group has identified low-value imaging as one area that is ripe for improvement across many of its sites. Now, the group is developing and implementing interventions that will improve practice in this area.

MEDIC’s research on imaging practices relied on a registry containing data on more than 1 million ED visits, 23% of which involved children.

Looking at the data collected through MEDIC from 16 participating EDs, investigators found that there is ample room for improvement in the way imaging studies are used according to clinical rules or guidelines pertaining to when specific studies should be conducted.

Reviewing all data for imaging tests conducted between June 1, 2016, and Oct. 31, 2017, investigators concluded that thousands of CT scans and X-rays could have been avoided, saving millions of dollars.

It can be challenging to push quality improvement (QI) across a range of EDs that are run by different healthcare organizations, serve different populations, and often are shaped by different cultures. But a unique payer-funded collaborative in Michigan is endeavoring to do just that. Established in 2015, the Michigan Emergency Department Improvement Collaborative (MEDIC) is seeking to identify opportunities for improvement among participating EDs and to drive practice changes that will improve outcomes and efficiency.

Among the group’s first targets for improvement is low-value imaging, an area MEDIC investigators have found to be ripe for improvement among many of its participating sites.1 Now, the group is using its scale to develop and drive quality improvement interventions at a much faster clip than might otherwise be possible in such a diverse group of EDs.

It is an intriguing model, one that MEDIC’s leaders believe can offer valuable lessons to health systems and EDs in other states.

The MEDIC collaborative includes EDs from across Michigan that have agreed to show their data to investigators and to work diligently toward quality improvement. Further, there are two major requirements for participation in the collaborative, according to Michele Nypaver, MD, MEDIC co-director. a professor of emergency medicine and pediatrics at the University of Michigan and C.S. Mott Children’s Hospital, both in Ann Arbor.

“One is [that] a site has to have a contract with Blue Cross and Blue Shield of Michigan [because] it is a funder of the program,” she says. “The second requirement is that we need a site to have a basic number of visits so that we have enough data from the site to be able to work with.”

The MEDIC team recruits members annually. The goal is to create a registry that represents the demography of the state, including people from all different regions, Nypaver shares. For instance, MEDIC’s research on imaging practices relied on a registry containing data on more than 1 million ED visits, 23% of which involved children.

Collecting meaningful data from the participating EDs certainly is a challenge, so MEDIC goes about the task in two different ways. First, the participating sites gather operational data such as how long patients spent in the ED, how they arrived, and similar metrics. “Those operational data points are in a data dictionary, which we ask our sites to map and give us on every single patient that visits their ED, not just the quality initiative cohorts,” Nypaver says.

In addition, MEDIC uses abstractors to explore specific patient records to pull up relevant data pertaining to a particular quality initiative. “For instance, if we are looking at CT use for head injuries, there will be abstractors at [the sites] who go into the ED records of the patients who are eligible and fit that cohort. They then will [collect] additional information from the charts and add that to the clinical registry,” Nypaver explains.

The idea of focusing on low-value imaging was made by MEDIC’s clinical leaders in partnership with the collaborative’s membership, explains Keith Kocher, MD, MPH, an assistant professor of emergency medicine at the University of Michigan and the other MEDIC co-director. “There is a lot of opportunity to do better around high-cost imaging decisions,” he explains.

The point is bolstered by the fact that professional organizations, such as the American College of Emergency Physicians (ACEP), have long recognized that several imaging tests sometimes are used in circumstances where they are not in line with consensus guidelines or the available evidence.2 This is important not only from the standpoint of cost, but also because providers do not want to unnecessarily expose patients to the potential harms associated with certain imaging tests. For example, there has been a push in recent years to limit the amount of radiation to which patients are exposed during CT scans.

Looking at the data collected through MEDIC from 16 participating EDs, investigators found that there is ample room for improvement in the way imaging studies are used according to clinical rules or guidelines pertaining to when specific studies should be conducted.

“Where there is high-quality evidence in particular use scenarios, we relied on that,” observes Kocher, spelling out how investigators determined when an imaging test was potentially avoidable. “For example, in children with minor head injuries, there [is guidance] that informs whether a CT should be done or not. This is called the PECARN [Pediatric Emergency Care Applied Research Network] rule.”3

Similarly, for adult patients, investigators relied on the Canadian Head CT Rule to determine whether a CT was appropriate in each case that it was used.4

Kocher acknowledges that there was not a comparable clinical decision rule for determining whether specific imaging tests should be used in some other clinical cases that were evaluated. However, there was enough literature and evidence for investigators to conclude that certain tests are overused. For instance, in the case of children who present with respiratory illnesses such as asthma, bronchiolitis, and croup, Kocher notes that X-rays typically are not required to reach a diagnosis.

Reviewing all the data for imaging tests conducted between June 1, 2016, and Oct. 31, 2017, investigators concluded that 1,519 head CT scans for minor head injury; 3,308 chest X-rays for children with asthma, bronchiolitis, or croup; and 4,254 CT scans for suspected pulmonary embolism could have been avoided. This translates into a potential savings of about $3.8 million per year, Kocher notes. Furthermore, researchers found substantial differences between the ED sites with respect to their imaging practices.

“Because there was so much variation that we saw across our sites, we know that some of them perform quite well ... we know it is possible to improve,” Kocher adds.

With these data in hand, the next step for MEDIC is to focus on its core mission, which is QI, Nypaver observes. “We set targets each year [for improvement] and we hold ourselves financially accountable, but the skeleton of the work is at the site level through quality improvement activity in order to drive improvements,” she explains. “Different sites are working on this in different ways.”

While MEDIC uses a framework for improvement that is based on the use of Plan-Do-Study-Act (PDSA) cycles, participating EDs design and leverage their own interventions to achieve collaborative goals. “We have instructed them in a small way about QI education at our collaborative-wide meetings. We teach them how to take these ideas and turn them into interventions ... and then understand whether they are working or not,” Nypaver shares.

Much of this QI work is ongoing. For example, Nypaver notes some sites are testing when and where to deliver guidance around the rules regarding specific imaging tests. The idea is to make it easy for physicians to absorb and understand this information at a time and place that is most effective at influencing decision-making. Other sites are looking at opportunities for shared decision-making with patients in cases where it is not clear-cut whether it is appropriate to conduct a specific imaging test.

“How can we standardize that process using evidence-based methods that are published in the literature?” Nypaver asks, alluding to a question that investigators at some of the ED sites are seeking to answer.

Over the next year or two, MEDIC leaders will be studying the specific imaging-focused interventions leveraged at different sites to see where improvement has been the greatest. “At the end of the day, we want to short-cut the sites’ quality work and figure out what is most successful at driving change,” Nypaver says. “We can [then] share [the successful interventions] with everybody in the collaborative so that they can apply these things internal to their own institutions.”

While not every successful intervention will necessarily work equally well at all the sites, Nypaver anticipates that ED leaders will be particularly interested in the tactics that have proven successful at the sites that are the most similar to their own departments.

Imaging is the focus of the most recent study, but MEDIC has other quality improvement initiatives in the works, too. For instance, the collaborative is developing alternatives to hospitalization for specific diagnoses or conditions that could be managed successfully in the outpatient setting. Kocher acknowledges that this work is challenging considering that much depends on what resources are available in the local healthcare environment and how care coordination can work within that locality.

“We are also moving to coalesce around uniformly working on low-risk chest pain because there is increasingly great evidence [regarding] how to risk-stratify this population,” Kocher explains. “One of the trends here is that many hospitals and EDs are moving to high-sensitivity troponin [assays], which also facilitate this kind of work. We think the timing is right to begin tackling this particular condition as an entire collaborative.”

MEDIC’s work shares many similarities to other QI efforts, but there is great power in the scale that the collaborative brings to the equation, Kocher explains. “We can learn ideally from everybody ... and shorten the cycle time to improvement,” he says. “But I don’t think it is otherwise different than what you could do locally with the tools at your disposal in general for QI.”

Kocher advises colleagues interested in developing their own approaches to QI to first develop a multidisciplinary team. It should be inclusive of all the people who would be affected by the specific care decisions that are a focal point of the effort.

“Have some regular cadence to the work so that you are adjusting in PDSA cycles,” he offers. Also, work hard to chart some early gains. “One lesson that we can share from MEDIC to all kinds of EDs is we are really over the hump now in terms of beginning to prove our value,” Nypaver notes. “I think that will help people tremendously. As you can imagine, all the different stakeholders in our various camps [need to know] why they should do something like this.”

While clinicians are most concerned with outcomes, the financial aspect always is a key factor in every conversation. Both administrative and clinical support are essential in any successful QI effort, Nypaver explains. “You cannot get physicians to the table unless they believe ... that they are doing the right thing for their patients. It is just a nonstarter [otherwise],” she says. “But it is a nonstarter for all the other stakeholders [when the financials don’t work].”

Nypaver adds that one reason why she is so interested and involved with MEDIC is because of its importance in pushing advances in pediatric emergency medicine beyond the confines of children’s hospitals. “The ED is the great equalizer. That is where everyone shows up in an acute emergency. Most children in the U.S. ... are not cared for in children’s hospitals. They are cared for in general hospitals,” she explains. “If you are going to do anything to improve children’s emergency care, you have to be at the table with the general emergency physicians.”

While all the research coming out of children’s hospitals is outstanding, it will not make a dent if that information does not land on the radar screen at general hospitals daily, Nypaver explains. “There has got to be some give and take. We have just not had the platform to do that in the United States. MEDIC is one such entity that allows us the ability to do that.”


  1. Kocher K, Arora R, Bassin B, et al. Baseline performance of real-world clinical practice within a statewide emergency medicine quality network: The Michigan Emergency Department Improvement Collaborative (MEDIC). Ann Emerg Med 2019; Jun 25. pii: S0196-0644(19)30356-7. doi: 10.1016/j.annemergmed.2019.04.033.
  2. American College of Emergency Physicians. Ten things physicians and patients should question. Updated June 18, 2018. Available at: http://bit.ly/2Zy4hle. Accessed Aug. 29, 2019.
  3. PECARN Pediatric Head Injury/Trauma Algorithm. Available at: http://bit.ly/2ZnsN9q. Accessed Aug. 29, 2019.
  4. Canadian CT Head Injury/Trauma Rule. Available at: http://bit.ly/2L2JE8w. Accessed Aug. 29, 2019.