Using Logistics to Put AEDs Where They’re Most Needed

Automated external defibrillators (AEDs) and first response care have a proven track record in saving lives after a heart attack, and with this in mind many communities purchase and make AEDs publicly accessible. But, are they placed effectively?

AEDs are expensive equipment and while some placement seems obvious—train stations, airports and bus terminals—the question of where next to put resources remains a challenge. Globally, researchers have taken up the banner to try to help cities identify other locations where AEDs can be deployed.

One such initiative developed by a Toronto-based team employed a mathematical equation and data on cardiac arrests collected through a joint U.S.-Canada effort called the Resuscitation Outcomes Consortium (Circulation 2013;127[17]:1801-1809). The Canadian Institutes of Health Research and the National Institutes of Health provided grants as  part of a combined effort to better understand and manage trauma and cardiac arrest cases.

The problem

As part of her grant, Laurie Morrison, MD, MSc, of St. Michael’s Hospital in Toronto engaged first responders and hospitals in the region to collect the data on cardiac arrest incidents and outcomes. In 2006, Morrison says, “The survival rate, initially, was horrible: it was 2 percent.” Morrison describes that survival rate as “almost hopeless.”

Part of the problem, they found after further analyzing the data, was bystander assistance. Bystanders performed cardiopulmonary resuscitation (CPR) or applied AEDs to assist patients in only 27 percent of cardiac arrests reported in 2006. Similar assistance rates were seen across Southern Ontario (Can Fam Physician 2010;56[6]:e213-218).

“The average response time in urban areas is five minutes,” she says. “Can you imagine standing there or sitting there and watching a loved one or a fellow human on the floor of a mall and doing nothing for five minutes?” This happened in seven out of 10 cases.

Minutes count. Survival is affected by 10 percent for each minute a cardiac arrest patient goes without treatment after collapsing (Circulation 2013;127[17]:1801-1809). Although emergency medical service (EMS) dispatchers were trained to coach callers through assisting patients while waiting for responders, dispatch stated in many cases, neither they nor the bystander knew where nearby defibrillators might be located.

“It’s a complete disconnect,” Morrison says. “You pour out some money and then you don’t tell anybody where you put [defibrillators]. You don’t tell the EMS dispatcher where you put them.”

In addition, when municipal and federal governments purchased AEDs, Morrison realized they weren’t using data to place the devices. They disbursed AEDs based on arbitrary criteria. While some places made sense, some, based on the data, did not.

Morrison figured there had to be a better way.

The solution

That’s where Timothy C. Y. Chan, PhD, of the University of Toronto, came in. Chan has worked with computer modeling in other areas of healthcare and is, as he described it, a classically trained mathematician.

“One of the questions they posed was if there was a way to identify cardiac arrest hot spots in [Toronto] and direct more resources to those hot spots,” Chan says. “When they described the problem to me, I started thinking about my field operations research mathematical modeling. Sounds like similar problems in different domains.”

To Chan, the problem was a matter of logistics. “This type of problem has been studied a lot in manufacturing, supply chain and logistics for decades. Companies have asked ‘Where do we locate our distribution centers to minimize transportation costs? Where do we locate facilities so that we’re closest to our customers who want to buy our products?’ A very similar idea came to mind when they described this cardiac arrest hot-spot problem. Cardiac arrest victims are our customers and AEDs are the companies providing services to the customers.”

The solution, as Chan sees it, is simply math. Collaborating with Morrison and other colleagues through St. Michael’s Hospital and the University of Toronto, he developed an equation to identify where AEDs were most needed in Toronto and the surrounding suburbs. Instead of hockey arenas and sports complexes, they found that in some communities additional AEDs offered more benefit on particular street corners (Resuscitation 2013;84[7]:904-909.)

The equation provided a way to map, geospatially, where deploying AEDs would do the most good.

Going forward

While no changes to AED placement have yet been put into place in the city of Toronto, the team is confident that this is a winning method. They hope to take these findings to the streets to improve bystander assistance rates and patient outcomes.

Globally, there’s a push to improve AED deployment. Similar models were developed using data from several cities including Osaka, Japan, Copenhagen and most recently, Paris. Presentations at the European Society of Cardiology conference in August included the Parisian study and a strategy to calculate mean distances between historical cardiac arrests and available assistance to determine where defibrillators should be placed.

The ability to learn from peers, identify shortcomings and apply successful strategies likely will help raise survival statistics locally and overall.

Morrison says, when they started looking at Toronto’s cardiac arrest rate, “We compared ourselves to Seattle and when our survival rate was 2 percent, their survival rate was 14 percent. The difference is that their bystander CPR rate is 60 percent. So, ours is 27 (percent), theirs is 60 (percent). The teaching point for me was that if you stand around and do nothing, people die.”