The Conversation That Can Change the Course of a Cardiac Arrest

Triptych of close up of eye, hands delivering CPR, mouth
Cavan / RUNSTUDIO / Getty / The Atlantic

The call came in at 7:42:02 p.m. on March 21, 2019.

A man in his early 60s had just sat down to dinner with his daughter and her boyfriend at an otherwise empty North Brooklyn restaurant, when he suddenly slumped in his chair. The daughter shouted at a hostess to call 911. Within seconds—by precisely 7:42:16, according to my review of the incident—a New York City Fire Department emergency-response unit had acknowledged the assignment, and would arrive on the scene some two and a half minutes later. In the meantime, a dispatcher stayed on the line.

“Is this for you, or someone else?” the dispatcher asked the hostess.

“For someone else,” the hostess replied.  

“Is the person breathing?” the dispatcher asked.

Confusion. Was the man having a seizure? Before long, it was established that he was not seizing and was unconscious. He had no discernible pulse. The dispatcher instructed the daughter and boyfriend, both in their 30s, to ease the man down to the hardwood floor, belly-up, and expose his chest.

The event was one of the more than 350,000 out-of-hospital cardiac arrests that occur annually in the United States. They are a leading cause of death, and only about one in 10 victims survives. Without early 911 access and cardiopulmonary resuscitation (CPR)—the first two links, followed by early defibrillation, in the out-of-hospital “chain of survival”—death is certain.

Recommended Reading

Over-the-phone CPR instruction by a dispatcher, also known as telephone CPR or T-CPR, can enable a caller to become a lay rescuer, and by doing so make the difference between life and death. Early CPR performed by a lay rescuer is associated with a roughly twofold increase in the chances of survival.

However, T-CPR is not as widespread as most 911 callers might expect. I would know. The boyfriend in this story? That’s me. The man was my girlfriend’s dad, Todd. For him to have a shot at survival, either my partner or I would need to intervene.

I was about to perform CPR on my future father-in-law.


Many dispatchers are trained to recognize signs of cardiac arrest from an oral description and then direct callers to begin CPR—even callers who might be in shock, as my partner and I were. But there is no universal requirement for dispatchers to do this. Few of the dispatch centers that have implemented T-CPR protocols deliver instructions consistently, and fewer still have strict quality-improvement measures in place. On the night of Todd’s cardiac arrest, I was fortunate that my hands were guided by the right dispatcher.

According to Robert Fazzino, a paramedic and the FDNY medical-affairs representative who procured our incident report, the hostess handed the cordless phone to my partner, Lex, who then handed it to me. Kneeling over Todd’s tensed body, I wedged the receiver between my right ear and shoulder. The dispatcher told me to interlock my hands—one atop the other, at the midpoint of the nipple line—and get ready to start pumping up and down, hard and fast.

The clock was ticking.

This wasn’t the first time I’d been involved in an emergency that required CPR. When I was a teenage pool lifeguard, a 74-year-old swimmer fell unconscious one summer afternoon. After I pulled her out of the water, five other guards and I performed CPR on her for several long minutes until paramedics arrived. She died days later.  

Now here I was again, face-to-face with someone clinging to life—only this time, it was a loved one, and my training was rusty.

In my lifeguarding days, I was regularly drilled on the CPR procedures for infants, children, and adults. Was it 15 compressions to two breaths for an adult? Or 30 to two? I was blanking. “What are the ratios?” I blurted out.

The dispatcher, realizing I was at least somewhat CPR conversant, seized the moment. No breaths necessary, he said. “Just stay on my count.”

That’s exactly what I did, according to the call audio. I counted aloud with the dispatcher, using my upper-body weight to press down on Todd’s sternum, before releasing: down and release, down and release. One and two and three and four and five and six …

Time slowed. I closed my eyes. Don’t stop, I thought. After what felt like an eternity, I heard sirens approaching.

“The public assumes that if they call 911 and someone’s in cardiac arrest that they’re going to get [CPR] instructions,” says Michael Kurz, an emergency-medicine physician at the University of Alabama at Birmingham and the volunteer chair of the American Heart Association’s T-CPR Task Force. “That’s not the case. It is the minority of cardiac arrests that receive that instruction.”

If I’ve learned anything in the weeks and months I’ve spent reconstructing the events of that evening, and researching the availability of T-CPR nationwide, it’s that we were very, very lucky. Dial 911 to report a cardiac arrest, and depending on where you are—a big city, a rural town, or somewhere in between—you may be told to wait until help arrives, to stand idle as your loved one’s fate hangs in the balance. Why didn’t that happen to us?


One day in August 1974, a panicked mother called the fire department in Phoenix, Arizona. A dispatcher listened as the woman explained that she and her husband had just pulled their 2-year-old son from the family swimming pool, and that the toddler was turning purple. “He’s not breathing!” she shouted.

“I want you to stay on the line,” the dispatcher said, noting the caller’s address. “I have a medic that is going to give you some help while I send someone.” The phone was passed to a department paramedic, Bill Toon, who had just clocked in.

As Toon wrote in Principles of Emergency Medical Dispatch years later, the department’s single paramedic unit was far enough away that the odds of it arriving in time to help were slim. “The dispatchers had little or no training in this area at this point in time,” he added.

Toon took it upon himself to assist the family until the paramedic unit arrived.

“I began to give the caller a crash course in CPR because the only real chance the child had of surviving was with his family doing the saving,” Toon wrote. After about a minute of over-the-phone instructions, Toon heard the toddler start to cry—a relief, because if he could cry, he could breathe. “That was a pretty sweet sound for everyone involved,” Toon recalled.

The three-minute, eight-second call was a signal moment in the emerging field of pre-arrival instruction and T-CPR. Toon’s ad hoc actions were remarkable because T-CPR protocols did not yet exist, making the episode’s recording an instant historical artifact. As Audrey Fraizer wrote in The Journal of Emergency Dispatch in 2019, word of the event made the national rounds and, as she later told me, helped in the push to standardize care in emergency dispatching.

By the early 1980s, the emergency medical system in King County, Washington, had become the first to implement a T-CPR script for dispatchers fielding cardiac-arrest calls. In the time since, T-CPR’s spread has been significant, albeit somewhat haphazard. A 2015 evidence review conducted by the American Heart Association suggested that, despite “rapid and widespread adoption,” dispatcher CPR instruction “does not lead to more successful resuscitations or improved survival.” But in Arizona, the birthplace of the practice, out-of-hospital cardiac-arrest victims who were provided with T-CPR were almost 65 percent more likely to survive than those who didn’t receive T-CPR, according to a February 2020 AHA T-CPR policy statement. Those who survived were also far less likely to have suffered brain damage.

Eight states—Indiana, Kentucky, Louisiana, Maryland, Tennessee, Virginia, West Virginia, and Wisconsin—currently require emergency dispatchers to provide T-CPR. But other states and jurisdictions—Arizona and New York among them—do not. The dispatchers in these states, says April Heinze of the National Emergency Number Association (NENA), a nonprofit that works to standardize 911 services, are going to send help, but until the ambulance arrives they may not be able to assist callers much.

To be sure, about one-third of the emergency dispatch centers in the U.S.—approximately 2,000—provide some sort of medical advice via telephone, helping bystanders assist someone who is choking, seizing, or even giving birth. Of those, “many do so without being required by law,” Heinze told me last spring. In her home state of Michigan, she added, more than 70 percent of dispatch centers provide these services, despite no mandate to do so. “Many others probably also do telephone CPR just because they know that’s the right thing to do,” Heinze, a former longtime 911 dispatcher, said.

Only recently has there been a proper drive, spearheaded largely by the AHA, to integrate T-CPR into the national 911 system, which itself dates back to only the late 1960s. “The push for telephone CPR just happened within the last year or two,” Heinze told me. “Legislation is very slow. It doesn’t happen overnight.” That at least eight states have T-CPR-specific legislation on the books, well, “I think that’s actually pretty good, to be honest with you.”


Still, nearly 50 years since Bill Toon’s impromptu T-CPR guidance, fewer than half of those who experience cardiac arrest outside of a hospital in America receive bystander CPR. Lay-rescuer rates are especially low in minority communities, due to both a lower overall availability of T-CPR and a widespread fear that involvement with a 911 call will lead to encounters with police or immigration authorities. The main obstacle to scaling up T-CPR, however, remains the patchwork nature of 911 itself. Though the national system is coordinated by the Federal Communications Commission, 32 states have adopted “home rule,” meaning that 911 and other services fall under local or regional control. As a result, implementing universal, consistent T-CPR programs is slowed by funding and staffing shortages and communication problems.

The AHA has argued that T-CPR is overwhelmingly cost-effective compared with other measures designed to reduce the time to first chest compression. Yet states and localities have limited budgets for new emergency-services initiatives. And 911 dispatchers, who are in short supply nationwide, were overburdened even before the start of the coronavirus pandemic. They are at the crux of a tightly choreographed feat of adrenaline, transportation, and communication; a high-stress job performed on marathon shifts, with varying degrees of training, and for low pay. Given the existing demands on dispatchers, who are disproportionately women, some are understandably wary of being held accountable for negative outcomes. “If anything goes wrong,” Heinze said, “the liability then falls more on the dispatcher than it does on the organization.”

Complicating matters is the fact that none of the nearly 6,000 emergency dispatch centers in the U.S. operates in exactly the same way. So when a 911 call from one area is routed to a dispatch center in the next town over—a not-uncommon occurrence—a caller may be transferred from a dispatcher trained in T-CPR to one who is not. And by the time responders arrive, it could be too late for someone like Todd in the throes of cardiac arrest.

Even when a bystander is lucky enough to connect with the right dispatcher, there are many points when things can go wrong. The reality is that not all bystanders can or will act.

Some simply aren’t physically capable of doing so. CPR requires two hands and has been compared to shoveling snow or walking through sand; one must push down 2.5 inches on the victim’s chest 100 to 120 times a minute in order to generate enough cardiac output. “It’s very tiring,” Fazzino, the FDNY liaison, explained.

Others might be concerned about infection risk, a worry inflamed by the coronavirus pandemic. Still others might be hesitant to perform CPR for fear of inadvertently causing physical harm, or of interfering with what they believe fate has decided for the victim.

To avoid these pitfalls in the course of T-CPR, dispatchers such as Adolfo Bonafoux don’t ask many questions once they’ve established that someone is calling on behalf of a person who is not breathing. “I will tell you what to do,” says Bonafoux, who fields emergency medical calls at the heavily fortified, Bronx-based PSAC II, one of New York City’s two public-safety answering centers.

By not asking questions or for a caller’s permission, “it takes the option away from you,” Bonafoux explained to me. “You’re more willing to act and follow my direction. Because if I give you the choice, you’re gonna stop and think. You’re gonna start to weigh all the variables. And that time is very valuable.”

Bonafoux is a former U.S. Army medic with 20 years of emergency-medical-service experience. He joined the FDNY in 2007, first as a paramedic and then, after being injured in the field, as a dispatcher. (He has formal training in, among other things, T-CPR protocols, a requirement instituted by the department’s medical directors.) Technically, he’s what’s known as an ARD, or assignment-receiving dispatcher. He isn’t the first person a caller talks to—that would be a police dispatcher, who discerns whether the caller needs to speak with the police, the fire department, or emergency medical services, and notes the caller’s location. Bonafoux receives the medical calls transferred from that police dispatcher, and handles the pre-arrival medical portion of the relay.

“My philosophy is if you’re not willing to do it, you’re going to stop me,” Bonafoux said. “Obviously I can’t force anybody over the phone to do anything. So I take an aggressive stance. A lot of times, people in a pressure situation, they’ll just do. They won’t hesitate, they won’t think about it, they’ll just do.”

Following Todd’s accident, I suspected that simply asking “What are the ratios?” had indicated to our dispatcher that I was familiar with CPR. And because of that baseline, it didn’t take much to get me to go.

Bonafoux later confirmed that hunch. He was the voice on the other end of the line, who walked us out of the depths of what Lex and I have taken to calling the Bad Night. “Muscle memory,” he said. “Once you have done it before, you remember it. Your brain starts remembering it. Your body remembers how to do it. That all contributed to the success of your father-in-law.”


The first responders, a paramedic team, arrived on the scene at 7:44:55 p.m., followed by the engine company, an FDNY lieutenant, and basic and advanced life-support units.

“From the time that the call comes in to the time that somebody is actually standing there, a professional provider, is [about] four minutes,” Fazzino told me over the phone, as he paged through our case file. The “real magic,” he said, is in that response time.

I remember a paramedic from the first unit crouching beside me, slinging a life-support bag off her shoulder and asking how long I’d been going at it, before relieving me. “Would you have guessed that was, you know, two and a half minutes of CPR you did?” Fazzino asked. “You get that serious fight-or-flight adrenaline rush. Your sense is enhanced. It becomes very surreal.” What felt like an eternity was really 150 seconds.

By 8 p.m., Todd had been shocked seven times with a portable defibrillator—typical, Fazzino said, for ventricular fibrillation, the kind of electrical disturbance of the heart that Todd experienced. Responders, now numbering at least a half-dozen, ran Todd’s electrocardiogram. That included the multiple defibrillation attempts, medications administered, and intubation.

Start to finish, the event clocked in at about 35 minutes, on par for this type of resuscitative effort in the field. Total call duration, including the T-CPR? Six minutes.

By 8:20 p.m., Todd was loaded into an ambulance. Lex and I got into a second ambulance, which followed closely as our caravan sped toward NYU Langone, the nearest hospital, about 10 minutes away. Port Authority officials temporarily halted Queens-Midtown tunnel traffic to allow us to slip through. I remember the lights streaking past our windows.

Todd was shocked an eighth time after being reeled into the emergency room. A long night was still ahead of us. But he now had a pulse—a testament to the help we’d gotten in those crucial first moments.  

Without T-CPR, “God forbid, what could have happened to your father-in-law?” asked Democratic Representative Norma Torres of California. “You wouldn’t have had somebody talking you through that.”

Torres, a former 911 operator in Los Angeles, is the lead sponsor of the 911 SAVES Act, a bipartisan bill that aims to reclassify 911 operators and other public-safety telecommunicators as “protective service occupations” under the Office of Management and Budget’s Standard Occupational Classification System. As it stands, dispatchers like Bonafoux are classified more as office secretaries. Torres wants to change that federal labor designation—with no disrespect to secretarial workers, she said—to encourage states to recognize dispatchers as crucial workers, recognition that in some states could exempt them from government furlough. Without dispatchers, Torres said, “we can’t get the help that we need.”   

A recently formed NENA-AHA working group, meanwhile, is focused on further standardizing T-CPR. (The AHA, for its part, has also launched Don’t Die of Doubt, a campaign to address the “alarming drop” since the start of the pandemic in 911 calls and ER visits by people needing urgent medical care after a stroke or heart attack.) But it would seem that scaling up T-CPR is as much about recognizing and supporting dispatchers as it is training lay people in CPR, something Lex and I have undertaken in the aftermath of our experience.

Here’s what I do know: The FDNY location from which the responding units were dispatched is mere blocks from the restaurant. NYU Langone happens to be one of the country’s top cardiac-care hospitals, too. Not only did I have the advantages of previous CPR training and Bonafoux’s experienced help, but we were in the right place at the right moment. Both luck and privilege—our well-appointed location, my previous training—were on my family’s side.

But the further I dig into the night of March 21, 2019, the clearer it becomes that I won’t ever be able to fully account for what transpired. I’d been on the fence about joining Lex and Todd for dinner, but made the last-minute decision to go. What if I hadn’t? It’s also entirely possible that, had Lex not immediately cried out for someone to call 911, kick-starting the “chain of survival,” this story would have a much different finale.

I asked Fazzino how many out-of-hospital cardiac arrests were reported in New York City in 2019, and of those patients, how many survived until either emergency services arrived or they reached a hospital. He couldn’t say for sure, but noted that the “vast majority” of patients behind such emergency requests that come into the city’s two call centers ultimately do not make it.

Of the minority of people who do survive, how many of them get to go home? That number, Fazzino said, is even smaller. Todd did what many people have not, “which is to cross the line and then come back to tell the story about it.”

The first night at the hospital, Todd was put into therapeutic hypothermia—“on ice,” the doctors called it—in an attempt to redirect blood from the rest of his body to his brain. We were told he would stay in this medically induced coma for up to 72 hours. The next morning, less than 18 hours after his heart gave out, he woke up on his own. I can still see the look of surprise and excitement on the attending nurse’s face. “Who did the bystander CPR?” one of his doctors asked. Lex pointed at me. “Well done.”

The following evening, in a quiet moment in a hospital lounge, Lex and I decided to get married. Todd was able to come to the wedding, three months later.

“By the way,” he said, shortly before being discharged. “Thanks.”