They had no idea what was wrong.
The patient, a 41-year-old female, lay on the kitchen floor unresponsive. Her husband said that she’d collapsed suddenly while complaining of a headache. Immediately, the two paramedics began their assessment. At least she was breathing and had a pulse.
“What the fuck would have happened to Autumn? She’s younger than me. What could have happened?”
Paramedic Steve Mason got a quick summary from the husband Robert Ferrante. He said his wife Autumn Klein was a neurologist at the University of Pittsburgh-Medical Center. She had returned home at about 11:30 p.m. after a grueling day at work and tumbled to the floor just a few minutes later. Mason asked him about a big Ziploc bag of white powder on the kitchen counter. It was Creatine, Ferrante said; his wife was taking it for infertility. As they spoke, Mason’s partner Jerad Albaugh got his attention.
Their patient was crashing. Her blood pressure and pulse were dropping fast.
“She’s not responding,” Albaugh said.
The men loaded the unconscious woman onto a gurney and rushed her to their ambulance, parked in front of the home. They called ahead to the hospital and raced the half mile there, with Albaugh unable to do anything other than start an IV during the short trip.
They pulled up to the emergency entrance of University of Pittsburgh Medical Center-Presbyterian hospital — called Presby by local residents — at 12:21 a.m. Just over an hour after she had left work there, Klein lay prone on the gurney, her arms contorted and her face twisted up and over her left shoulder.
Emergency department resident Dr. Andrew Farkas met the stretcher in the hallway. Klein’s eyes were open and glassy, and her breaths were shallow. There was a vacant look on her face.
They rushed her to curtain area 32 in the emergency department. Her heart rate was measured in the low 40s, and her blood pressure was a mere 48 over 36. Although her pupils were reactive to light, the patient — whom they now knew was one of their own — was unresponsive.
The team of nurses and technicians assisting Farkas put in another IV to push fluids and boost Klein’s blood pressure. Her respirations were starting to slow — as low as four per minute — and Farkas knew she needed to be put on a ventilator immediately to help her breathe. He inserted the breathing tube into her mouth and down her trachea, then ran to get his attending physician, Dr. Thomas Martin.
Just two minutes after Farkas checked his patient’s pupils, Martin checked them again. They were no longer reacting to light.
Farkas had ordered a broad panel of blood tests, gases, and chemistries be sent out to check Klein’s organ function. The problem, though, was that because she was so slim, the staff was having trouble getting a blood draw from her arm. Using a larger needle, they moved to the femoral artery in her leg, but still couldn’t successfully take her blood.
As the staff continued to work on Klein, Ferrante arrived with his friend and colleague Dr. Robert Friedlander, who had driven him there. Like his wife, Ferrante, 64, was also connected to Presby; he was co-director of the Center for Amyotrophic Lateral Sclerosis (ALS) Research at the University of Pittsburgh’s Medical Center.
Farkas pulled back the curtain surrounding the bed where Klein lay, the vent pushing air in and out of her lungs. Ferrante took a long look and then screamed, “No!”
Farkas continued his assessment while he listened to Ferrante describe what had happened at the house. Klein had been complaining of headaches in recent weeks, he said, and when she arrived home from work that night she’d complained of not feeling well. He said she grasped her head in her hands and then dropped to the floor.
Farkas and Martin now suspected she was having a brain hemorrhage. They needed to get a CT scan of her head immediately, so they momentarily scrapped their plans for the blood draw to move her to the 3-D imaging machine. It was only one hundred feet from her bed, but Klein’s blood pressure was below 60, and her pulse was in the 30s. Just putting her in the scanner was too risky. Ignoring protocol, Martin went in with her. Draped in a protective vest, he pushed epinephrine every one to two minutes to keep her heart pumping.
As the scan ran, the images immediately appeared on a computer monitor in the control room.
They were normal.
“It looks completely clear,” Farkas said. “There’s no explanation for her symptoms. There’s no evidence of any disease state whatsoever.”
The emergency team then switched focus, trying to think of what else could cause such a dramatic decline so fast. They ordered additional CT scans of Klein’s chest, abdomen, and pelvis. Her EKG showed no abnormalities in her heart’s electrical activity. There was no aneurysm in her abdomen. She had no aortic tear that could have caused blood to spill into her chest cavity. There was no blood clot in her lungs.
The treatment team had no idea what was wrong.
At 1:20 a.m., Martin paged the hospital’s on-call intensivist — a doctor who specializes in treating critically-ill patients. Dr. Lori Shutter returned the page.
“This is Tom Martin. I’m one of the ED docs.”
“Yeah,” Shutter answered. “What do you need?”
“I have a patient down here. You may have heard of her: Autumn Klein. She’s one of our neurology attendings.”
“Autumn?” Shutter responded. Klein, who was chief of the division of women’s neurology at the hospital, was her colleague, neighbor, and friend.
“Yeah. She got here around midnight. I’ve been working on her since then, and I just need help. I don’t know what’s going on.”
Shutter hung up the phone, thinking to herself, “What the fuck would have happened to Autumn? She’s younger than me. What could have happened?” She rushed to the elevators to make the trip down the 9 floors to the ground level of the hospital and hurried around the corner to the trauma room where Klein lay, a huge team of nurses, technicians, and doctors working on her. Ferrante, Friedlander, and Klein’s neurology chair Dr. Lawrence Wechsler, were there as well.
“Bob, what’s going on?” Shutter asked Ferrante. He said again that his wife had collapsed at home.
Martin was at the bedside, still pushing syringes full of epinephrine to try to sustain blood pressure. He told Shutter about Klein’s condition, the need to intubate her, and the puzzling fact that all the CT scans were clear.
“Bob, what’s going on?” Shutter asked Ferrante. He said again that his wife had collapsed at home.
“I can’t figure out what’s happening.”
The nurses were still having difficulty getting blood from Klein, who at nearly 5 feet 7 inches, weighed only 107 pounds. Everyone agreed she needed to have a central line so that they could more quickly administer medications and draw blood.
Using an ultrasound to locate the patient’s internal jugular vein — the largest in the neck — Farkas inserted a triple-lumen catheter. As he placed the tube inside the vein, the blood that came out was bright red — noticeably so, since venous blood, which has less oxygen in it, should be much darker.
The resident alerted Martin to what he’d seen, wondering if he’d accidentally placed the line in an artery, where the blood would be a much brighter red. But the more experienced physician had seen the ultrasound and knew the catheter was in the right location. Still, Farkas did a manometer test, placing a plastic tube over the needle to see what would happen to the blood inside. If it pulsed up, then the line was in an artery. If it dropped down, then it was in a vein. When Farkas held the line up, Klein’s blood dropped back down. They had definitely tapped a vein; Farkas deftly stitched the catheter in place.
The treatment team, which now included a neurologist, a cardiologist, and an intensive-care-unit physician, had received Klein’s initial lab results, looking at her electrolytes, white blood cell count, hemoglobin, red blood cell volume, and coagulation factors. Again, they were normal. The only things that came back abnormal were the patient’s pH and oxygen levels. There was a lot of acid in her blood, indicating a severe metabolic dysfunction, and her oxygen levels were more than double what they should have been. Klein’s cells were unable to use the oxygen in her blood, but the doctors didn’t know why.
They continued to give her sodium bicarbonate to lower the acid levels and increased the ventilator speed to make her breathe faster. They were also still pushing huge doses of epinephrine regularly.
“We’ve been doing this a long time,” Martin said. “We need to come up with some ideas.”
Shutter suggested running a toxicology screen that would check for a standard set of drugs and poisons. As the team was treating Klein, they heard her husband talking about her.
“She really enjoyed what she was doing,” Ferrante said. “She loved her job, and she would never want to survive if she couldn’t be herself and continue her work.”
At 2:17 a.m., as the physicians continued to try to stabilize her, Klein went into cardiac arrest. They called a code. A respiratory therapist disconnected the ventilator and began bagging her manually as a team of technicians and nurses took turns performing chest compressions. Each person got up on top of Klein, straddling her. They clasped their hands one over the other with arms fully extended, pounding on her chest to try to compress it one to two inches with each pump.
Martin had his hand on Klein’s femoral artery in her groin to ensure her blood was still circulating with each compression.
“Shit. Bloody secretions,” Martin exclaimed as he saw bloody fluid coming up from the breathing tube. They suspected her ribs had been broken from the compressions and maybe punctured her lungs.
For 22 minutes, they took turns performing CPR, continuing to push epinephrine to try to sustain her heart and blood pressure.
“This has been going a long time,” Martin said. “This isn’t going well. Do you think we should call this?”
“We probably should, but I’m not going to be the one to go out and tell the husband,” Shutter responded. She told Martin she couldn’t tell Ferrante they had let his wife die. “Do you mind going out to talk to him?”
Shutter took over running the code as Martin went to speak to Bob. He sobbed at the news.
She replaced Martin in feeling for Klein’s femoral pulse and was struck for the first time by how small the woman was, lying on the bed wearing nothing but a pair of pink, animal-print panties.
“No one should ever be in this position with a friend of theirs,” Shutter thought, a wave of despair flashing through her as she looked at her hand in her friend’s groin. Just as quickly, though, she told herself to stop staring.
“The next round is when we’ll stop if she’s not back,” she told the team.
They gave Klein another round of epinephrine and started a last set of compressions. About 30 seconds in, Martin returned to the room.
“Her husband understands we’re going to call it,” he said.
A short time later, a nurse keeping track of the time announced, “It’s been two minutes.”
Everyone stopped what they were doing.
“Oh, my gosh, there’s a pulse,” Shutter said. “Someone confirm.” A touch to Klein’s carotid artery confirmed.
Her heart was beating — not well, but like a quivering bag.
Using electric paddles, they shocked the heart with 120 joules — which Shutter compared to being hit in the chest with a baseball thrown at 50 miles per hour.
The shock steadied the heart rate, though it remained very slow. She was not conscious or responsive. But she was alive.
“Now what do we do?” Shutter asked. “What’s the next step?”
The treatment team knew they needed to take measures to ensure Klein’s heart continued beating, and so they paged the on-call cardiothoracic surgeon. As the physicians discussed their ideas, at about 3 a.m., Friedlander suggested he and Ferrante go get a cup of coffee.
When the cardiothoracic surgeon and his fellow arrived, they discussed a few different options, including placing Klein on extracorporeal membrane oxygenation (ECMO which removes blood from the body, filters and puts oxygen in it, and then returns it so the organs can use the oxygenated blood to continue to function. They also considered installing an intra-aortic balloon pump.
The physicians worried that, because of their patient’s small frame, they would not be able to place the ECMO lines. The tubing for the procedure is the size of a small garden hose and must be pushed several inches into the patient’s veins and arteries.
Dr. Jay Bhama decided they should do the balloon pump, but as they made their plans — and discussed whether Klein could sustain a trip to the cardiac catheterization lab — her vitals again began to fall.
“While you guys are standing there, we’re losing her blood pressure,” said Dr. Jeremiah Hayanga, who was working with Bhama that night.
They immediately decided they had to go with ECMO. The physicians knew it was possible that Klein could lose her limbs from the procedure because they would not get proper blood flow with the large lines inside of her.
But without it, they knew, she would die.
There was no time to get to an operating room, so the doctors called the emergency ECMO team on the hospital’s second floor, telling the perfusionist to bring all of the equipment directly to the emergency department. Within minutes, the team wheeled the 18-inch-wide, 36-inch-tall machine to Klein’s bedside.
Hayanga and Bhama worked together. They prepped both the right and left femoral regions. Using a scalpel, they nicked the skin on each side, and then used a dilator catheter to widen the holes to accommodate the width of the tubes. Once they were big enough, they fed an 11-inch-long plastic tube up into the femoral vein on the left and into the femoral artery on the right. When they turned on the ECMO machine, it would work by drawing the blood out of the femoral vein, running it through the filtration and oxygenation system, and then return it to Klein’s body through her right femoral artery.
They then stitched the tubes into place.
Once Klein was on ECMO, the staff was able to maintain her heart rate at 60 to 70 beats per minute. They gave her a transfusion of two units of packed red blood cells and planned to cool her body because hypothermia can help recovery after cardiac arrest. However, doctors noted that she was already four degrees cooler than the 33 degrees Celsius that is recommended.
When Ferrante and Friedlander returned to the emergency room 30 minutes after they’d left, they learned Klein was on ECMO.
Ferrante appeared to relax at the news, and Martin worried that he’d given the man false hope. Although her heart was beating, Klein’s condition remained grave.
Throughout the night, the physicians talked about what they thought might have caused their colleague’s collapse. They suggested a number of ideas, including Brugada syndrome, which is an abnormal heart rhythm, and even the possibility of an electrical brain storm, which is when brain cells fail to properly discharge energy, causing a surge of it through the brain.
They consulted local, national, and international experts to try to get the best information possible to guide her care.
At 5:30 a.m., it was agreed that Klein could be moved from the emergency department to the cardiothoracic intensive-care unit (CTICU) on the second floor of the hospital. On arrival, the staff removed Klein’s white-gold wedding band to return to her husband and hooked her up to a continuous EEG machine with 32 small electrodes placed around the head to measure her brain activity. There was none.
Dr. Frank Guyette, a consultant with the hospital’s post–cardiac arrest team who evaluated Klein at 8:21 a.m., noted that her blood pressure was still just 70 over 30. Her pupils were not responding to light. She had no gag reflex. Her extremities showed no spontaneous movement, and there was no response to painful stimuli. He put her chances of a meaningful recovery at less than 4 percent.
Later that morning, Dr. Jon Rittenberger, another member of the post–cardiac arrest team at Presby, took over Klein’s care. He noted that she was “profoundly comatose” with a “flat, nonresponsive EEG,” but said that from a “neuro-prognostic standpoint, it is simply too early to tell.”
The doctors treating Klein agreed that it would not be fair to fully assess her brain activity until she had been properly rewarmed since the cooling protects brain function. Shutter spoke with a colleague at Yale, who suggested they give their patient at least 72 hours at average body temperature before reaching any conclusions. She passed that information along to Ferrante.
“Don’t call things too early,” she told him. “You never know.”
Rittenberger, still stumped by what had precipitated Klein’s collapse, called his colleague, Dr. Clifton Callaway, an emergency physician, who was traveling. He filled him in on the case and recounted the patient’s symptoms — including the bright-red blood in the central line— and noted that she still had very high levels of acid, even with improved blood flow and increased levels of oxygen.
Callaway suggested running a test for cyanide toxicity, even though the chances of that causing the collapse were slim.
“Please also note that due to her profound acidosis on initial arrival, we had added on a toxicologic screen along with serum alcohols and cyanide (although this is unlikely),” Rittenberger wrote in Klein’s chart. The blood was drawn at 2:32 p.m. on April 18, and the tube sent to Quest Diagnostics in Chantilly, Virginia, to be tested.
About two hours later, Ferrante met with a social worker at the hospital. He told the woman that his wife’s parents had arrived in Pittsburgh from their home in Baltimore, but that he had not talked to them and was uncertain if they planned to visit their daughter.
Lois and Bill Klein had driven through the night from their home in Towson, Maryland, after receiving a call from Ferrante. He told them he did not want to take them to see their daughter until his adult children (from a previous marriage), Michael and Kimberly, arrived from Boston and San Diego. Until then, the Kleins remained at the couple’s house with Ferrante and he and his wife’s six-year-old daughter, Cianna.
That evening, though, the adults gathered together and finally went to the hospital.
When the group entered Klein’s room in the cardiac intensive care unit, her grave condition was obvious.
Over the next two days, Klein’s family members spent their time shuttling back and forth from the couple’s home to the hospital. Lois and Bill sat for hours at a time at the bedside. But they were increasingly unhappy with the situation. They felt like they were being left out of the discussions about their daughter’s care and condition. When Bob Ferrante met with his wife’s doctors, he would include Kimberly, who was a physician, but not the Kleins. Finally, when Ferrante met with the doctors in the hallway, his mother-in-law started forcing her way into the conversations, determined not to be left out.
She was then frustrated by her son-in-law’s behavior during the discussions with the doctors. Instead of letting them talk about their suspicions, Bob kept making his own diagnoses.
“Don’t you think this is what it is?” he would ask.
The team of physicians continued to talk about what might have caused such a drastic, untreatable condition.
The team of physicians continued to talk about what might have caused such a drastic, untreatable condition. Klein had been in touch with her family physician six weeks earlier, noting that she had been having significant hair loss and that she thought she might need to have her hormone levels checked, but otherwise, there were no medical complaints reported.
At the hospital, Ferrante told several people he believed his wife had suffered an electrical brain surge.
“Autumn’s two-year history of headache-migraine may have caused this,” he wrote in an e-mail to Friedlander on April 20. “While we were certain of four events that I witnessed, there may have been more that she did not discuss or were subclinical. There remains the hypothesis that this was myogenic and that there was a defect in the heart.”
As it became clear that Klein was not going to recover, the family started talking about her wish to be an organ donor. Intensive-care physicians treating her believed that the liver and kidneys could be transplanted.
Ferrante e-mailed Friedlander: “This may be the last physical gift that she can provide and would be a wonderful one.”
There was also talk about an autopsy.
Rittenberger suggested to Ferrante that if his wife had died from a heart-rhythm disturbance, then an autopsy ought to be performed. It was likely that such a condition would be genetic, which meant their young daughter could be at risk.
Lois Klein also wanted an autopsy. She wanted to know what had happened. “A healthy 41-year-old woman doesn’t just come home, drop on the floor, and die,” she said.
But Ferrante said no. He was so insistently against the procedure that several physicians noted his rejections in Klein’s chart. He didn’t even want a limited one — an external examination and toxicology screening — that would have allowed for genetic testing. He talked to Shutter about it, suggesting that the ECMO might wash anything out of his wife’s system that would point to cause of death.
“What would an autopsy tell us?” he asked.
He reiterated the point in the e-mail to Friedlander: “Discussion with all the principals suggests that autopsy, limited or full, will not resolve this issue.”
However, in Pennsylvania, with any unnatural death, the law requires an autopsy. Whether Ferrante wanted it or not, the Allegheny County Medical Examiner’s Office was going to do one.
As Klein entered her third day in the hospital, her status remained unchanged.
“Autumn’s EEG remains without any activity and has been so twenty-four hours after rewarming. The ICU team is ready to call her death,” Ferrante wrote to Friedlander. “The fact that the EEG has not change[d] suggests little chance of recovery. Bedside brainstem exam is negative. It would be wonderful to get another CT but they are unable to take her off ECMO to do this. Unclear whether her heart would sustain the workload.”
Ferrante took Cianna to the hospital to see her mother Saturday morning. For the two previous days he had sent his daughter to school to try to sustain normalcy, telling her that her mother had gone away to a meeting but would return soon. By Saturday, though, he explained to his little girl that her mom had a medical problem, was very ill, and was in the hospital.
“Cianna was there all morning,” Bob wrote in his e-mail to Friedlander. “That was the best thing to do. She was pleased to see her mom, but when she left she commented that she did not think that mommy would be coming back home. This is all so very sad to me. It breaks my heart.”
Ferrante agreed that day to remove Klein from life support.
At 11:06 a.m. on April 20, Rittenberger performed the first of two required brain-death exams on Klein with her parents and Ferrante at her side. He called to her, announcing his presence. She did not respond. He pinched her. There was nothing. He checked her pupils, shining a light directly into her eyes. No response.
No gag reflex.
“Throughout the entirety of my observations to my examination, I observed no respiratory efforts whatsoever.”
A doll’s-eye reflex — where the patient’s eyes spontaneously move to the side when the head is moved in the opposite direction — was absent.
And Klein showed no reaction when he dripped cold water into her ear canal.
At 12:10 p.m., he declared her brain dead.
Fifteen minutes later, Dr. Joseph Darby, a critical-care physician, conducted the same tests with the patient’s parents at her bedside.
He also took Klein off the ventilator for a full 5 minutes. “Throughout the entirety of my observations to my examination, I observed no respiratory efforts whatsoever,” he wrote.
Klein was pronounced dead at 12:31 p.m. on April 20, but remained on the equipment to keep her heart beating and her lungs functioning in order to facilitate organ donation.
That afternoon, staff from the Center for Organ Recovery and Education met with Ferrante to obtain permission for organ donation. The family remained with Klein throughout the evening.
Then, at 2:26 a.m. on April 21, staff arrived in the room to prep her for organ harvesting, and she was transported to the operating room.
The surgery began at 3 a.m. and ended two hours later.
Her body arrived at the medical examiner’s office six hours after that.
Paula Reed Ward has been a reporter with the Pittsburgh Post-Gazette since 2003. Scenes and dialog in the foregoing excerpt have been recreated based on extensive examination of medical records, emails, court documents, and interviews with the parties involved.