When Tess Camp was pregnant with her second child, she knew she would need to get to the hospital fast when the baby came. Her first labor had been short for a first-time mother (seven hours), and second babies tend to be in more of a hurry. Even so, she was not prepared for what happened: One day, at 40 weeks, she started feeling what she thought was just pregnancy back pain. Then her water broke, and 12 minutes later, she was holding a baby in her arms.
Needless to say, she didn’t make it into the hospital in time. But the first contraction after Camp’s water broke at home had been so intense—“immediate horrific pain; I could barely talk”—that she and her husband rushed into the car. He drove through town like a madman, running red lights. They were turning into the ER when she saw the baby’s head between her legs. Her husband tore out of the car, yelling for help. A security guard ran over to a terrified Camp in the passenger’s seat, and in that moment, her son slipped out and into the security guard’s hands. His umbilical cord was wrapped around his neck. An ER nurse finally appeared to take the baby—still blue and limp—and resuscitated him right on the curb.
What Camp experienced is called “precipitous labor,” when a baby is born after fewer than three hours of regular contractions. It is uncommon but not entirely rare, occurring in about 3 percent of deliveries, usually in second, third, or later labors. Having had a previous fast birth, like Camp did, increases the chances of a precipitous labor. But otherwise, doctors can’t predict for sure who will have one, especially among first-time moms with no previous birth experience. Like many topics in pregnancy and childbirth, precipitous labor remains understudied.
Counterintuitively, perhaps, an extremely fast labor is not always a better one. It can even be a terrible one. “It felt like being hit by a truck and dragged along behind,” says Stephanie Spitzer-Hanks, a doula and childbirth-class instructor who had precipitous labors with her two children. “People would tell me I was lucky, and I don’t feel like that. I tell my students, ‘I don’t really wish for you to have this kind of labor.’” In normal labor, each contraction gradually opens the cervix and prods the baby out. In a precipitous labor, the cervix still has to open just as wide, and the baby still has to move just as far—but in much less time. It’s like running the length of a marathon at the punishing pace of a sprint.
Babies born through precipitous labor tend to do just fine, but the process can be traumatic for the mother’s body. In the normal course of labor, says Tamika Auguste, an ob-gyn at MedStar Washington Hospital Center, the back-and-forth movement of the baby’s head during contractions stretches the perineum, a layer of tissue especially likely to tear in childbirth. In one study, precipitous labor multiplied the odds of a severe third-degree perineal tear by 25 and the odds of postpartum hemorrhaging by almost 35. (Precipitous labor is also responsible for one of the most horrifying case reports I have ever come across, whose title contains the phrase “severed external anal sphincter.”)
Even for ER doctors, “a precipitous delivery is right up there with some of the most stressful events that we managed,” says Joelle Borhart, an emergency-medicine doctor also at MedStar Washington Hospital Center. Precipitous labor can happen so fast that even if the mother makes it to the hospital, there is sometimes no time to transfer her from the ER to the labor-and-delivery unit. ER staff are trained in childbirth, but it’s not what they do on a daily basis. Borhart says the emergency department at her large hospital in Washington, D.C., gets about one case a month. Brian Sharp, an emergency-medicine physician at UW Health—a large academic hospital in Madison, Wisconsin—told me his hospital averages a little over once a year; the smaller community site where he also works just had their first case of precipitous labor in years. The rarity of these events means that hospitals aren’t always the most prepared. When Camp arrived with her baby almost born at the entrance of the ER, the hospital sent out the wrong code, mistakenly suggesting that there had been an abduction. No one from labor and delivery came to meet her, because they were counting babies to make sure none had gone missing. The hospital later reviewed her case, Camp told me, to figure how to improve the response in future situations.
All of this means that precipitous labor can be psychologically distressing too. When Bryn Huntpalmer, who runs the podcast The Birth Hour and a childbirth course, talks with postpartum mothers, “more times than not, the person who shares their precipitous labor has that shell-shocked view of it.” Some of the mothers I interviewed talked about feeling out of control and deeply disconnected from their bodies. “I couldn’t get words out. I couldn’t open my eyes. I couldn’t control what my arms were doing,” says Shannon Burke, who had a precipitous labor with her second child. “I couldn’t do anything.” For many people, the experience of childbirth is an experience of ceding control, of letting our most animal instincts take over. But in normal labor, this is at least a gradual process; you can joke and laugh and walk in the early phases, and only hours in, when you’ve mentally prepared yourself, do the screaming and vomiting take over. Burke remembers her 24-hour first labor fondly, in fact; she had spent the early phase at home with her mother and sister, readying the house for the baby. With her precipitous labor, she had no time for any of that. She plunged straight into full-blown pain.
“There’s no buildup to prepare your mind and body,” Huntpalmer, the podcaster who herself went through precipitous labor, told me. “Everything was so compressed.” But in talking about her experience—and talking since on The Birth Hour with hundreds of women about their experiences—she ultimately came to see her precipitous labor as affirming, too: Her body knew what to do. “It was so hands-off from my midwife. I was able to just kind of do it all myself,” she says. Emily Geller, who delivered her second baby during a precipitous labor in a car, told me the same. She had what she felt was an unnecessary C-section with her first child, so she wanted a natural vaginal birth this time—and she did have one, just faster than she planned. It was empowering, she said, to know that she could do it after all.
When Camp got pregnant with her third child, though, she did not want to give birth in the car again. Her husband was terrified too—he kept saying he was going to rent a trailer so they could spend the final weeks of her pregnancy sleeping in the hospital parking lot. “It’s $150 a week to rent a trailer,” she remembers him telling her. They didn’t do that, but she did schedule an induction at 39 weeks. Her daughter was born after two pushes.
Sarah Zhang
Source link
This article was originally published by Kaiser Health News.
Amanda Shelley was sitting in her dentist’s waiting room when she received a call from the police. A local teenage girl had been sexually assaulted and needed an exam.
Shelley, a nurse in rural Eagle County, Colorado, went to her car and called a telehealth company to arrange an appointment with a sexual-assault nurse examiner, or SANE. The nurse examiners have extensive training in how to care for assault survivors and collect evidence for possible criminal prosecution.
About an hour later, Shelley met the patient at the Colorado Mountain Medical urgent-care clinic in the small town of Avon. She used a tablet to connect by video with a SANE about 2,000 miles away, in New Hampshire.
The remote nurse used the video technology to speak with the patient and guide Shelley through each step of a two-hour exam. One of those steps was a colposcopy, in which Shelley used a magnifying device to closely examine the vagina and cervix. The remote nurse saw, in real time, what Shelley could see, with the help of a video camera attached to the machine.
The service, known as “teleSANE,” is new at Shelley’s hospital. Before, sexual-assault patients faced mountains of obstacles—literally—when they had to travel to a hospital in another county for care.
“We’re asking them to drive maybe over snowy passes and then [be there] three to four hours for this exam and then drive back home—it’s disheartening for them,” Shelley says. “They want to start the healing process and go home and shower.”
To avoid this scenario, teleSANE services are expanding across the country in rural, sparsely populated areas. Research shows that SANE programs encourage psychological healing, provide comprehensive health care, allow for professional evidence collection, and improve the chance of a successful prosecution.
Jennifer Pierce-Weeks is the CEO of the International Association of Forensic Nurses, which created the national standards and certification programs for sexual-assault nurse examiners. She says every sexual-assault survivor faces health consequences. Assaults can cause physical injuries, sexually transmitted infections, unwanted pregnancies, and mental-health conditions that can lead to suicide attempts and drug and alcohol misuse.
“If they are cared for on the front end, all of the risks of those things can be reduced dramatically with the right intervention,” Pierce-Weeks says.
Pierce-Weeks says there are no comprehensive national data on the number and location of health-care professionals with SANE training. But she says studies show that there’s a nationwide shortage, especially in rural areas.
Some rural hospitals struggle to create or maintain in-person SANE programs because of staffing and funding shortfalls, Pierce-Weeks says.
Training costs money and takes time. If rural hospitals train nurses, they still might not have enough to provide round-the-clock coverage. And nurses in rural areas can’t practice their skills as often as those who work in busy urban hospitals.
Some hospitals without SANE programs refer sexual-assault survivors elsewhere because they don’t feel qualified to help and aren’t always legally required to provide comprehensive treatment and evidence collection.
Avel eCare, based in Sioux Falls, South Dakota, has been providing telehealth services since 1993. It recently added teleSANE to its offerings.
Avel provides this service to 43 mostly rural and small-town hospitals across five states and is expanding to Indian Health Service hospitals in the Great Plains. Native Americans face high rates of sexual assault and might have to travel hours for care if they live in one of the region’s large, rural reservations.
Jen Canton, who oversees Avel’s teleSANE program, says arriving at a local hospital and being referred elsewhere can be devastating for sexual-assault survivors. “You just went through what is potentially the worst moment of your life, and then you have to travel two, three hours away to another facility,” Canton says. “It takes a lot of courage to even come into the first hospital and say what happened to you and ask for help.”
Patients who receive care at hospitals without SANE programs might not receive trauma-informed care, which focuses on identifying sources of trauma, determining how those experiences may affect people’s health, and preventing the retraumatizing of patients. Emergency-department staffers may not have experience with internal exams or evidence collection. They also might not know about patients’ options for involving police.
Patients who travel to a second hospital might struggle to arrange for and afford transportation or child care. Other patients don’t have the emotional bandwidth to make the trip and retell their story.
That’s why some survivors, such as Ada Sapp, don’t get an exam.
Sapp, a health-care executive at Colorado Mountain Medical, was assaulted before the hospital system began its SANE program. She was shocked to learn that she would need to drive 45 minutes to another county for an exam. “I didn’t feel comfortable doing that by myself,” Sapp says. “So my husband would have had to come with me, or a friend. The logistics made it feel insurmountable.”
Sapp’s experience inspired her to help bring SANE services to Colorado Mountain Medical.
Shelley and several other of the hospital system’s nurses have SANE training but appreciate having telehealth support from the remote nurses with more experience. “We are a rural community, and we’re not doing these every single day,” Shelley says. “A lot of my nurses would get really anxious before an exam because maybe they haven’t done one in a couple months.”
A remote “second set of eyes” increases the confidence of the in-person nurse and is reassuring to patients, she says.
Avera St. Mary’s Hospital in Pierre, South Dakota, recently began using teleSANE. Rural towns, farms, and ranches surround this capital city, home to about 14,000 people. The nearest metropolitan area is more than a two-and-a-half-hour drive.
Taking a break from a recent busy morning in the emergency department, the nurse Lindee Miller rolled out the mobile teleSANE cart and colposcope device from Avel eCare. She pulled out a thick binder of instructions and forms and opened drawers filled with swabs, evidence tags, measuring devices, and other forensic materials.
“You’re never doing the same exam twice,” Miller said. “It’s all driven by what the patient wants to do.”
She said some patients might want only medicines to prevent pregnancy and sexually transmitted infections. Other patients opt for a head-to-toe physical exam. And some might want her to collect forensic evidence.
Federal laws provide funding to pay for these sexual-assault exams, but some survivors are billed because of legal gaps and a lack of awareness of the rules. A proposed federal law, the No Surprises for Survivors Act, would close some of those gaps.
SANE programs, including telehealth versions aimed at rural communities, are expected to continue expanding across the country.
President Joe Biden signed a bill last year that provides $30 million annually through 2027 to expand SANE services, especially those that use telehealth and help rural, tribal, and other underserved communities. The law also requires the Justice Department to create a website listing the locations of the programs and grant opportunities for starting them.
Arielle Zionts
Source link
At the height of the coronavirus pandemic, as lines of ambulances roared down the streets and freezer vans packed into parking lots, the pediatric emergency department at Our Lady of the Lake Children’s Hospital, in Baton Rouge, Louisiana, was quiet.
It was an eerie juxtaposition, says Chris Woodward, a pediatric-emergency-medicine specialist at the hospital, given what was happening just a few doors down. While adult emergency departments were being inundated, his team was so low on work that he worried positions might be cut. A small proportion of kids were getting very sick with COVID-19—some still are—but most weren’t. And due to school closures and scrupulous hygiene, they weren’t really catching other infections—flu, RSV, and the like—that might have sent them to the hospital in pre-pandemic years. Woodward and his colleagues couldn’t help but wonder if the brunt of the crisis had skipped them by. “It was, like, the least patients I saw in my career,” he told me.
That is no longer the case.
Across the country, children have for weeks been slammed with a massive, early wave of viral infections—driven largely by RSV, but also flu, rhinovirus, enterovirus, and SARS-CoV-2. Many emergency departments and intensive-care units are now at or past capacity, and resorting to extreme measures. At Johns Hopkins Children’s Center, in Maryland, staff has pitched a tent outside the emergency department to accommodate overflow; Connecticut Children’s Hospital mulled calling in the National Guard. It’s already the largest surge of infectious illnesses that some pediatricians have seen in their decades-long careers, and many worry that the worst is yet to come. “It is a crisis,” Sapna Kudchadkar, a pediatric-intensive-care specialist and anesthesiologist at Johns Hopkins, told me. “It’s bananas; it’s been full to the gills since September,” says Melissa J. Sacco, a pediatric-intensive-care specialist at UVA Health. “Every night I turn away a patient, or tell the emergency department they have to have a PICU-level kid there for the foreseeable future.”
I asked Chris Carroll, a pediatric-intensive-care specialist at Connecticut Children’s, how bad things were on a scale of 1 to 10. “Can I use a Spinal Tap reference?” he asked me back. “This is our 2020. This is as bad as it gets.”
The autumn crush, experts told me, is fueled by dual factors: the disappearance of COVID mitigations and low population immunity. For much of the pandemic, some combination of masking, distancing, remote learning, and other tactics tamped down on the transmission of nearly all the respiratory viruses that normally come knocking during the colder months. This fall, though, as kids have flocked back into day cares and classrooms with almost no precautions in place, those microbes have made a catastrophic comeback. Rhinovirus and enterovirus were two of the first to overrun hospitals late this summer; now they’re being joined by RSV, all while SARS-CoV-2 remains in play. Also on the horizon is flu, which has begun to pick up in the South and the mid-Atlantic, triggering school closures or switches to remote learning. During the summer of 2021, when Delta swept across the nation, “we thought that was busy,” Woodward said. “We were wrong.”
Children, on the whole, are more susceptible to these microbes than they have been in years. Infants already have a rough time with viruses like RSV: The virus infiltrates the airways, causing them to swell and flood with mucus that their tiny lungs may struggle to expel. “It’s almost like breathing through a straw,” says Marietta Vazquez, a pediatric-infectious-disease specialist at Yale. The more narrow and clogged the tubes get, “the less room you have to move air in and out.” Immunity accumulated from prior exposures can blunt that severity. But with the pandemic’s great viral vanishing, kids missed out on early encounters that would have trained up their bodies’ defensive cavalry. Hospitals are now caring for their usual RSV cohort—infants—as well as toddlers, many of whom are sicker than expected. Infections that might, in other years, have produced a trifling cold are progressing to pneumonia severe enough to require respiratory support. “The kids are just not handling it well,” says Stacy Williams, a PICU nurse at UVA Health.
Coinfections, too, have always posed a threat—but they’ve grown more common with SARS-CoV-2 in the mix. “There’s just one more virus they’re susceptible to,” Vazquez told me. Each additional bug can burden a child “with a bigger hill to climb, in terms of recovery,” says Shelby Lighton, a nurse at UVA Health. Some patients are leaving the hospital healthy, only to come right back. There are kids who “have had four respiratory viral illnesses since the start of September,” Woodward told me.
Pediatric care capacity in many parts of the country actually shrank after COVID hit, Sallie Permar, a pediatrician at NewYork-Presbyterian and Weill Cornell Medicine, whose hospital was among those that cut beds from its PICU, told me. A mass exodus of health-care workers—nurses in particular—has also left the system ill-equipped to meet the fresh wave of demand. At UVA Health, the pediatric ICU is operating with maybe two-thirds of the core staff it needs, Williams said. Many hospitals have been trying to call in reinforcements from inside and outside their institutions. But “you can’t just train a bunch of people quickly to take care of a two-month-old,” Kudchadkar said. To make do, some hospitals are doubling up patients in rooms; others have diverted parts of other care units to pediatrics, or are sending specialists across buildings to stabilize children who can’t get a bed in the ICU. In Baton Rouge, Woodward is regularly visiting the patients who have just been admitted to the hospital and are still being held in the emergency department, trying to figure out who’s healthy enough to go home so more space can be cleared. His emergency department used to take in, on average, about 130 patients a day; lately, that number has been closer to 250. “They can’t stay,” he told me. “We need this room for somebody else.”
Experts are also grappling with how to strike the right balance between raising awareness among caregivers and managing fears that may morph into overconcern. On the one hand, with all the talk of SARS-CoV-2 being “mild” in kids, some parents might ignore the signs of RSV, which can initially resemble those of COVID, then get much more serious, says Ashley Joffrion, a respiratory therapist at Baton Rouge General Medical Center. On the other hand, if families swamp already overstretched hospitals with illnesses that are truly mild enough to resolve at home, the system could fracture even further. “We definitely don’t want parents bringing kids in for every cold,” Williams told me. The key signs of severe respiratory sickness in children include wheezing, grunting, rapid or labored breaths, trouble drinking or swallowing, and bluing of the lips or fingernails. When in doubt, experts told me, parents should call their pediatrician for an assist.
With winter still ahead, the situation could take an even darker turn, especially as flu rates climb, and new SARS-CoV-2 subvariants loom. In most years, the chilly viral churn doesn’t abate until late winter, which means hospitals may be only at the start of a grueling few months. And still-spotty uptake of COVID vaccines among little kids, coupled with a recent dip in flu-shot uptake and the widespread abandonment of infection-prevention measures, could make things even worse, says Abdallah Dalabih, a pediatric-intensive-care specialist at Arkansas Children’s.
The spike in respiratory illness marks a jarring departure from a comforting narrative that’s dominated the intersection of infectious disease and little children’s health for nearly three years. When it comes to respiratory viruses, little children have always been a vulnerable group. This fall may force Americans to reset their expectations around young people’s resilience and recall, Lighton told me, “just how bad a ‘common cold’ can get.”
Katherine J. Wu
Source link
This article was originally published in Undark Magazine.
Ten years ago, 12-year-old Rory Staunton dove for a ball in gym class and scraped his arm. He woke up the next day with a 104-degree Fahrenheit fever, so his parents took him to the pediatrician and eventually the emergency room. It was just the stomach flu, they were told. Three days later, Rory died of sepsis after bacteria from the scrape infiltrated his blood and triggered organ failure.
“How does that happen in a modern society?” his father, Ciaran Staunton, asked me.
Each year in the United States, sepsis kills more than a quarter million people—more than stroke, diabetes, or lung cancer. One reason for all this carnage is that if sepsis is not detected in time, it’s essentially a death sentence. Consequently, much research has focused on catching sepsis early, but the condition’s complexity has plagued existing clinical support systems—electronic tools that use pop-up alerts to improve patient care—with low accuracy and high rates of false alarm.
That may soon change. Back in July, Johns Hopkins researchers published a trio of studies in Nature Medicine and npj Digital Medicine showcasing an early-warning system that uses artificial intelligence. The system caught 82 percent of sepsis cases and significantly reduced mortality. While AI—in this case, machine learning—has long promised to improve health care, most studies demonstrating its benefits have been conducted using historical data sets. Sources told me that, to the best of their knowledge, when used on patients in real time, no AI algorithm has shown success at scale. Suchi Saria, the director of the Machine Learning and Healthcare Lab at Johns Hopkins University and the senior author of the studies, said in an interview that the novelty of this research is how “AI is implemented at the bedside, used by thousands of providers, and where we’re seeing lives saved.”
The Targeted Real-Time Early Warning System scans through hospitals’ electronic health records—digital versions of patients’ medical histories—to identify clinical signs that predict sepsis, alert providers about at-risk patients, and facilitate early treatment. Leveraging vast amounts of data, TREWS provides real-time patient insights and a unique level of transparency in its reasoning, according to the Johns Hopkins internal-medicine physician Albert Wu, a co-author of the study.
Wu says that this system also offers a glimpse into a new age of medical electronization. Since their introduction in the 1960s, electronic health records have reshaped how physicians document clinical information; nowadays, however, these systems primarily serve as “an electronic notepad,” he added. With a series of machine-learning projects on the horizon, both from Johns Hopkins and other groups, Saria says that using electronic records in new ways could transform health-care delivery, providing physicians with an extra set of eyes and ears—and helping them make better decisions.
It’s an enticing vision, but one in which Saria, the CEO of the company developing TREWS, has a financial stake. This vision also discounts the difficulties of implementing any new medical technology: Providers might be reluctant to trust machine-learning tools, and these systems might not work as well outside controlled research settings. Electronic health records also come with many existing problems, from burying providers under administrative work to risking patient safety because of software glitches.
Saria is nevertheless optimistic. “The technology exists; the data is there,” she says. “We really need high-quality care-augmentation tools that will allow providers to do more with less.”
Currently, there’s no single test for sepsis, so health-care providers have to piece together their diagnoses by reviewing a patient’s medical history, conducting a physical exam, running tests, and relying on their own clinical impressions. Given such complexity, over the past decade, doctors have increasingly leaned on electronic health records to help diagnose sepsis, mostly by employing a rules-based criteria—if this, then that.
One such example, known as the SIRS criteria, says a patient is at risk of sepsis if two of four clinical signs—body temperature, heart rate, breathing rate, white-blood-cell count—are abnormal. This broadness, although helpful for catching the various ways sepsis might present itself, triggers countless false positives. Take a patient with a broken arm: “A computerized system might say, ‘Hey, look, fast heart rate, breathing fast.’ It might throw an alert,” says Cyrus Shariat, an ICU physician at Washington Hospital in California. The patient almost certainly doesn’t have sepsis but would nonetheless trip the alarm.
These alerts also appear on providers’ computer screens as a pop-up, which forces them to stop whatever they’re doing to respond. So, despite these rules-based systems occasionally reducing mortality, there’s a risk of alert fatigue, where health-care workers start ignoring the flood of irritating reminders. According to M. Michael Shabot, a surgeon and the former chief clinical officer of Memorial Hermann Health System, “It’s like a fire alarm going off all the time. You tend to be desensitized. You don’t pay attention to it.”
Already, electronic records aren’t particularly popular among doctors. In a 2018 survey, 71 percent of physicians said that the records greatly contribute to burnout, and 69 percent said that they take valuable time away from patients. Another 2016 study found that, for every hour spent on patient care, physicians have to devote two extra hours to electronic health records and desk work. James Adams, the chair of the Department of Emergency Medicine at Northwestern University, calls electronic health records a “congested morass of information.”
But Adams also says that the health-care industry is at an inflection point to transform the files. An electronic record doesn’t have to simply involve a doctor or nurse putting data in, he says; instead, it “needs to transform to be a clinical-care-delivery tool.” With their universal deployment and real-time patient data, electronic records could warn providers about sepsis and various other conditions—but that will require more than a rules-based approach.
What doctors need, according to Shabot, is an algorithm that can integrate various streams of clinical information to offer a clearer, more accurate picture when something’s wrong.
Machine-learning algorithms work by looking for patterns in data to predict a particular outcome, like a patient’s risk of sepsis. Researchers train the algorithms on existing data sets, which helps the algorithms create a model for how that world works and then make predictions on new data sets. The algorithms can also actively adapt and improve over time, without the interference of humans.
TREWS follows this general mold. The researchers first trained the algorithm on historical electronic-records data so that it could recognize early signs of sepsis. After this testing showed that TREWS could have identified patients with sepsis hours before they actually got treatment, the algorithm was deployed inside hospitals to influence patient care in real time.
Saria and Wu published three studies on TREWS. The first tried to determine how accurate the system was, whether providers would actually use it, and if use led to earlier sepsis treatment. The second went a step further to see if using TREWS actually reduced patient mortality. And the third interviewed 20 providers who tested the tool on what they thought about machine learning, including what factors facilitate versus hinder trust.
In these studies, TREWS monitored patients in the emergency department and inpatient wards, scanning through their data—vital signs, lab results, medications, clinical histories, and provider notes—for early signals of sepsis. (Providers could do this themselves, Saria says, but it might take them about 20 to 40 minutes.) If the system suspected organ dysfunction based on its analysis of millions of other data points, it flagged the patient and prompted providers to confirm sepsis, dismiss the alert, or temporarily pause the alert.
“This is a colleague telling you, based upon data and having reviewed all this person’s chart, why they believe there’s reason for concern,” Saria says. “We very much want our frontline providers to disagree, because they have ultimately their eyes on the patient.” And TREWS continuously learns from these providers’ feedback. Such real-time improvements, as well as the diversity of data TREWS considers, are what distinguish it from other electronic-records tools for sepsis.
In addition to these functional differences, TREWS doesn’t alert providers with incessant pop-up boxes. Instead, the system uses a more passive approach, with alerts arriving as icons on the patient list that providers can click on later. Initially, Saria was worried this might be too passive: “Providers aren’t going to listen. They’re not going to agree. You’re mostly going to get ignored.” However, clinicians responded to 89 percent of the system’s alerts. One physician interviewed for the third study described TREWS as less “irritating” than the previous rules-based system.
Saria says that TREWS’s high adoption rate shows that providers will trust AI tools. But Fei Wang, an associate professor of health informatics at Weill Cornell Medicine, is more skeptical about how these findings will hold up if TREWS is deployed more broadly. Although he calls these studies first-of-a-kind and thinks their results are encouraging, he notes that providers can be conservative and resistant to change: “It’s just not easy to convince physicians to use another tool they are not familiar with,” Wang says. Any new system is a burden until proven otherwise. Trust takes time.
TREWS is further limited because it only knows what’s been inputted into the electronic health record—the system is not actually at the patient’s bedside. As one emergency-department physician put it, in an interview for the third study, the system “can’t help you with what it can’t see.” And even what it can see is filled with missing, faulty, and out-of-date data, according to Wang.
But Saria says that TREWS’s strengths and limitations complement those of health-care providers. Although the algorithm can analyze massive amounts of clinical data in real time, it will always be limited by the quality and comprehensiveness of the electronic health record. The goal, Saria adds, is not to replace physicians, but to partner with them and augment their capabilities.
The most impressive aspect of TREWS, according to Zachary Lipton, an assistant professor of machine learning and operations research at Carnegie Mellon University, is not the model’s novelty, but the effort it must have taken to deploy it on 590,736 patients across five hospitals over the course of the study. “In this area, there is a tremendous amount of offline research,” Lipton says, but relatively few studies “actually make it to the level of being deployed widely in a major health system.” It’s so difficult to perform research like this “in the wild,” he adds, because it requires collaborations across various disciplines, from product designers to systems engineers to administrators.
As such, by demonstrating how well the algorithm worked in a large clinical study, TREWS has joined an exclusive club. But this uniqueness may be fleeting. Duke University’s Sepsis Watch algorithm, for one, is currently being tested across three hospitals following a successful pilot phase, with more data forthcoming. In contrast with TREWS, Sepsis Watch uses a type of machine learning called deep learning. Although this can provide more powerful insights, how the deep-learning algorithm comes to its conclusions is unexplainable—a situation that computer scientists call the black-box problem. The inputs and outputs are visible, but the process in between is impenetrable.
On the one hand, there’s the question of whether this is really a problem: Doctors don’t always know how drugs work, Adams says, “but at some point, we have to trust what the medicine is doing.” Lithium, for example, is a widely used, effective treatment for bipolar disorder, but nobody really understands exactly how it works. If an AI system is similarly useful, maybe interpretability doesn’t matter.
Wang suggests that that’s a dangerous conclusion. “How can you confidently say your algorithm is accurate?” he asks. After all, it’s difficult to know anything for sure when a model’s mechanics are a black box. That’s why TREWS, a simpler algorithm that can explain itself, might be a more promising approach. “If you have this set of rules,” Wang says, “people can easily validate that everywhere.”
Indeed, providers trusted TREWS largely because they could see descriptions of the system’s process. Of the clinicians interviewed, none fully understood machine learning, but that level of comprehension wasn’t necessary.
In machine learning, although the specific algorithmic design is important, the results have to speak for themselves. By catching 82 percent of sepsis cases and reducing time to antibiotics by 1.85 hours, TREWS ultimately reduced patient deaths. “This tool is, No. 1, very good; No. 2, received well by clinicians; and No. 3, impacts mortality,” Adams says. “That combination makes it very special.”
However, Shariat, the ICU physician at Washington Hospital in California, was more cautious about these findings. For one, these studies only compared patients with sepsis who had the TREWS alert confirmed within three hours to those who didn’t. “They’re just telling us that this alert system that we’re studying is more effective if someone responds to it,” Shariat says. A more robust approach would have been to conduct a randomized controlled trial—the gold standard of medical research—where half of patients got TREWS in their electronic record while the other half didn’t. Saria says that randomization would have been difficult to do given patient-safety concerns, and Shariat agrees. Even so, he says that the absence “makes the data less rigorous.”
Shariat also worries that the sheer volume of alerts, with about two out of three being false positives, might contribute to alert fatigue—and potentially overtreatment with fluids and antibiotics, which can lead to serious medical complications such as pulmonary edema and antibiotic resistance. Saria acknowledges that TREWS’s false-positive rate, although lower than that of existing electronic-health-record systems, could certainly improve, but says it will always be crucial for clinicians to continue to use their own judgment.
The studies also have a conflict of interest: Saria is entitled to revenue distribution from TREWS, as is Johns Hopkins. “If this goes prime time, and they sell it to every hospital, there’s so much money,” Shariat says. “It’s billions and billions of dollars.”
Saria maintains that these studies went through rigorous internal and external review processes to manage conflicts of interest, and that the vast majority of study authors don’t have a financial stake in this research. Regardless, Shariat says it will be crucial to have independent validation to confirm these findings and ensure the system is truly generalizable.
The Epic Sepsis Model, a widely used algorithm that scans through electronic records but doesn’t use machine learning, is a cautionary example here, according to David Bates, the chief of general internal medicine at Brigham and Women’s Hospital. He explains that the model was developed at a few health systems with promising results before being deployed at hundreds of others. The model then deteriorated, missing two-thirds of patients with sepsis and having a concerningly high false-positive rate. “You can’t really predict how much the performance is going to degrade,” Bates says, “without actually going and looking.”
Despite the potential drawbacks, Orlaith Staunton, Rory’s mother, told me that TREWS could have saved her son’s life. “There was complete breakdown in my son’s situation,” she said; none of his clinicians considered sepsis until it was too late. An early-warning system that alerted them about the condition, she added, “would make the world of difference.”
After Rory’s death, the Stauntons started the organization End Sepsis to ensure that no other family would have to go through their pain. In part because of their efforts, New York State mandated that hospitals develop sepsis protocols, and the CDC launched a sepsis-education campaign. But none of this will ever bring back Rory, Ciaran Staunton said: “We will never be happy again.”
This research is personal for Saria as well. Almost a decade ago, her nephew died of sepsis. By the time it was discovered, there was nothing his doctors could do. “It all happened too quickly, and we lost him,” she says. That’s precisely why early detection is so important—life and death can be mere minutes away. “Last year, we flew helicopters on Mars,” Saria says, “but we’re still freaking killing patients every day.”
Simar Bajaj
Source link