Jared Kaltwasser

home

authors

jared-kaltwasser

A patient rests comfortably in a hospital bed. Everything seems fine. Neither the patient nor the doctor or nurses realize he is about to take a life-threatening turn for the worse. Within hours, he will be septic, and his care team will race to begin aggressive fluid resuscitation and antibiotics

Yet, although the patient’s doctor and nurses do not see sepsis coming, someone else does — or, more accurately, some

 else does. The patient’s electronic health record (EHR) includes an artificial intelligence (AI) tool that detects patterns imperceptible to the patient’s human caretakers. It duly sends an alert to his doctor and nurses. What could have been a shocking deterioration is instead averted.

It sounds futuristic, this kind of thing that AI could do. But algorithms are already detecting hard-to-anticipate conditions such as sepsis in hospitals, and lives are already being saved. In fact, the underlying idea — that a set of data, put together in context, can help predict outcomes — is already very much a part of medical thinking.

“A lot of these concepts, they’re not new to medicine,” says Yasir Tarabichi, M.D., MSCR, director of clinical informatics for research support at MetroHealth, a safety net healthcare system in Cleveland.

AI and machine learning based on huge medical data sets are just the latest iteration of risk stratification, Tarabichi says, which is something as old as medicine itself. Before AI, patients could be given risk scores based on simple scoring tools and scales. AI has created the opportunity to score patients using thousands of measurements not previously available, much less analyzable. “It was simpler before, because it was more of a point-based system,” Tarabichi says. “Now it (has become) a lot more complicated.”

With opportunity comes challenges. One of the main challenges with AI is dealing with the onslaught of new products. Prem Thomas, M.D., medical director of data analytics and medical information officer at Yale New Haven Health in Connecticut, said he and colleagues get visits from software representatives the way doctors have historically received visits from drug company sales reps. “It’s interesting,” he says. “We have our weekly medical information officers’ meeting, and typically there’s somebody (from a technology vendor who) wants to present to us during that time.”

The companies hawking their AI wares range from entrenched players such as Epic Systems, the largest EHR company in the United States, which markets a sepsis prediction model, to tiny startups that have little, if any, experience in healthcare. Increasingly, tech giants such as Apple and Google have tried to gain a foothold in the healthcare technology sector, often with an
AI focus.

The utility of AI in healthcare settings extends beyond clinical decision support tools such as the sepsis model. One area Thomas has been examining uses AI to make operating room scheduling more efficient. Currently, the time slots that surgeons request for routine procedures such as appendectomies are often imprecise.

“The surgeon asked for two hours,” Thomas says, “but based on the data for that surgeon and the location where they are going to be (performing the operation), can we be more accurate in how long it’s going to take and, hopefully, minimize things like delays for other patients and other surgeons?”

Houston Methodist Hospital turned to AI when it was preparing to receive COVID-19 vaccines, according to LeTesha Montgomery, M.H.A., RN, the hospital’s vice president for operations and patient access. The hospital knew they would be facing a “tsunami of phone calls” from patients and members of the public asking questions or wanting to sign up for the vaccines, Montgomery says. “With predictions of increased volume reaching 300% to 400%, hospital leaders knew we needed a solution that would manage the flood of vaccine-related phone calls without impacting usual operations,” she says.

The hospital used Syllable, a conversation AI voice assistant, to automate work flow and conserve resources by making it easier for patients to find a wide array of answers without needing to speak to a human being. It also used AI to handle in-person traffic. Visitors to the facilities were screened by the temperature-monitoring platform Care.ai by standing in front of a tablet equipped to scan for fevers. “This contactless technology not only helped with conserving staff resources, but it also speeds up screening measures at point of entry,” Montgomery says.

One reason Houston Methodist was able to implement its COVID-19 strategies fast is because it has set up an infrastructure to quickly review and vet new technologies. The Center for Innovation, launched in 2018, has a group called DIOP, which stands for Digital Innovation Obsessed People, made up of roughly half operational staff and half information technology staff. The group is charged with developing strategies for broad-scale implementation or “failing fast” — quickly identifying technology that is not a good fit. “The results of focusing on innovation before it was essential to a pandemic such as COVID-19 enabled our hospital system to embrace new technology, including AI, more seamlessly,” Montgomery says.

Still, the stakes for AI implementation vary depending on its proposed use. A temperature-screening device that stops working might mean a staff member gets temporarily reassigned to thermometer duty. An errant AI system designed to help clinicians make treatment decisions could harm patients, even kill them.

Tarabichi said it is important that the public understands that hospitals do not take AI lightly.

“We don’t just blindly adopt new technology,” he says. “We make sure it’s going to work for us.”

For instance, before implementing the Epic sepsis model, MetroHealth ran the program in the background for an entire year, disabling the alerts but collecting data. Before they toggled it live, the health system wanted to make sure it worked and had a meaningful benefit. At the end of the year, they adopted the technology but limited its use to the emergency department.

Thomas says that kind of slow, evaluative approach is key to AI integration. It means taking the time to understand how an AI system works, getting an idea of the data sets used by the vendor and tailoring the program’s settings to the health system’s needs.

In other words, it takes time and money — and more time. But Thomas says the expense is worth it.

“I would recommend that community hospitals or academic hospitals invest the resources in having a redesign committee review what this model is purported to do for you and then think through how to operationalize it,” he says. Yale New Haven’s redesign committee, for instance, is an interdisciplinary team whose members include intensive-care unit nurses, emergency department doctors, inpatient nurses and other professionals, as well as analytics staff.

But a health system’s ability to carefully evaluate an AI system isn’t just a matter of staff expertise and representation. It also requires having access to enough data and software to fully understand a product. Because AI products are proprietary software marketed by for-profit companies, there can be tension between the company’s desire to keep proprietary information private and a healthcare system’s need to fully understand the tools they are using to take care of patients. 

An example of that kind of conflict had a public airing this summer when investigators from the University of Michigan reported findings of their independent evaluation of Epic’s sepsis prediction tool in 

. Their research suggested that the product is less accurate at predicting sepsis than what the company claims in its marketing materials. The findings were controversial, in part because the sensitivity of the AI tool depends on the settings put in place. Epic disputed the findings, saying that it works closely with users to ensure the product meets their needs and to provide access to the data necessary to evaluate the model.

Tarabichi has submitted his own paper, soon to be published, which contradicts some of what the Michigan investigators found. Yet, even as experts disagree about the particular product, this raises important questions about the limits of using proprietary models.

“The increase and growth in deployment of proprietary models has led to an underbelly of confidential, non-peer-reviewed model performance documents that may not accurately reflect real-world model performance,” wrote the corresponding author of University of Michigan study, Karandeep Singh, M.D., M.M.Sc., and his colleagues.

In the case of Epic’s sepsis model, both Thomas and Tarabichi said they were satisfied that they had access to the information they needed to evaluate the product, though they both said information from the company was just one component of their implementation process.

Tarabichi cautions against hospitals “jumping on the bandwagon,” of using popular AI without first validating it and testing it in their own system. In a smaller community hospital setting, validation may not need to be as complicated as it might at a top academic medical center with a diverse patient population.

“Randomized controlled trials are the highest level of evidence, but lower down that hierarchy is pre- and post- assessment,” he says. “Run it for a couple of months and then evaluate. What changed? Can you do better?”

For Thomas, it is not just about choosing the right AI system but also about being thoughtful about when and where AI is actually needed. He and his colleagues reference a “Mad Lib” created by Michael Draugelis, the chief data scientist at Penn Medicine in Philadelphia. It goes like this, Thomas says: “If I were [a physician, a physician’s assistant, a nurse, an operating room scheduler] in the health system and I knew ____, I would do ___ to change ____. “Before we start on any project now, we say, ‘fill in the blanks,’ ” Thomas says. “Because if we don’t have a good answer for those, then we shouldn’t be investing the time and effort to go through with the project.”

Montgomery agrees, saying technology cannot displace patients’ status “at the center of everything we do.”

“AI isn’t meant to solve all the challenges in the healthcare setting,” she says, “but where these applications make sense, where they provide real benefits to our patients, that means it’s reaching its full potential.”

Articles

Artificial intelligence is improving healthcare methods and insights. But when used incorrectly, AI is rife with pitfalls.

AI: Beyond the Buzz 

A new indication for a well-known drug could lead to relief for patients with a rare sleep condition, but the approval is also calling to mind some of the problem associated with the drug, which will only be dispensed through a risk-mitigation protocol.

The FDA last month granted a new indication for Xywav (calcium, magnesium, potassium, and sodium oxybates), a drug whose active ingredient is a form of gamma hydroxybutyrate (GHB). Xywav had already been approved to treat cataplexy or excessive daytime sleepiness in patients ages 7 and older. The new indication is for idiopathic hypersomnia in adults, a rare condition in which patients experience excessive daytime sleepiness even when they get plenty of sleep.

“We are excited that with today's approval Xywav will become the first and only medicine indicated to treat idiopathic hypersomnia, a unique medical condition that can have significant effects on the lives of those diagnosed with the condition," said Bruce Cozadd, MBA, the chief executive at Jazz Pharmaceuticals, 

In clinical trials, the drug out-performed placebo, generating a statistically significant improvement in Epworth Sleepiness Scale and Idiopathic Hypersomnia Severity Scale scores among patients .

Xywav is an oral solution similar to Xyrem (sodium oxybate), which has also been approved to treat cataplexy and excessive daytime sleepiness. However, Xywav has 92% less sodium in each nightly dose compared to Xyrem. Both Xywav and Xyrem are marketed by Jazz Pharmaceuticals.

“Idiopathic hypersomnia is a life-long condition, and the approval of Xywav will be instrumental in providing treatment for symptoms such as excessive sleepiness and difficulty waking, and in effectively managing this debilitating disorder,” 

 Eric Bastings, M.D., the deputy director of the Office of Neuroscience at the FDA’s Center for Drug Evaluation and Research.

The drug comes with the potential for side effects, the most common adverse events reported in the trial were nausea (21.4%), headache (16.2%), dizziness (11.7%), anxiety (10.4%) and vomiting (10.4%), the FDA said.

The drug’s potential for abuse that has caused some concern. Before it was approved for sleep-related conditions, illicit versions of GHB were used recreationally, and sometimes maliciously. As a 2014 report from the 

reported, the drug was one of the first known “date-rape drugs.”

GHB is now a Schedule 1 controlled substance.

The FDA approval came with conditions. The drug is subject to a risk evaluation and mitigation strategy (REMS) protocol, meaning prescribers must be certified before they can write prescriptions for the drug, and only certified pharmacies can dispense it. The drug will be available only at pharmacies that ship directly to patients.

Jazz Pharmaceuticals has also launched a Nurse Case Management program, in which nurses are available to answer patient questions and support patients taking Xywav or Xyrem.

The company has also set up a savings program and a patient assistance program to help patients obtain the drug, which 

 can cost up to $100,000 per year for nightly treatments.

FDA Grants New Indication to Sleep-Disorder Drug Xywav

The research is in: people who live sedentary lifestyles face a higher risk of obstructive sleep apnea (OSA). However, getting the research from “the bench to the clinic” in a meaningful way can be unusually challenging in this increasingly central corner of sleep medicine.

At least two issues are at play. First, the remedy for the risk factor is exercise and an active lifestyle, something patients who do not lead active lifestyles are all too used to hearing from their doctors. The second problem, though, is that most of the research has focused on risk factors and on treating patients who already have OSA, rather than on preventing the condition before it develops.

Tianyi Huang, ScD, MSc, said it is clear exercise has a number of health benefits, including improving inflammation and reducing obesity. However, Huang told 

 that there are also some specific mechanisms by which exercise can help OSA.

“For example, an active lifestyle during the day helps reduce excessive body fluid retention in the lower body,” said Huang, an assistant professor and associate epidemiologist at Brigham and Women’s Hospital and Harvard Medical School. “At night, when people lie down to sleep, these excessive body fluids in the lower body can flow back to the upper part of the body, pressure the lung to reduce its volume and cause sleep apnea.”

Huang and colleagues recently investigated links between exercise, a sedentary lifestyle, and OSA risk. The results, 

 were based on more than 138,000 people who were tracked for between 10 and 18 years. About 8,700 of the patients were eventually diagnosed with OSA.

Huang and his co-authors found that people who did exercise equivalent to 3 hours of running each week had a 54% lower risk of OSA compared to people who did activities equivalent to 2 hours of walking per week.

Similarly, people who sat and watched TV for more than 4 hours per day had a 78% higher risk than those in the least sedentary category of people in the study. Those who sat for long periods of time for work had a 49% higher risk than the least sedentary group.

The good news is that all of these risk factors are modifiable. People have the power to reduce their OSA risk by becoming more active, the authors found. Huang noted that while obesity is a risk factor for OSA, even people who are not obese can benefit from increasing their activity.

“[W]e found more physical activity and less sedentary behavior are also associated with lower OSA risk even among normal-weight individuals,” Huang said.

However, recommendations for more physical activity are already a standard part of primary care for patients with sedentary lifestyles or obesity. Huang said physicians need to be on the lookout for patients who specifically are at risk for OSA.

“Clinicians should be aware of this evidence and recommend exercise-based interventions to those at risk of OSA (e.g., heavy snorers),” Huang said.

In fact, other research has suggested that even if a patient develops OSA, increasing physical activity can lower the severity of OSA and improve related symptoms.

Though physicians already have evidence to help them lower OSA risk, Huang said more should be done. While OSA has garnered significant attention from the research community, the research is usually focused on risk factors or treatment.

“I think there is not sufficient focus on prevention of OSA, either in clinical practice or in research,” Huang said. “This is in contrast to other chronic diseases, for which large efforts have been put to identify modifiable risk factors for effective prevention strategies.”

Huang said ultimately OSA prevention warrants a greater share of attention, both from researchers and from the general healthcare community.

A new study gives physicians the power to help patients at risk of obstructive sleep apnea, but it also highlights the relative lack of public attention on prevention.

Lowering Risk of Obstructive Sleep Apnea Prevention Is Possible, But More Work Is Needed

A new study finds that a potential new treatment strategy for certain patients with non-small cell lung cancer (NSCLC) does not increase survival, but the study’s co-author says the research may help pave the way for other new strategies.

The report concerns the use of hypofractionated image-guided radiotherapy in patients with stage II/III NSCLC who are unfit for concurrent chemotherapy due to comorbidities or performance status.

Puneeth Iyengar, MD, PhD, of the University of Texas Southwestern Medical Center, told 

 the preferred treatment for patients with this type of lung cancer is chemotherapy and radiotherapy followed up with immunotherapy.

“The problem is, there are quite a few of our lung cancer patients who, because of their lung function, because of other medical morbidities, they cannot tolerate doing chemotherapy and radiation at the same time,” he said.

Since these patients could not tolerate chemotherapy, Iyengar and colleagues wanted to know whether escalating the dosage of radiation for these patients might yield better results than conventionally fractionated radiotherapy. Their idea was to give concurrent chemoradiotherapy-ineligible patients radiation doses of 60 Gy in 15 fractions instead of the conventional 60 Gy in 30 fractions.

“And the idea was that if we could somehow increase the dose per treatment, we could make up—in terms of controlling the local disease—for the fact that there was no chemotherapy on board,” he said.

The investigators had already completed a phase I study showing the escalated dosage was tolerable. The new study, 

, was a nonblinded phase 3 randomized clinical trial designed to find out if the therapy would boost one-year overall survival.

The investigators enrolled 103 patients, 96 of whom were able to be analyzed. All of the patients had phase II/III NSCLC, Zubrod performance statuses of at least 2, and greater than 10% weight loss in the previous 6 months, or they were otherwise ruled ineligible for concurrent chemoradiotherapy by oncologists.

Fifty patients were given hypofractionated radiotherapy and 46 were treated with conventionally fractionated radiotherapy. Patients were enrolled between between 2012 and 2018 and were followed for a median of 8.7 months.

Despite the positive phase I study results, Iyengar said a planned interim analysis of the new phase III study made clear that the strategy was not going to reach the primary endpoint of prolonging survival. There was no statistically significant difference in one-year survival between the two groups. Those receiving hypofractionated radiotherapy had a one-year survival rate of 37.7% (95% CI, 24.2%-51.0%) versus 44.6%(95% CI, 29.9%-58.3%) in the conventionally fractionated group. Similarly, the new strategy showed no benefit in median overall survival or progression-free survival.

Though the study did not meet its primary endpoint, Iyengar said the trial generated meaningful data that could someday be used to improve patient outcomes.

“I would argue this is a good foundation to build off of and this could be the tip of the iceberg,” he said.

Iyengar noted that there were some signs in the study that the therapy was doing a good job of controlling disease, even if that did not translate to better survival.

“So it may have been doing its job with respect to controlling the disease, but not with respect to survival because of the toxicity relevant to that patient population,” he said.

Thus, Iyengar added, the research may help inform different strategies and different combinations of treatments for particular populations, including potentially a combination with immunotherapy. He said other research is already underway using different “formulas” of treatment, but all probing the same underlying questions about when and whether hypofractionation might improve outcomes.

Iyengar said despite failing to meet its endpoint, the study could be a long-term success, because it generated data that could help optimize treatment in the future.

“I sometimes feel that we don't take away enough from the studies that don't work the way that we anticipated they would work,” he said. “I think there's a lot of information that can be garnered from those types of studies.”

Though the study did not meet its endpoint, the strategy behind it remains under investigation.

Jared Kaltwasser

Articles

AI: Beyond the Buzz

FDA Grants New Indication to Sleep-Disorder Drug Xywav

Lowering Risk of Obstructive Sleep Apnea Prevention Is Possible, But More Work Is Needed

Hypofractionated Radiotherapy Fails to Extend Survival in NSCLC Patients Ineligible for Concurrent Chemoradiotherapy