What happened in health care technology this week – and why it’s important.
The baby was gravely ill. Rapid sequencing turned around a diagnosis in 13 hours — and pointed to a treatment
Andrew Joseph published an article in STAT reporting on the use of rapid DNA sequencing in a pediatric case at Rady Children’s Hospital in San Diego. When the boy was brought to the San Diego emergency department one night last October, he was inconsolable. Within about half an hour, clinicians had a clue of what was wrong: A CT scan showed signs of disease in the 5-week old’s brain.
Less than 36 hours after the boy showed up to the ER, they had their diagnosis: a rare disease called THMD2 that was easily treatable. From starting the sequencing to the diagnosis, the process took just 13 hours — a figure unimaginable even a few years ago. That highlights the remarkable progress in sequencing technology and how quickly it can turn around answers. It wasn’t just that the boy got his genome decoded so quickly. The sequencing flagged misspellings in his DNA that in turn pointed to a specific diagnosis; often, reading someone’s genes doesn’t turn up anything so clear-cut.
Why it’s important – It’s a proof of concept of what can happen when you have access to this compelling test. The case underscores the power of rapid sequencing when used for the correct type of patient. It’s important to note, however, that it is not typical of every child’s experience. Less-resourced hospitals might not have access to the technology. Insurance can also get in the way. Only about a third of sick babies with a suspected genetic disease who have their genomes sequenced get a firm diagnosis. And research suggests that only 10% of those babies have treatment options once the condition is identified. The technology is so new that experts are still trying to determine how widely rapid sequencing should be made available and to which patients. To get insurance companies on board, they’ll need to show that broader access to sequencing can save money overall or lead to better outcomes.
Epic’s AI algorithms, shielded from scrutiny by a corporate firewall, are delivering inaccurate information on seriously ill patients
In a STAT investigative report (subscription required), Casey Ross reported on issues that have surfaced on the use of algorithms developed by Epic on their electronic health record systems. Several major health systems employees said they were particularly concerned about Epic’s algorithm for predicting sepsis, a life-threatening complication of infection. The algorithm, they noted, routinely fails to identify the condition in advance and triggers frequent false alarms. Some hospitals reported a benefit for patients after fine-tuning the model, but that process took at least a year. (If you don’t have a STAT subscription, you can get a look at the issue in this accessible article from Fast Company from earlier this year.
STAT’s investigation, based on interviews with data scientists, ethics experts, and many of Epic’s largest and most influential clients, underscores the need for extreme caution in using artificial intelligence algorithms to guide the care of patients. Errant alarms may lead to unnecessary care or divert clinicians from treating sicker patients in emergency departments or intensive care units where time and attention are finite resources.
Even more disturbing, STAT’s investigation also found that Wisconsin-based Epic, whose software holds the medical records of some 250 million people, has paid some health systems up to $1 million in cash incentives based, in part, on their adoption of predictive algorithms developed by Epic or others. Those payments may create a conflict between duties to deliver the best care to patients and preserve their bottom lines.
“Epic will crow about its [algorithm] adoption rates. But if the adoption rates are essentially paying for adoptions, it’s a little bit of a different problem.”
Glenn Cohen, director of the Petrie-Flom Center for Health Law Policy, Biotechnology and Bioethics at Harvard University.
Why it’s important – This latest report reinforces why it is vital to have complete transparency in how algorithms are developed, what data sets were used to train the algorithms, and how they perform in individual cases at the front lines of health care. Companies like Epic need to better explain the limitations and potential hazards of algorithms and make sure hospitals know they should be evaluated on their patients. Some also said that it should no longer be up to Epic, or any other algorithm developer, to decide precisely how much information is disclosed or the extent of testing that ought to be performed.
King’s College Researchers Develop Deep Learning Approach to Labeling MRI Scans At Scale
Researchers at King’s College in London have made a breakthrough in automating the process, which means up to 100,000 MRI brain examinations can be labeled in a matter of 30 minutes rather than the years it would take to complete manually. This was done by training the systems to obtain the labels from radiology reports and correctly assign them to their corresponding MRI examinations.
Dr. David Wood, the study’s lead author, said the research builds on recent breakthroughs in natural language processing, which had used huge collections of unlabeled text, including all of the English language Wikipedia and (open-access research repository) PubMed Central abstracts and full-text articles.
Why it’s important – Besides the approach to training the algorithms using a deep learning approach, the researchers have made their code and models available to other researchers so that as many organizations can benefit from their work as possible.
Investigational Magnetic Device Shrinks Glioblastoma In First-In-World Human Test
In a news release from the Houston Methodist Neurological Institute, researchers from the department of neurosurgery shrunk a deadly glioblastoma tumor by more than a third using a helmet generating a noninvasive oscillating magnetic field that the patient wore on his head while administering the therapy in his own home. The 53-year-old patient died from an unrelated injury about a month into the treatment, but during that short time, 31% of the tumor mass disappeared. The autopsy of his brain confirmed the rapid response to the treatment.
The treatment consisted of intermittent application of an oscillating magnetic field generated by rotating permanent magnets in a specific frequency profile and timing pattern. First administered for two hours under supervision in the Peak Clinic, ensuing treatments were given at home with help from the patient’s wife, increasing treatment times up to a maximum of only six hours per day.
Why it’s important – Glioblastoma is the deadliest of brain cancers in adults, nearly always fatal, with a life expectancy of a few months to two years. If proven in additional clinical trials, this technique could treat brain cancer without radiation therapy or chemotherapy. It could open a new world of non-invasive and nontoxic treatment for brain cancer, with many interesting possibilities for the future.
Dollar General Announces Plans to Increase Access to Health Care Products & Services, Expanding Health Offerings in Rural Communities Across America
Add Dollar General to the list of companies looking to expand their footprint in health care. In a press release this week, the company announced plans to expand its offerings in health care. This will include an increased assortment of cough and cold, dental, nutritional, medical, health aids, and feminine hygiene products across many of its Dollar General stores. This effort marks the first significant step in Dollar General’s strategic journey toward increasing access to healthcare offerings.
The company also announced the appointment of Albert Wu, M.D., to the newly created position of Vice President, Chief Medical Officer. In this role, Dr. Wu will help develop DG’s new health care services efforts by establishing and strengthening relationships with current and prospective healthcare product and service providers to build a comprehensive network of affordable services for DG customers. Dr. Wu joins Dollar General from McKinsey & Company, where he has worked since 2016. During that time, he led a team to create a total cost of care model for 250,000 rural healthcare patients; oversaw multiple hospital provider turnarounds totaling $2-5 billion in revenue; designed a digitally-driven healthcare insurance product offering; and provided guidance to analysts modeling ventilation and pharmaceutical needs to support pandemic relief efforts.
Why it’s important – With 75% of the U.S. population living within approximately five miles of one of Dollar General’s 17,000+ stores, the Company provides access to rural communities often underserved by other retailers as well as the existing healthcare ecosystem. Couple this work with Walmart’s health care initiatives (even though they have been slowed by the pandemic and a change in top management), and you have two major commercial players that have deep roots in underserved rural markets that could make a big dent in the critical access issues facing those people.
Ultrasound Patch Monitors Blood Flow
Medgadget’s Conn Hastings reported on research conducted at the University of California San Diego. Researchers created an ultrasound patch that can measure blood flow in vessels as deep as 14 cm within the body. The stretchy patch can be applied to the skin and may help clinicians monitor and diagnose various conditions, including blockages that could cause an infarct. The patch contains an array of ultrasound transducers that can measure blood flow in vessels directly beneath it, and the ultrasound beam can also be steered to assess vessels that are nearby but not directly below.
“Just stick it on the skin, then read the signals. It’s not operator dependent, and it poses no extra work or burden to the technicians, clinicians or patients.”
Sai Zhou, Researcher, University of California San Diego
Why it’s important – Although very early in development, using this flexible device allows for the monitoring of blood flow in specific vessels and can help clinicians diagnose various cardiovascular diseases. For instance, measuring changes in blood flow in the carotid artery could show that someone is at a risk of stroke and help to initiate treatment before a stroke occurs. So far, the researchers have tested the patch and have shown that it is as accurate as a commercial ultrasound device.