What happened in health care technology this week, and why it’s important.
The Next Frontier For Large Language Models Is Biology
Rob Toews in Forbes writes that large language models like GPT-4 have taken the world by storm thanks to their astonishing command of natural language. Yet the most significant long-term opportunity for LLMs will entail an entirely different type of language: the language of biology. One striking theme has emerged from the long march of research progress across biochemistry, molecular biology, and genetics over the past century: it turns out that biology is a decipherable, programmable, in some ways even digital system. “Protein language models”—LLMs trained not on English words but instead on protein sequences—have demonstrated an astonishing ability to intuit the complex patterns and interrelationships between protein sequence, structure, and function: say, how changing certain amino acids in certain parts of a protein’s sequence will affect the shape that the protein folds into. Protein language models can, if you will, learn the grammar or linguistics of proteins.
Why it’s important – Pointing large language models at biological data—enabling them to learn the language of life—will unlock possibilities that will make natural language and images seem almost trivial by comparison. In the near term, the most compelling opportunity to apply large language models in the life sciences is to design novel proteins. All the proteins that exist today represent but an infinitesimally tiny fraction of all the proteins that could theoretically exist. Herein lies the opportunity. This domain is thus ripe for massive scaling efforts powered by LLMs—efforts that may result in astonishing emergent insights and capabilities in protein science.
Infographics of the week – Chen et al. have presented a vision of future home-centered geriatric care powered by digital technologies and devices. A network of internet-connected sensors on the body and distributed around the home monitors the health conditions of older adults and transmits rich, dynamic data to cloud servers. Machine learning algorithms then analyze the data to coordinate with the remote caregiver and with autonomous wearable therapeutic devices toward optimal health care.
This week’s second infographic is from Dr. Bertalan Mesko and the Medical Futurist Institute, highlighting the key differences between traditional drug-based therapies and digital therapeutics. This graphic visualization captures how this new approach to treatment could help us achieve the same old goal of assisting patients in getting better.
A Diagnostic Test Can Detect a Stroke Within 15 Minutes
Seeking a way to reduce the mortality rate for large vessel occlusions (LVOs) – the most common stroke – Cambridge, England-based startup Upfront Diagnostics has created a rapidly deployed, mobile, and highly accurate stroke detection kit. The company discovered new blood biomarkers that can detect strokes and, with that knowledge, built the LVOne. The handheld device uses a single drop of blood from a finger prick to diagnose an acute ischaemic stroke caused by LVO. It is designed for use on the go in an ambulance on the way to a hospital A&E. Paramedics administering the test receive results in 15 minutes or less.
Why it’s important – The World Stroke Organization’s Global Fact Sheet 2022 revealed that the absolute number of stroke cases increased “substantially” between 1990 and 2019. And with that growth, a correspondingly high financial cost is borne by health services. With LVOs responsible for up to 96 percent of stroke deaths, identifying an occurrence en route to the hospital relieves some of the burden on the receiving care team. It allows the patient to access time-sensitive treatment as quickly as possible.
Podcast of the week – This “Unfiltered” episode of Fixing Healthcare welcomes Dr. Jonathan Fisher, a respected cardiologist and renowned advocate for physician well-being. Alongside co-host Jeremy Corr, Dr. Robert Pearl starts the conversation by noting that many healthcare conferences focus on clinician burnout, including several events hosted by Dr. Fisher, which strive to relieve the pain doctors and nurses experience daily. And simultaneously, Pearl cites the pain of patients who are increasingly unable to afford the cost of health insurance and healthcare. You can listen to the podcast here: https://www.fixinghealthcarepodcast.com/2023/07/18/fhc-99-unfiltered/
Ticks and the Diseases They Carry Are Spreading. Can This Drug Stamp Them Out?
A small study showed that feeding deer a type of ivermectin reduced the number of ticks drinking their blood. Wired’s Maryn Mckenna reports that a team of scientists working within the Connecticut state government has found a new application for something ivermectin was always good at killing parasites that live in or on animals. In a small study, they found that feeding deer corn dosed with an ivermectin derivative builds up enough drug in their blood that ticks biting them will (probably) die. That’s the thinking, anyway. There are several experimental steps still to go—and not everyone in the world of ticks is convinced it’s a practical approach.
Why it’s important – This will be good news if the finding holds up while the researchers expand the study into more and larger sites. That’s because drinking blood from deer is what gives adult female ticks the boost they need to lay thousands of eggs that will turn into more ticks—and because the combination of erupting deer populations and expanding tick ranges is driving an explosion of tick-borne diseases across the US. The work has attracted the interest of the Centers for Disease Control and Prevention, whose Division of Vector-Borne Diseases has given the group a five-year grant. “The preliminary work looked good at the proof-of-concept stage,” says Lars Eisen, a research entomologist in the Vector-Borne Diseases Division. “This is additional funding to do a larger-scale field trial in Connecticut, in an inland setting, and an island setting in Maine.”
‘Hospital-at-home’ trend means family members must be caregivers — ready or not
Hospital-at-home programs are for people sick enough to need the attention a hospital provides but stable enough to be cared for at home. NPR’s Kat McGowan reports that research on outcomes is inconclusive, but it shows promise that it can provide good care and save healthcare dollars. But a big question looms: What about the family? Are unpaid, untrained family caregivers ready to take on the responsibility of overseeing a critically ill person at home — even with backup from visiting clinicians? This question is about to become more critical. For decades, hospital-at-home was a small-scale experiment. During the COVID pandemic, the idea went mainstream. In November 2020, the federal government changed rules so that hospitals could be paid the same amount to treat patients at home. Today, 290 hospitals in 37 states have signed up.
“A huge part of the process is making sure that both patients and their caregivers have a really good idea of what they’re getting into.”
Margaret Paulson, Mayo Clinic Health System, Wisconsin
Why it’s important – For a caregiver, hospital-at-home eliminates the hardship that comes with a loved one’s stint in the hospital, everything from overpriced parking to hours spent at bedside waiting for a doctor to make rounds. But it brings new concerns. These programs don’t ask caregivers to handle medical tasks like dealing with an IV. But they might need to bring glasses of cold water in the middle of the night, help a weak person turn over in bed, change clothes, or get to the bathroom. In the hospital, nursing aids do those tasks. Programs can arrange home health aides for help, usually for limited hours. It could be part of the reason why between 10 and 62 percent of people turn down the option to participate in these new programs. A recent AARP policy briefing delved into the implications. Hospital-at-home has strong potential, but the effects on caregivers need more attention, the report advises. “The family caregivers are completely invisible,” says Reinhard of AARP. “They’re not turning to wife, daughter, or husband and saying: Can you handle this? That’s the discussion we think needs to happen.” The briefing recommends federal policymakers require programs to ensure family members know what will happen and are on board.
Hey Tech, It’s Time To Build. In Healthcare.
Here’s what sounds crazy in 2023: Trying to fix the United States’ gigantic mess of a healthcare system. But Daisy Wolfe and Vijay Pande argue that now is the time to do just that in their post on a16z.com. The most impactful companies are built at the frontier, and healthcare is the next frontier. It’s time to put our tech skills to work. The mother of all markets is ripe for disruption. Healthcare needs tech. Healthcare, at its core, is (1) a data, operations, and logistics problem and (2) a consumer experience and engagement problem. Both are areas where the tech world excels. Moreover, healthcare provides the best opportunity to use the hottest tools in tech–especially AI–to displace huge incumbents.
Why it’s important – As the authors point out: When you think about the limited number of days you have on this planet and the insane number of hours you’ll pour into building a startup, wouldn’t it be great if you were working on something that really mattered? You don’t have to be a doctor or cancer researcher to save lives. A company detecting medical errors or helping people afford medical care might save countless lives. Working to improve healthcare is an obvious way to find purpose in one’s work. Every American can relate personally to how broken the American healthcare system is. The advantages of coming from tech are clear. Tech moves faster than healthcare. The tech world produces visionaries who don’t accept that the ways things work today will be how things work in the future. Your fresh eyes are exactly what healthcare needs.
Finally, next week is MRI Safety Week. Despite over two decades of dedicated efforts in MRI safety, incidents, accidents, and near misses continue to occur with unsettling frequency. While our systems and processes have undoubtedly progressed, certain aspects of our outcomes have not kept pace. Metrasens offers a week-long series of events to explore our challenges and seize the opportunities that lie ahead, forging a path toward enhanced MRI safety standards. You can access those events here.
