What happened in health care technology this week, and why it’s important.
A surprising finding about taking blood pressure lying down
The simple act of having someone lie down for a blood pressure reading might reveal more than expected about their heart health, preliminary research has found. Michael Merschel, from American Heart Association News, reports on the investigation. Using data from a large, long-running study, researchers discovered that when compared with readings taken while someone was sitting, readings that showed high blood pressure in people who were lying down did a better job of predicting stroke, serious heart problems, and death.
Why it’s important – The findings imply that checking supine blood pressure might unveil hypertension that would otherwise be missed in the doctor’s office. Previous work has shown that nighttime blood pressure independently predicts cardiovascular outcomes. Still, it has been unclear whether that was related to the timing of the readings or the position of the person being measured.
Infographic of the week – This week’s selection comes from PatientOne online. The Internet of Medical Things (#IoMT) is reshaping the healthcare landscape, particularly in remote patient care. By seamlessly connecting #medicaldevices and #sensors, the IoMT enables real-time monitoring and data sharing, transcending geographical barriers. This translates to more proactive and personalized care for patients, as healthcare providers can remotely track vital signs, medication adherence, and overall health trends. Empowering patients to be active participants in their health, IoMT fosters a sense of ownership and accountability. As we embrace this digital evolution, the healthcare system will become more patient-centric, accessible, and responsive than ever before.
Microsoft and Paige are building the world’s largest AI model for detecting cancer
CNBC’s Ashley Capoot reports that Microsoft announced Thursday it is teaming up with digital pathology provider Paige to build the world’s largest image-based artificial intelligence model for identifying cancer. The AI model is training on an unprecedented amount of data that includes billions of images, according to a release. It can identify both common cancers and rare cancers that are notoriously difficult to diagnose, and researchers hope it will eventually help doctors who are struggling to contend with staffing shortages and growing caseloads. On Thursday, Paige and Microsoft will publish a paper on the model through Cornell University’s preprint server, arXiv. The paper quantifies the impact of the new model compared with existing models.
“The more you go away from academic medical centers, especially in community clinics where pathologists are completely overwhelmed across all cancer types with so many cases, there, the impact is quite drastic. That really helps to democratize access to health care in these places.”
Thomas Fuchs, co-founder and chief scientist at Paige
Why it’s important – Despite pathologists’ essential role in medicine, Fuchs said their workflow has not changed much in the last 150 years. Paige’s original AI model used over 1 billion images from 500,000 pathology slides. The Microsoft/Paige model is training on 4 million slides to identify both common and rare cancers, which can be difficult to diagnose. Paige said it is the largest computer vision model ever publicly announced.
HCA, One Of The Largest Healthcare Organizations In The World, Is Deploying Generative AI
Healthcare Corporation of America (HCA), which operates nearly 180+ hospitals and is one of the largest healthcare entities in the world, announced that it will be deploying generative AI solutions to improve its care delivery models. To do so, the organization will be expanding its existing partnership with Google Cloud, which has been helping HCA advance its IT and data/analytics infrastructure. Forbes contributor Sai Balasubramanian, M.D., J.D. reports on the announcement. HCA developed a tool using one of Google Cloud’s LLMs, which helps automatically generate handoff reports. Prompts were developed to ensure that the LLM “[prioritizes] details, such as medication changes, laboratory results, vital sign fluctuations, patient concerns, and overall response to treatment.” Additionally, leadership worked with nursing teams to collect feedback and refine the product; after initial tests, nursing staff reportedly were “pleased with the speed, accuracy, and relevance of the draft reports” generated by the tool and “expressed high interest in putting the tool into practice.”
Why it’s important – Both parties also agree that the use cases for this technology are numerous: from optimizing the handoff process to automatically creating discharge summaries, improving the electronic health record experience, and driving better insight generation from large data sets, the applications for generative AI in healthcare are endless.
Health Technologies Driving Hospital-at-Home Programs
The hospital-at-home model is gaining momentum, supported by various types of health technologies, including remote patient monitoring, telehealth, and analytics. Anuja Vaidya, Senior Editor at xintelligent Media, covers the story. Healthcare is moving increasingly outside the walls of hospitals, spurred by the popularity of outpatient and virtual care modalities. During the COVID-19 pandemic, healthcare provider organizations stood up or scaled various care delivery options to extend care access, including hospital-at-home programs. Though some organizations implemented the model in the following decade or so, adoption received a significant boost in November 2020. To bolster care access amid in-person care restrictions and lockdowns, the Centers for Medicare & Medicaid Services (CMS) launched the Acute Hospital Care at Home initiative. The article does an excellent job covering the technologies supporting the evolving hospital-at-home model. The technologies covered in the article include:
- Remote patient monitoring (RPM) underpins most hospital-at-home programs. Hospital-at-home programs employ a wide array of RPM tools. These can include wearable devices such as blood pressure cuffs, pulse oximeters, and biosensors.
- Telehealth is a critical component of at-home hospitals, providing a direct connection between clinicians and patients. Most at-home hospital programs employ a hybrid telehealth and in-person care model. The telehealth aspect of this model allows clinicians to observe patients remotely and engage with them regarding the treatment plan and potential changes.
- Hospital-at-home programs require data analytics to be successful. The large amounts of structured and unstructured data generated from the RPM tools and telehealth solutions must be analyzed to allow clinicians to track patients’ progress and make clinical decisions. A common type of data analytics used in hospital-at-home programs is clinical risk prediction, according to the 2023 npj Digital Medicine article. Clinical risk prediction involves an analysis of retrospective observational data and using statistical methods to predict a patient’s likelihood of a certain clinical outcome.
- Efficient inventory management is critical for hospital-at-home program operations. As described above, hospital-at-home programs utilize various types of technology, making inventory management complex. According to a whitepaper by healthcare consultancy Chartis, hospital-at-home programs need to transport the technology and equipment necessary for treatment to patient homes, set up and test the devices and tools, replenish any equipment that can no longer be used, deploy diagnostic testing resources and coordinate reverse logistics for specimens, troubleshoot technical issues with patients, and collect the technology and equipment and prepare it for re-use.
Why it’s important – As these technologies continue to improve, their use in hospital-at-home programs will increase utilization, provide a more comprehensive data-driven approach to care delivery and make it easier for patients and their families to function as active participants in their care, leading to better outcomes and a better quality of life for all.
Scientists Just Tried Growing Human Kidneys in Pigs
In a first, researchers in China have used pigs to grow early-stage kidneys made up of mostly human cells. The advance is a step closer to producing organs in animals that could one day be transplanted into people. Wired’s Emily Mullen reports that in the current study, a team led by scientists at Guangzhou Institutes of Biomedicine and Health injected more than 1,800 pig embryos with human stem cells and then transferred them into the wombs of 13 female pigs. They allowed the chimeric embryos to grow for up to 28 days, then stopped the pregnancies to remove and examine the embryos. They collected five, which all had kidneys that were developing normally and contained up to 65 percent human cells. The research was published September 7 in the journal Cell Stem Cell. (The study authors didn’t respond to WIRED’s request for an interview.)
Why it’s important – More than 100,000 people in the United States are on the national transplant waiting list, and 17 people across the country die each day waiting for a donor organ, according to the Organ Procurement and Transplantation Network. Kidneys are the most in-demand, with nearly 89,000 Americans needing one as of September. Though these results are encouraging, even if scientists manage to grow full-fledged humanized organs inside pigs, there’s no guarantee they’d be compatible with the human immune system. And that’s the big question facing any technique that aims to generate transplant organs for patients: “Will an organ, regardless of how you make it, be accepted by the recipient?”
The Shrinking Number of Primary Care Physicians Is Reaching a Tipping Point
Finally, this week is this important article from Elisabeth Rosenthal, Senior Contributing Editor, KFF Health News. The percentage of U.S. doctors in adult primary care has been declining for years and is now about 25% — a tipping point beyond which many Americans won’t be able to find a family doctor at all. Already, more than 100 million Americans don’t have usual access to primary care, a number that has nearly doubled since 2014.
Why it’s important – The United States already ranks last among wealthy countries in certain health outcomes. The average life span in America is decreasing, even as it increases in many other countries. According to Rosenthal, Some relatively simple solutions are available if we care enough about supporting this foundational part of a good medical system. Hospitals and commercial groups could invest some of the money they earn by replacing hips and knees to support primary care staffing; giving these doctors more face time with their patients would be good for their customers’ health and loyalty if not (always) the bottom line. Reimbursement for primary care visits could be increased to reflect their value — perhaps by enacting a national primary care fee schedule, so these doctors won’t have to butt heads with insurers. Policymakers could consider forgiving the medical school debt of doctors who choose primary care as a profession. They deserve support that allows them to do what they were trained to do: diagnosing, treating, and getting to know their patients.