Support Pullquote, upgrade to Pro!
(Or just tweet your Pullquote for free!)
With Pullquote Pro, you'll get to:
- share on Facebook
- schedule tweets
- tweet from multiple accounts
- edit quotes
- customize colors
- change fonts
- save and index quotes
- private quotes
Choose a plan: $5/month $50/year (includes free access to any new features)
Recent quotes:
ChatGPT Defeated Doctors at Diagnosing Illness - The New York Times
The chatbot, from the company OpenAI, scored an average of 90 percent when diagnosing a medical condition from a case report and explaining its reasoning. Doctors randomly assigned to use the chatbot got an average score of 76 percent. Those randomly assigned not to use it had an average score of 74 percent.
The study showed more than just the chatbot’s superior performance.
It unveiled doctors’ sometimes unwavering belief in a diagnosis they made, even when a chatbot potentially suggests a better one.
Fragmented physical activity linked to greater mortality risk -- ScienceDaily
The researchers found that for each 10 percent higher activity fragmentation there was a 49 percent increase in the risk of mortality. The researchers defined activity fragmentation as the probability of transitioning from an active state to a sedentary state for each participant, so shorter average activity periods meant higher fragmentation.
The researchers also analyzed the duration of each participant's bouts of activity, and found that "percent of activity spent in bouts of less than five minutes" appeared to be another good marker of mortality risk. Each additional 10 percent of active time spent in such short bouts was associated with a 28 percent increase in the chance of mortality. Percent of active time spent in 5- to 10-minute bouts was not a significant indicator of mortality risk.
AI system accurately detects key findings in chest X-rays of pneumonia patients within 10 seconds -- ScienceDaily
Researchers found that the CheXpert model outperformed the current system of using a radiologist to create radiology reports for all key pneumonia findings, plus NLP. It also did so in less than 10 seconds, compared to the 20 minutes to hours from NLP. NLP of radiology reports was the most frequent cause of errors within ePNa.
"A 2013 study published in JAMA Internal Medicine found that 59 percent of errors made by ePNA were due to NLP processing of radiologist reports, so we're eager to replace it with a better, faster system," Dr. Dean said.
Outside of ePNa concerns, emergency department physicians looking at radiology reports often are challenged to understand the unstructured language used by radiologists in interpreting shadows on chest X-rays, Dr. Dead added.
Fertility app 'Dot' found to be as effective as other family planning methods -- ScienceDaily
The researchers found that the app had a typical-use failure rate of 5 percent and a perfect-use failure rate of 1 percent, which makes Dot comparable to family planning methods such as the pill, vaginal ring, and other fertility awareness-based methods.
"More and more women are using apps as a family planning method, so having an option backed by strong evidence of effectiveness is critical," says Victoria Jennings, PhD, principal investigator of the Dot effectiveness study and director of the IRH. "Women must be able to base their app choice on solid evidence about how well the method works and what's involved in using it. That's why it was so important that an app like Dot undergo a rigorous effectiveness trial conducted according to established study guidelines used to study other methods."
Next for Apple Watch: a clinical trial with J&J to track heart health - STAT
Typically, clinical trials use their own research infrastructure to track how patients do, but the HEARTLINE trial will plug into insurance claims databases to track patients. This approach, called a pragmatic clinical trial, could be cheaper and more efficient than the way studies are conducted now, without sacrificing the clarity and certainty that comes from having a control group in a study. But this is also risky: Using insurance claims data this way is new and untested.
“It’s certainly a vast and gross departure form the bricks and mortar model,” said Gibson. “This is one of the most exciting things, the idea that you’re going to find participants through the media, the news, potentially through insurers and even health care providers.”
Sweat holds most promise for noninvasive testing -- ScienceDaily
Last year the lab created the world's first continuous-monitoring sensor that can record the same health information in sweat that doctors for generations have examined in blood. The milestone is remarkable because the continuous sensor allows doctors to track health over time to see whether a patient is getting better or worse. And they can do so in a noninvasive way with a tiny patch applied to the skin that stimulates sweat for up to 24 hours at a time.
"This is the Holy Grail. For the first time, we can show here's the blood data; here's the sweat data -- and they work beautifully together," Heikenfeld said.
Heikenfeld and his students published their latest experimental findings in December in the journal Lab on a Chip. UC's study tracked how test subjects metabolized ethanol. The study concluded that sweat provided virtually the same information as blood to measure a drug's presence in the body.
Blood test to diagnose heart attacks is flawed, warn researchers: One in 20 patients had test levels higher than recommended limit -- study results could help to avoid misdiagnosis and inappropriate treatment -- ScienceDaily
This recommended level is used as the upper limit of normal (ULN). In other words, if the value of troponin is above the 99th percentile, that is considered to be abnormal, and would indicate a heart attack in appropriate clinical circumstances.
But little is known about the true distribution of the troponin level across a whole hospital population that includes inpatients, outpatients, patients undergoing surgery, in intensive care etc.
So researchers measured levels of high sensitivity cardiac troponin I (hs-cTnI) in 20,000 inpatients and outpatients undergoing blood tests for any reason at University Hospital Southampton between 29 June and 24 August 2017.
The average age of participants was 61 and 53% (10,580) were women.
The researchers found that the 99th centile of troponin for the whole study population was 296 ng/L compared with the manufacturer's recommended level of 40 ng/L.
One in 20 (1,080; 5.4%) of all 20,000 patients had a troponin level greater than 40 ng/L, but in most of these patients there was no clinical suspicion of a heart attack.
Overall, 39% of all patients from the critical care units, 14% of all medical inpatients, and 6% of all patients from the emergency department had a troponin concentration greater than the recommended ULN.
Applying a network perspective to human physiology: Physicist describes 'network physiology,' which looks at different organ systems and how they relate to each other -- ScienceDaily
"We need to show how the different systems communicate with each other and adjust, coordinate and stay in sync," said Ivanov. The human body, according to this view, can be thought of as a network, with each organ serving as a node connected to other organs and other nodes. "The nodes are not just dots," he said. "They're dynamical systems, constantly changing in time, as are the connections between them."
Today's best hospitals aren't equipped to monitor the inter-organ interactions. "Separate devices keep track of separate functions, but no single monitor can observe a multitude of functions," he said. To improve health monitoring techniques, Ivanov and his colleagues have spent the past decade developing the computational tools and biomedical devices needed to capture data streams from different organ systems and see how they relate to each other.
Genomics could better match treatments to pancreatic cancer patients -- ScienceDaily
"Every pancreatic cancer is different, and performing molecular profiling of each patient's tumor could help determine the best treatment options," said lead author Aatur Singhi, M.D., Ph.D., surgical pathologist at UPMC and assistant professor of pathology at Pitt. "Rather than blindly giving patients the same chemotherapy, we want to tailor a patient's chemo to their tumor type. A one-size-fits-all approach isn't going to work. Therefore, we would like to make molecular profiling standard-of-care for patients with pancreatic cancer."
This hyperlocal news site in San Francisco is reinventing itself with an automated local news wire » Nieman Journalism Lab
“There are so many stories news organizations could potentially do, that nobody can afford to, because it’s expensive and time-consuming to have that many people on the ground,” Eldon sad. “We’re starting with the simple stuff right now: New business openings, rental price trends — simple story types that we can produce using data sets that cover a lot of geographical places and then distribute to a lot of people. Over time, we’ll want to get more sophisticated with how we analyze the data we have.”
Comparative Accuracy of Diagnosis by Collective Intelligence of Multiple Physicians vs Individual Physicians. | Medical Education and Training | JAMA Network Open | JAMA Network
Using the top 3 diagnoses given by each user to adjudicate solutions, the diagnostic accuracy of all users was 62.5% (95% CI, 60.1%-64.9%). The accuracy of individual residents and fellows was 65.5% (95% CI, 63.1%-67.8%) compared with 55.8% for medical students (95% CI, 53.4%-58.3%; P < .001 for difference vs residents and fellows by z test for proportions) and 63.9% for attending physicians (95% CI, 61.6%-66.3%; P = .10 for difference vs residents and fellows).
The frequency of diagnostic errors in outpatient care: estimations from three large observational studies involving US adult populations | BMJ Quality & Safety
Data sources included two previous studies that used electronic triggers, or algorithms, to detect unusual patterns of return visits (primary care study) or lack of follow-up of abnormal clinical findings for colorectal cancer (CRC) (colon cancer study), both suggestive of diagnostic errors.4 ,6 A third study examined consecutive cases of lung cancer in two institutions (lung cancer study).5 In all three studies, diagnostic errors were confirmed through chart review and defined as missed opportunities to make a timely or correct diagnosis based on available evidence. The criteria for diagnostic errors were comparable across the three studies and excluded atypical presentations and appropriate decisions to watch and wait.
Psychiatry: Case notes indicate impending seclusion -- ScienceDaily
The retrospective examination encompassed the records of 26 adult patients who ultimately required seclusion. A patient group of equal size with identical diagnoses, for whom coercive measures were not required, was used as a comparison. The basis for the qualitative and quantitative text analysis was the notes from the three days preceding an escalation.
The results showed that the case notes before seclusion were more extensive, meaning they contained a significantly higher number of words. "It seems that the attending staff describe problematic behavior more comprehensively in order to improve information transfer between different shifts, justify upcoming coercive measures and ensure that they are legally protected," says Dr. Clara Stepanow, author of the study.
Health data tools to rapidly detect sepsis in newborns: Study uses automated models to identify sepsis before clinical recognition -- ScienceDaily
"Because early detection and rapid intervention is essential in cases of sepsis, machine-learning tools like this offer the potential to improve clinical outcomes in these infants," said first author Aaron J. Masino, PhD, who led the study team's machine-learning efforts. Masino is an assistant professor in the Department of Anesthesiology and Critical Care Medicine and a member of the Department of Biomedical and Health Informatics at Children's Hospital of Philadelphia (CHOP). "Follow-up clinical studies will allow researchers to evaluate how well such systems perform in a hospital setting."
Infectious diseases could be diagnosed with smartphones in sub-Saharan Africa -- ScienceDaily
Still, the report's authors remain optimistic. As of 2016, global smartphone adoption has reached 51 per cent and is predicted to keep growing -- particularly in resource limited settings such as sub-Saharan Africa.
This means more and more of the world's population is equipped with a powerful pocket computer that can connect patients and share healthcare data.
Professor Stevens said: "This is an exciting opportunity for researchers and policy makers to develop new tools and systems that could drastically improve human health and wellbeing in the future."
THC found more important for therapeutic effects in cannabis than originally thought: Researchers measure product characteristics and associated effects with mobile app -- ScienceDaily
Since its release in 2016, the commercially developed ReleafApp has been the only publicly available, incentive-free app for educating patients on how their type of product (e.g., flower or concentrate), combustion method, cannabis subspecies (indica, sativa, and hybrid), and major cannabinoid contents (THC and CBD) affect their symptom severity levels, essentially providing invaluable user feedback on their health status, medication choices, and the clinical outcomes of those choices as measured by symptom relief and side effects.
The study aimed to address the practical questions of knowing how fundamental characteristics of currently available and frequently used cannabis products, characteristics that often influence consumer choices, affect health symptom intensity levels. The average patient, across the roughly 20,000 measured user sessions and 27 measured symptom categories ranging from depression to seizure activity, showed an immediate symptom improvement of 3.5 points on a 0-10 scale. Dried flower was the most commonly used product and generally associated with greater symptom improvement than other types of products.
Machine learning could eliminate unnecessary treatments for children with arthritis: An algorithm predicted disease outcome in children suffering from arthritis, helping doctors better tailor treatment -- ScienceDaily
"We had to use machine learning just to detect these seven patterns of disease in the first place," says Morris, whose team modified the technique known as multilayer non-negative matrix factorization. "And then we realized there are some children who do not fall into any of the patterns and they have a very bad version of the disease. Now we understand the disease much better we can group children into these different categories to predict response to treatment, how fast do they go into remission and whether or not we can tell they are in remission and remove therapy."
