Thursday, September 11, 2025

 

Clocks created from random events can probe ‘quantumness’ of universe



A newly discovered set of mathematical equations describes how to turn any sequence of random events into a clock, scientists at King’s College London reveal




King's College London




A newly discovered set of mathematical equations describes how to turn any sequence of random events into a clock, scientists at King’s College London reveal. 

The researchers suggest that these formulae could help to understand how cells in our bodies measure time and to detect the effects of quantum mechanics in the wider world. 

Studying these timekeeping processes could have far-reaching implications, helping us to understand proteins with rhythmic movements which malfunction in motor neurone disease or chemical receptors that cells use to detect harmful toxins. 

Einstein famously said that “Time is whatever a clock measures” and while a wristwatch keeps time because it ticks at regular intervals, events which don’t follow a pre-determined pattern can also be used to measure time.  

For example, some processes consist of inherently random ‘jumps’ at irregular times. If each jump only depends on the previous jump, the process is called Markovian. Examples can be seen throughout nature, from fluctuating stock prices to the beating of a heart. 

By analysing these jumps, the scientists can estimate how much time has passed and place the strictest mathematical bound to date on how accurate that ‘clock’ is. 

If the clock behaves differently than the equations suggest, then it isn’t a classical Markovian process and there might be underlying quantum effects in the system. By the same token, clocks that use quantum physics are not restricted by the bound, which explains why quantum technologies such as atomic clocks can do better than any classical clock like those commonly used by the public. 

Dr Mark Mitchison, Proleptic Senior Lecturer in the Department of Physics at King’s and lead author explains, “Our goal was to find out the minimum ingredients you need to build a clock. For example, could you still measure time precisely even when stranded on a desert island? We found equations that tell you how to create a ‘clock’ by counting random events around you, like waves lapping on the shore or your heartbeats. 

“This turns out to be the best possible clock you can build by counting Markovian events in a system governed by classical physics. So, if you find a system that doesn’t follow the expected pattern, you can be sure something else is going on like underlying quantum behaviour.” 

The team also hope that these mathematical procedures can be used to study how biological systems operate efficiently in the presence of random fluctuations. For example, the motor protein kinesin transports other proteins within the cell, walking across small ‘microtubules’ which cross-cross the cell using two ‘feet’ to take directed steps along the tube.

These ‘molecular machines’ convert random thermal energy into a repeated, regular motion, like the ticking of a clock. They are also crucial for biological function: malfunctioning kinesin has been implicated in motor neurone disease.   

Dr Mitchison said, “Thinking about molecular machines as ‘clocks’ gives us insight into how some natural processes spontaneously generate order from chaos. We see this occurring at many different scales in our universe, from biological organisms and ecosystems down to the microscopic world. By establishing a fundamental limit on how well clocks can operate in the realm of classical physics, we also gain a better understanding of what makes quantum clocks different. 

“Time lies at the heart of many unsolved mysteries in quantum physics. Why does time seem to flow in only one direction? Why do we only remember the past and not the future? Is time quantised in discrete chunks, in the same way as energy? By thinking about what clocks can do, we ultimately hope to answer some of these questions about the nature of time itself.” 

Schaeffer Center white paper outlines FDA reforms to boost pharmaceutical innovation and expand access 



Harnessing modernized drug development tools could accelerate patient access to new treatments and cut costs 




University of Southern California





Rapid scientific advances are accelerating the development of medical innovations, from personalized treatments to curative gene therapies and advanced diagnostic tools. But significant policy and regulatory reforms at the Food and Drug Administration are needed to fully harness the game-changing potential of these technologies, according to a new white paper from the USC Schaeffer Center for Health Policy & Economics.

The white paper offers comprehensive recommendations for how FDA can modernize drug development by streamlining clinical trials, providing clearer guidance to drugmakers about newer technologies, and improving agency efficiency amid substantial staff reductions. Taken together, these steps could reduce the considerable time and cost of bringing new drugs to market, pushing down prices and improving patient access while maintaining rigorous safety and effectiveness standards.

“We are entering a transformative era in medical innovation that could improve the lives of millions of Americans. To fully realize this potential, FDA must provide clear, dependable guidance that helps drugmakers navigate the growing complexities around evaluating new technologies,” said Schaeffer Nonresident Senior Scholar Lowell Schiller, the lead author and former acting chief counsel and principal associate commissioner for policy at FDA.

Supporting newer clinical trial methods

Technology has made it easier to generate evidence for new drugs outside of traditional randomized controlled trials, which can be time-consuming, costly and are not always feasible, particularly for rare diseases. Although use of innovative approaches like real-world evidence and external control arms has become more common, they remain difficult to implement, and drugmakers often face uncertainty about whether they will pass muster with FDA.

That uncertainty could discourage drugmakers from pursuing riskier projects or inflate development costs.

The authors recommend that FDA increase early interactions with drug developers who are using innovative trial designs, systematically share insights and best practices, and formalize guidance documents to help developers better identify appropriate use of these methods.

“Before spending hundreds of millions of dollars on developing a new treatment, drugmakers want to know that FDA won’t eventually send them back to square one,” Schiller said. “Proactive guidance on emerging technologies and challenging areas of clinical development will help build certainty, driving investment in innovation.”

Better leveraging data on how drugs perform in real-world settings could also improve FDA’s accelerated approval program, which has provided expedited access to hundreds of treatments for serious diseases based on early evidence of effectiveness. However, required follow-up studies to confirm treatment benefit are notoriously challenging. As part of a broader set of recommended reforms for the accelerated approval program, FDA should make clear how drugmakers can incorporate real-world evidence in these studies.

Investing in AI use and oversight
Strategic expansion of advanced computing technologies like artificial intelligence and machine learning at FDA can make new drug reviews better and faster.

FDA’s recent rollout of a generative AI tool to assist with drug reviews is an important step in this direction, but the agency’s implementation of this and other advanced tools is still in early stages. Going forward, FDA can maximize the potential of such tools by establishing strong principles and guardrails. For instance, FDA should optimize its tools for high-value use cases and require that any “algorithmic-informed” decisions are transparent and ultimately made by human reviewers who understand the limitations and risks of AI systems.

Deployed responsibly, these tools can do more than just summarize documents. Analytical platforms could help agency reviewers more effectively identify data quality issues or conduct sophisticated analyses within the increasingly large data sets that FDA receives, allowing them more time to ensure agency guidance and precedent are applied consistently.

FDA should also clearly define goals and use cases for integrating AI and similar tools into drug reviews, helping drugmakers and the public understand how the technology is being used responsibly.

Focusing on what matters most to patients
Regulators and patient communities have at times clashed over how some treatments are evaluated and approved. This tension is most common with rare diseases, which affect 30 million Americans and often have no FDA-approved treatments.

Better incorporating patient perspective into how trials for drugs and biologics are designed and evaluated could build patient trust and encourage more efficient trials focused on outcomes that matter most to patients. Particularly with rare diseases, this would better account for patients’ willingness to accept less certain benefits and risks in exchange for faster access to new therapies.

While FDA has been more flexible in evaluating rare disease treatments, it should work with Congress to formalize its approach to ensure consistency, including in how the agency accounts for patient preferences. Data on patient preferences should also inform Medicare coverage of new treatments and may be especially helpful for drugs approved through the accelerated pathway.

About this white paper
The white paper’s recommendations are organized into six sections: (1) Modernize Evidence Generation, (2) Advance Innovation for Rare Disease, (3) Enhance Supply Chain Oversight, (4) Strengthen the Accelerated Approval Pathway, (5) Invest in FDA’s Use and Oversight of Artificial Intelligence and Other Advanced Computing Technologies and (6) Advance Drug Competition.

Eunjoo Huisung Pacifici, associate professor at the USC Mann School of Pharmacy and Pharmaceutical Sciences, and Barry Liden, director of public policy at the USC Schaeffer Center, are also co-authors. This white paper was supported by the Schaeffer Center. A complete list of supporters of the Schaeffer Center can be found in our annual report, available here

 D.E.I. IS MERIT

Scientist, advocate and entrepreneur Lucy Shapiro to receive Lasker-Koshland special achievement award



Stanford Medicine





In 1959, Lucy Shapiro, a freshman honors student at Brooklyn College majoring in arts and literature, had signed up for a course in inorganic chemistry on a lark. It didn’t interest her, and she stopped attending. But at the end of the term, having failed to officially drop the class, she was notified she had to take the final exam.

“It was a multiple-choice test, so I just circled all the B's,” recalled Shapiro, PhD, professor emerita of developmental biology and director of Stanford Medicine’s Beckman Center recalled. She received a D.

Three years later, the budding artist met physical chemist and violinist Theodore Shedlovsky, PhD, at an art exhibition that included one of Shapiro’s paintings. “He had this thing of finding young people in the arts and, if he thought they were smart and creative, urging them to go into science,” Shapiro said. She fit the bill.

Shedlovsky convinced Shapiro to take a course in organic chemistry. But the only available class at the college was taught at an honors level.

“I went to the chemistry department office for permission to take the course,” Shapiro said. “The assistant to the chair asked about my chemistry course record, and I had to explain that I had none, and that furthermore the D in inorganic chemistry on my record was the result of circling all the B's.” As the assistant was about to throw Shapiro out, the department chair came out of his adjoining office. He stared silently at her awhile, then asked why she wanted to take his honors course.

“I assured him that I was quite smart,” Shapiro said. “He saw this as challenge, and I was in! At that time, girls studying chemistry were rare.”

She aced the class.

“That course changed the trajectory of my life,” Shapiro said.

An influential career

It’s not an exaggeration to say that Shapiro’s extraordinary powers of persuasion also altered the field of biology. Shapiro has served as the first female department chair at three universities; launched Stanford Medicine’s department of  developmental biology; with her physicist husband Harley McAdams, PhD, a professor of developmental biology, emeritus, created the field of systems biology; served as an advisor on bioterrorism, antibiotic resistance and pandemic preparedness to two U.S. presidential administrations; addressed the U.S. Senate’s Armed Services Committee and the world’s leading economists on global health threats; founded companies that developed novel antibiotic and antifungal drugs; and been awarded a bevy of national and international awards including the National Medal of Science and Canada’s Gairdner International Award.

Next week, Shapiro will be awarded the Lasker-Koshland Award for Special Achievement in Medical Science. The Lasker Award, known as the “American Nobel,” is this country’s most distinguished honor for researchers in basic and clinical medical sciences. The Special Achievement award, which has been renamed in honor of the late biochemist Daniel Koshland Jr., is given only once every two years to commemorate a life of scientific contribution and service.

Shapiro will be the 18th recipient of the award and the third woman. In 2008, the late Stanley Falkow, PhD, a professor of microbiology and immunology at Stanford Medicine, received the same award for his work on the molecular basis of microbial pathogenesis. The 2025 awards were announced Sept. 11 by the Lasker Foundation and are to be presented at a ceremony Sept. 19 in New York City.

“Lucy has been described as a force of nature,” said Lloyd Minor, MD, the Carl and Elizabeth Naumann Professor for the Dean of the School of Medicine and vice president for medical affairs at Stanford University. “I wholeheartedly agree. In her 36 years at Stanford Medicine, she transformed the way biologists think of bacteria and developmental biology; she pioneered a truly interdisciplinary lab that brought together physicists, chemists and biologists; met with multiple heads of state; launched two extraordinarily successful biotech companies; and mentored dozens of students who now lead successful research projects around the world. And she does it with a smile on her face and an engagement with her colleagues that makes everyone feel valued. We are incredibly fortunate to have her as a member of the Stanford Medicine community.”

Shapiro is stunned by the recognition. “I don’t even know how to answer that,” she responded when asked how she felt when she first learned she would receive the award. “It was just so shocking. Many of the previous recipients are my heroes — people who I’ve looked up to my entire scientific career. It means more than I can possibly say to be part of that cadre. It is really the highest honor I can imagine.”

Since her graduate school days, Shapiro has focused on understanding how bacteria compartmentalize proteins and structures that generally lack the intracellular membranes and organelles that more advanced cells (from yeast to human) use for organization. Discoveries made in her lab in the late 1990s overturned the prevailing idea that bacteria were just a “bag of enzymes” by proving that proteins involved in the regulation of the cell cycle travel to the poles of the cell in a dynamic, regulated fashion prior to cell division. 

She did so by studying a unique bacterium called Caulobacter crescentus that divides asymmetrically to create one stationary, stalked cell and one mobile, flagellated cell that swims away to new locations. This type of cell division is the fundamental basis of stem cell function and the generation of diversity in the living world.

Along the way, Shapiro sought out and cultivated interdisciplinary collaborators at Stanford Medicine and beyond, including McAdams, a former department head of systems engineering at Bell Labs who moved with her to Stanford in 1989 when Shapiro was recruited to be the founding chair of the new Department of Developmental Biology. Together, first in a spare bedroom in their house on the Stanford University campus and then in side-by-side laboratories when McAdams became a faculty member, biologists, geneticists, physicists and electrical engineers rubbed shoulders and drew parallels between genetic feedback networks and electrical circuits. “This revolutionized our understanding of the complex genetic and molecular interactions that govern how cells grow and divide,” Shapiro said.

“Lucy’s emotional intensity and focus, along with her willingness and desire to solve any problem, is inspirational” said W.E. Moerner, PhD, the Harry S. Mosher Professor, a professor of chemistry at Stanford University and the 2014 recipient of the Nobel Prize in chemistry for the development of a type of microscopy that enabled the detection of single molecules within a living cell. Shapiro and Moerner used the technique to track the movement of proteins associated with the cell cycle in Caulobacter — cementing the idea that bacteria actively regulate the location of proteins and DNA during cell division.

“People didn’t believe it at first,” Shapiro said. “I had to fight to get those papers published in Science.”

Never one to shrink from unconventional ideas, Shapiro leveraged what she’d learned about bacterial biology in a collaboration with  Steve Benkovic, PhD, a chemist at Pennsylvania State University, to explore whether replacing a carbon molecule in the active site of potential antibacterial and antifungal drugs with a boron molecule could lessen toxicity while maintaining efficacy.

“The boron-based molecules killed the pathogens, while sparing human cells,” Shapiro said.

She and Benkovic obtained the license to this new class of anti-infectives and launched a company called Anacor Inc. Anacor developed the first new antifungal in 50 years and an anti-inflammatory drug effective against eczema — both of which were approved by the Food and Drug Administration. In 2016, Anacor was sold to Pfizer for $5.2 billion. Shapiro co-founded Boragen, which subsequently merged with AgriMetis to form the agricultural company 5Metis that is designing boron-containing antifungals to protect bananas and rice crops worldwide. She has served on the boards of multiple biotech companies.

Lunching with a purpose

For decades, Shapiro participated in monthly lunches hosted by former U.S. Secretary of State George Shultz. During these meetings, Shultz would ask Shapiro, then-Stanford-provost Condoleezza Rice and nuclear nonproliferation expert Sid Drell what they had accomplished (or failed to accomplish) during the previous month, as well as what concerned them the most.

“Being forced to articulate, and be aware of, existential threats in my area of expertise expanded my concerns from how a cell goes about its business of being a living entity to ones of national policy,” Shapiro said. “I developed a new vision of an integrated world system — or circuit, if you will — in which the components were the economy, political national interests and technological developments that were rapidly changing our understanding of global dynamics. Underlying these issues was the specter of climate change that was clearly the elephant in the room.”

The lunches, which began in the mid-1990s, caused Shapiro to contemplate how to respond to challenges facing human health. As a result, she embarked on a three-pronged mission: speak publicly about the rising threats to global health, gain the political visibility and influence to champion the support of basic research, and design and obtain regulatory approval for new anti-infective drugs.

In true Shapiro fashion, she was remarkably successful. Over the years, she spoke with Soviet Union leader Mikhail Gorbachev about emerging infectious disease threats; met with then National Security Advisor Condoleezza Rice during the George W. Bush administration; briefed President Bill Clinton about how the threat of bioterrorism is similar to the threat of the natural emergence of dangerous pathogens; and, in 2019, addressed the Senate Armed Services Committee about the looming threat of a global pandemic. She met with Mark Zuckerberg and his wife, Priscilla Chan, to encourage the collaboration between Stanford University and the University of California, San Francisco, that would become the Chan Zuckerberg Initiative, and she counseled former California governor Jerry Brown about antibiotic resistance and SARS-CoV-2 and Microsoft founder Bill Gates about boron-containing anti-infectives.

“If Lucy sees something that is not right, she is not going to sit around and do nothing,” said Peter Kim, PhD, the Virginia and D.K. Ludwig Professor in Biochemistry and former president of Merck Research Laboratories. “Her role as a government adviser on bioterrorism and antimicrobial resistance has been remarkable. And the overall breadth of her accomplishments, along with her ability to influence so many different parts of society in such positive ways, is astounding.”

“She’s a true citizen scientist who feels responsible for the broader community,” said scientific colleague Jeremy Nathans, MD, PhD, professor of molecular biology and genetics at the Johns Hopkins School of Medicine. “Not just for her own lab, or department, or university but much more broadly. And she understands the consequences of large-scale public misunderstanding of science and how incredibly damaging that is.”

Early days

Shapiro graduated from Brooklyn College in 1962 with a bachelor’s in fine arts and a minor in biology. She launched into graduate school under the auspices of Shedlovsky and biochemist Jerry Hurwitz, PhD, first at New York University where she became part of the “jungle biochemists,” then at Albert Einstein College of Medicine. “They were rough, competitive, demanding and exquisitely able chemists. And now I was one of them,” Shapiro recalled in an unpublished reflection.

As a graduate student, Shapiro was challenged to determine how a bacterial virus that used RNA as its sole genetic material was able to hijack an infected bacterial cell and make numerous copies of itself. Her PhD thesis was the discovery of an RNA-dependent RNA polymerase encoded by a virus that infects the bacterium E. coli. Learning that RNA could be stable enough to serve as a primary genetic code, rather than just a messenger to carry instructions from DNA to the cell’s protein-making machinery, was a key step in the development of today’s mRNA vaccines.

“I had no way of knowing then that the test tubes and cold rooms with which I quickly became familiar would yield information that would be absolutely critical to battle a pandemic nearly six decades later,” Shapiro said.

Within 10 years of graduating, Shapiro rose to lead the department of molecular biology at Albert Einstein — the first woman to hold a department chair at the school. 

“Lucy burst like a bolt of lightning upon the scientific scene,” said colleague Carol Gross, PhD, professor of cell and tissue biology at UCSF. “Although her PhD thesis was purely biochemistry and almost all her coursework was in chemistry or physical chemistry, when faced with starting her own laboratory, Lucy determined that the most significant question she could address was in developmental biology.” 

Caulobacter crescentus was an obscure, understudied bacterium in a time when the vast majority of biologists were using E. coli — a common gut bacterium that divides symmetrically — as a model system. But Caulobacter was ideal for the questions Shapiro wanted to address — namely, how a linear DNA genome gives rise to a three-dimensional cell with specific architecture and function. The topic was incredibly daunting, but Shapiro had confidence in her “hydrogen atom” (the simplest possible unit of study that, when understood, could shed light on much more complex systems).

“At scientific meetings, where every session focused on E. coli, Lucy and I took to calling ourselves the ‘un-colis’ as a nod to the popular ad for 7 Up as the ‘un-cola,'” said Richard Losick, PhD, a professor of biology at Harvard University and a Howard Hughes Medical Institute investigator. Losick was also studying a little-known bacteria called Bacillus subtilis.

“For a long time our work was largely ignored. But her remarkable charisma convinced the planet to pay attention to her as she developed Caulobacter into an excellent model for research. And I came along for the ride,” Losick said. “It seems like she can do anything she sets her mind to. It’s amazing. This award is the crowning recognition for Lucy’s career. It’s a category that recognizes broad contributions to science and the world. It is a perfect fit for her.”

After 19 years at Albert Einstein, Shapiro moved to Columbia University as professor and chair of microbiology and immunology. She lasted three years before yielding to Stanford Medicine’s repeated offers to chair its new Department of Developmental Biology. The opportunity to recruit the best of the best to the new department was too appealing to turn down. At Stanford Medicine, Shapiro set about assembling a cadre of researchers studying similar questions in model organisms from bacteria to mammals.

A woman in science

Once again the first woman to chair a department at a school of medicine, Shapiro was used to the experience. Jarring moments during her early career — being heckled during scientific presentations and ostracized in classes — had made it clear that women face particular challenges when entering male-dominated fields.

“More than any other woman of her time, Lucy showed that women could compete in all aspects of the scientific enterprise at the level of their male counterparts, even without a male scientist partner to pave the way for her,” Gross said. “At a time when very few women were scientists, much less in leadership positions, Lucy’s brilliance, foresight and determination showed what is possible, paving the way for others to follow.”

Her determination included helping young scientists facing similar hurdles.

“Lucy is not just a brilliant scientist, but also a very caring mentor,” said Christine Jacobs-Wagner, PhD, a professor of biology and of microbiology and immunology as well as a former postdoctoral student in Shapiro’s lab. “When I first arrived, I had won an award and needed to attend a very fancy reception in Sweden where you had to have a cocktail dress. And Lucy, who is very stylish, just looked at me and said, ‘Let’s go to Stanford Mall together on Saturday and we’ll pick up a dress.’ And this is what we did. Once you are part of her lab, you are part of her family.”

Many of Shapiro’s former protégés, including Jacobs-Wagner, have launched highly successful laboratories of their own.

“She is the kind of mentor who helps you feel independent and empowered, while also giving you the structure to really thrive in such a seamless way that you think you’re coming up with all these great ideas yourself,” Jacobs-Wagner said. “She knows what you need before you know you need it.”

“The important thing about winning this award is that people realize that buried within a career is a full life, which includes a family,” said Shapiro, who raised three children with McAdams. “That isn’t often mentioned. We talk all about that company and this science, but raising my children and my grandchildren has been such an important part of my life. The same was true for Marie Curie and her daughters. We are all human. That tends to get lost when you start talking about the ‘American Nobel.’ It shouldn’t be lost.”

Now, 65 years after Shapiro talked her way into the course that would change her life, she has a message for the young scientists following in her footsteps.

“If you are confident in what you are talking about, and your science is excellent, there is no need to be intimidated by anyone.”

 

Mizzou economists: 2025 farm income boosted by high cattle prices and one-time payments



The University of Missouri Food and Agricultural Policy Research Institute (FAPRI) finds signs of strain in crop markets despite a strong year for livestock producers.




University of Missouri-Columbia

FAPRI 

image: 

FAPRI

view more 

Credit: University of Missouri





Net farm income in the United States is projected to reach $177 billion in 2025, a sharp increase from $128 billion in 2024. This is according to the latest update of the annual U.S. farm income and consumer food price report by the Food and Agricultural Policy Research Institute (FAPRI) at the University of Missouri’s College of Agriculture, Food and Natural Resources.

Record cattle prices and large one-time government payments have boosted 2025 income, but declines in crop prices and projected reductions in future government support raise concerns about the outlook for 2026.

“Despite strong income this year, much of the gain is temporary,” Pat Westhoff, director of FAPRI, said. “As emergency payments dry up and crop prices remain weak, we project a $31 billion decline in farm income next year.”

The report incorporates data available in August 2025, including United States Department of Agriculture (USDA) crop production estimates and economic forecasts from S&P Global. It also accounts for modifications in key farm programs and tax credits related to biofuel production that were included in the One Big Beautiful Bill Act signed into law in July.

Key findings from the update include:

  • Corn prices fall significantly due to record production, with the 2025-26 marketing year (Sept. 1 - Aug. 31) average price projected at $4.05 per bushel, slightly above USDA’s latest estimate. Modest price recovery is expected in 2026-27.
  • Soybean prices increase slightly to $10.16 per bushel in 2025-26 as reduced acreage and strong biofuel demand tighten supplies. Continued demand from the renewable fuels sector could drive further gains in 2026-27.
  • Other crop prices remain weak, with large global supplies pressuring wheat, rice, sorghum and barley. Cotton is a notable exception, with a smaller crop supporting prices at 66.5 cents per pound.
  • Cattle prices hit new records. Tight supplies and strong domestic demand push prices even higher in 2026 before increased production brings moderation.
  • Dairy production increases as both cow numbers and yields rebound. However, added supply has weighed on prices, especially for cheese and butter. Exports are expected to help offset the pressure.
  • Food price inflation rebounds to 2.9% in 2025, driven largely by beef prices, which are projected to rise by more than 10% for the year. Food-at-home inflation is expected to moderate in 2026, but costs at restaurants continue to rise.

Westhoff emphasizes that the projections reflect a snapshot in time and are subject to change as new information becomes available.

“These forecasts are conditional on current policies and market expectations,” Westhoff said. “They provide a useful benchmark for evaluating potential impacts of economic shifts, weather events and future policy changes.”

The update is part of FAPRI’s ongoing efforts to provide policymakers, industry stakeholders and the public with reliable economic analysis of the U.S. agricultural sector.

 

Additional insights from the farm income report

 

 

About FAPRI

FAPRI, a program of distinction in the College of Agriculture, Food and Natural Resources (CAFNR), develops and publishes baseline reports to highlight the impact of current events on agricultural market trends and projections.


AI tools fall short in predicting suicide, study finds



ALGORITHIMS DON'T DIE

Analysis of 53 studies using machine learning to predict suicide and self-harm finds low accuracy




PLOS





The accuracy of machine learning algorithms for predicting suicidal behavior is too low to be useful for screening or for prioritizing high-risk individuals for interventions, according to a new study published September 11th in the open-access journal PLOS Medicine by Matthew Spittal of the University of Melbourne, Australia, and colleagues.

Numerous risk assessment scales have been developed over the past 50 years to identify patients at high risk of suicide or self-harm. In general, these scales have had poor predictive accuracy, but the availability of modern machine learning methods combined with electronic health record data has re-focused attention on developing new algorithms to predict suicide and self-harm.

In the new study, researchers undertook a systemic review and meta-analysis of 53 previous studies that used machine learning algorithms to predict suicide, self-harm and a combined suicide/self-harm outcome. In all, the studies involved more than 35 million medical records and nearly 250,000 cases of suicide or hospital-treated self-harm.

The team found that the algorithms had modest sensitivity and high specificity, or high percentages of people identified as low-risk who did not go on to self-harm or die by suicide. While the algorithms excel at identifying people who will not re-present for self-harm or die by suicide, they are generally poor at identifying those who will. Specifically, the researchers found that these algorithms wrongly classified as low risk more than half of those who subsequently presented to health services for self-harm or died by suicide. Among those classified as high-risk, only 6% subsequently died by suicide and less than 20% re-presented to health services for self-harm.

“We found that the predictive properties of these machine learning algorithms were poor and no better than traditional risk assessment scales,” the authors say. “The overall quality of the research in this area was poor, with most studies at either high or unclear risk of bias. There is insufficient evidence to warrant changing recommendations in current clinical practice guidelines.”

The authors add, “There is burgeoning interest in the ability of artificial intelligence and machine learning to accurately identify patients at high-risk of suicide and self-harm. Our research shows that the algorithms that have been developed poorly forecast who will die by suicide or re-present to health services for the treatment of self-harm and they have substantial false positive rates.”

The authors note, “Many clinical practice guidelines around the world strongly discourage the use of risk assessment for suicide and self-harm as the basis on which to allocate effective after-care interventions. Our study shows that machine learning algorithms do no better at predicting future suicidal behaviour than the traditional risk assessment tools that these guidelines were based on. We see no evidence to warrant changing these guidelines.”

 

Citation: Spittal MJ, Guo XA, Kang L, Kirtley OJ, Clapperton A, Hawton K, et al. (2025) Machine learning algorithms and their predictive accuracy for suicide and self-harm: Systematic review and meta-analysis. PLoS Med 22(9): e1004581. https://doi.org/10.1371/journal.pmed.1004581

Author countries: Australia, Belgium, United Kingdom

Funding: see manuscript