Forge Friday Roundup - September 10, 2021

September 10, 2021

In today’s Roundup: sample sizes for external validation of models; cohort study affirms vax safety in pregnancy; diving deep for antimicrobials; risk factors: more than just genetics; pivotal role of women in vaccine development; investigation reveals privacy limitations for WhatsApp; latest challenge to Affordable Care Act may imperil preventive coverage; odds ratios vs. relative risk; retracted papers still get cited for while; Arkansas prisoners given ivermectin; who should regulate brain-computer interfaces?; UK government mines data for behavioral nudges; Pew surveys pandemic internet use; much more:


DEEP BREATHS

Photograph of multiple stacks of old hardcover books, turned so that the books’ spines are facing the viewer.
Image credit: Ed Robertson/Unsplash
  • A codex is what many of us think of when we think of a “book” – sequential written or printed pages bound between hard covers. But for a very long time, the continuous scroll was the dominant format for written material across much of the world. What changed to make the codex design the eventual winner across Europe? As this Twitter thread by @incunabula explains, the answer involves religion, ham & cheese, linen underwear, and marine snails – among a host of other esoteric factors.
  • “But here was the catch – by counting the layers of vertical crystalline rock, Powell estimated that this section should be 10,000ft (3,050m) thick. In reality, it measured just 500ft (152m). There were thousands of feet of missing rock – it had just vanished. He named this feature The Great Unconformity, and asked himself, ‘how can this be?’”. An article by the BBC’s Zaria Gorvett chronicles a lingering geological puzzle that revolves around a missing tract of geological record – the Great Unconformity - that amounts to a mission billion years of Earth’s history (H/T @ArielDora).
  • “The core from Rochette now rests in Perseverance’s belly, hermetically sealed and ready to wait many years until future spacecraft can retrieve it and any other cores the rover manages to collect. The goal is to gather about 35 cores representing the geological history of Jezero Crater, Perseverance’s landing site — which was home to a river delta billions of years ago and might contain evidence of ancient Martian life.” A rousing chorus of “I wanna rock!” for NASA’s Mars Perseverance rover, which has now, uh, persevered (despite some initial setbacks) in picking up its first sample of the Martian regolith, which will eventually be sent back to Earth for study by scientists. Nature’s Alexandra Witze has the story.

AI, STATISTICS & DATA SCIENCE

  • “…before consideration for use in clinical practice, it is important to evaluate a model's performance in separate data independent to that used for model development. This process is known as external validation and helps decide whether an existing model is fit for purpose in the target population and settings of interest.” A research paper by Riley and colleagues published in Statistics in Medicine presents methods for calculating sample sizes for performing external validation of a clinical prediction model with a binary outcome.
  • “As a PhD student focused on AI, it’s been hard to figure out how to keep up with it all. So, over the past few years I’ve sought out and subscribed to a ton of newsletters that help me do that. In this piece i’ll share what I consider to be the best currently active newsletters, with a bit of commentary on why I think they are good.” At his Medium page, engineering PhD candidate Andrey Kurenkov provides a list of AI-and-machine-learning newsletters that can help those interested stay on top of a rapidly evolving field.
  • Closeup photograph of an analog car speedometer gauge, with red-illuminated numbers and the indicator needle resting on zero.
    Image credit: Suraj Kardile/Unsplash

    “While it’s usually good for the government to achieve goals like reducing house fires or preventing cybercrime, Collier and his colleagues warn that the rise of ‘influence government’ could cause harm. Not only does it encourage departments to play fast and loose with personal data – using notes from an interview under caution to build a profile of a typical cybercriminal, for instance – it can also focus negative attention on vulnerable and disadvantaged groups in ways that could be counterproductive.” The Guardian’s Alex Hearn report on the British government’s far-reaching efforts to apply the results of extensive (and sometimes intrusive) data mining to “nudge” the behavior of citizens in what it considers a desirable direction.

  • “Many car companies use neural networks to identify objects on the road, but Tesla is relying more heavily on the technology, with a single giant neural network known as a ‘transformer’ receiving input from eight cameras at once.” Wired’s Will Knight reports on recent revelations that Tesla is developing its own customized processor chips to expedite training of the neural networks that the company’s cars rely on for its “Autopilot” self-driving systems.
  • “The OR should replace the RR in clinical research and meta-analyses though there should be conversion of the end product into ratios or differences of risk, solely, for interpretation.” Odds ratios (ORs) or relative risk (RR)? A paper published in the Journal of Clinical Epidemiology by Doi and colleagues – itself part of an ongoing conversation about measuring effects in clinical epidemiology – comes down on the side of the odds ratio for most applications.
  • “The software lets employers indicate requirements like degree requirements, certifications and licenses, along with negative attributes a candidate's application should be without, like criminal charges, for example. It becomes tricky when the negative attributes are more subtle -- like when the software spots long gaps between jobs on a resume, according to the report.” CNET’s Shelby Brown reports on a recent study that finds AI-enabled resume scanners are misfiring on qualified job candidates (H/T @hypervisible).
  • “A car’s speedometer tells you how fast that car is moving at any given time. The scientific term for this is velocity. Acceleration is a measure of the change in velocity, in other words how fast something is speeding up or slowing down….The daily count of new Covid-19 cases can be thought of as the pandemic’s velocity. If velocity tells us how bad the pandemic is at any given time, acceleration tells us how fast things are getting better or worse.” At STAT News, a set of data visualizations presented by Emory Parker allows readers to assess the degree of “acceleration” with regard to COVID developments (for better or for worse).

BASIC SCIENCE, CLINICAL RESEARCH & PUBLIC HEALTH

  • “Some people are still at risk because they can’t yet be vaccinated or because vaccines don’t work for them. They hope that their fellow Americans will take care of them. They hope that the rest of us will agree to sensible precautions or protections to keep them safe until they, too, can be immunized, or until the danger from exposure eventually subsides….Much of the public is refusing. That’s not new, though. In America, it’s always been like this.” At The Atlantic, pediatrician Aaron E. Carroll points out a potential lesson offered by the pandemic – a lesson that hopefully will lead to increased awareness of the need to protect people with disabilities or immune challenges that puts them at much greater risk during a pandemic.
  • Pregnant woman standing in profile in a meadow or field, with trees in the background.
    Image credit: Taylor Wright/Unsplash

    “More than 60% of respondents to a 2016 Nature survey said they had tried to repeat other scientists’ experiments and been unable to do so. A poll of members of the American Society for Cell Biologists similarly found that more than 70% had been unable to replicate a published experimental result, with incomplete detail in the original protocol given as the most common explanation.” A feature article at Nature by Monya Baker elucidates some key approaches and tools for ensuring that experimental protocols are sufficiently clear and unambiguous that they can allow others to replicate experimental results.

  • “Estimated vaccine effectiveness from 7 through to 56 d after the second dose was 96% (95% confidence interval 89–100%) for any documented infection, 97% (91–100%) for infections with documented symptoms and 89% (43–100%) for COVID-19-related hospitalization. Only one event of severe illness was observed in the unvaccinated group and no deaths were observed in either group.”  A brief communication in Nature Medicine by Dagan and colleagues reports findings from an Israeli cohort study of the effectiveness of the Pfizer/BioNtech COVID vaccine in pregnant persons (H/T @florian_krammer).
  • “Murphy isn’t ready to abandon ship. He is among those dedicated to pairing old-fashioned exploration with 21st-century technology, believing that biology will continue to drive next-generation drug discovery, that the real world is the ideal crucible for forging potent new compounds.” A fascinating article at STAT News by Peter Andrey Smith profiles scientists who are undertaking a literal deep dive for novel antimicrobial compounds – and as they do so, racing against time and rapidly encroaching “digital” competitors.
  • “Research is beginning to reveal that the reason children have fared well against COVID-19 could lie in the innate immune response — the body’s crude but swift reaction to pathogens. Kids seem to have an innate response that’s ‘revved up and ready to go’, says Herold. But she adds that more studies are needed to fully support that hypothesis.” An article by Nature’s Smriti Mallapaty examines why children have tended not to suffer from severe COVID nearly as often as adults – and why the spread of the delta variant may be changing the terms of the equation.
  • “There is more to disease risk than genetics. For most common, late-onset diseases, individual risk is heavily influenced by non-genetic factors. Often collectively labelled as environmental, these might include factors as varied as diet, socio-economic status, access to health care, the status of personal relationships and gut-microbiome diversity….It is not straightforward to measure and integrate these factors into risk estimates.” A Nature commentary by Mark McCarthy and Ewan Birney argues that predicting individual health risks is far more complex than simply relying on genetic indicators (H/T @cecilejanssens).
  • “In general, the world would be far better off if there had been fewer but better Covid clinical trials. I'm all for trying out new ideas - that's essential, in fact. But try them out for real. Don't throw something together just because you can, or because you might get lucky, or because you might get a paper out of it one way or another. If you're going to do research on human beings, you owe it to the subjects of your trial and to the rest of the medical community - and to the rest of the world, in this case - to do it right.” At In the Pipeline, Derek Lowe surveys the proliferation of COVID-related therapeutic clinical trials and asks whether too much is enough, considering the dubious quality of some of the investigations.
  • “Unless countries that have purchased vaccine doses and companies that have already brought vaccines into use agree to find ways to resolve the problem, manufacturers that trail the first wave of producers may not be able to prove that their vaccines work. Not only will that slow efforts to vaccinate the planet, it will block development of next-generation vaccines, and it will stymie efforts to answer key public health questions, like whether boosting with a different vaccine would generate better protection, or whether giving smaller — fractional — doses could protect more people more quickly.” At STAT News, Helen Branswell reports on potential barriers to further development of COVID vaccines and boosters.
  • “The odds have historically been dramatically stacked against women gaining access to science education and research careers, though, and if they did get through, against getting credit for their contributions – and then against being celebrated, too. All of which makes those who did achieve important breakthroughs and prominence all the more impressive.” In a series of posts at her PLOS blog “Definitely Maybe,” Hilda Bastian chronicles the contributions of women in developing and improving vaccines across the 20th Century.

COMMUNICATIONS & DIGITAL SOCIETY

  • “WhatsApp messages are so secure, [Mark Zuckerberg] said, that nobody else — not even the company — can read a word….Those assurances are not true. WhatsApp has more than 1,000 contract workers filling floors of office buildings in Austin, Texas, Dublin and Singapore, where they examine millions of pieces of users’ content. Seated at computers in pods organized by work assignments, these hourly workers use special Facebook software to sift through streams of private messages, images and videos that have been reported by WhatsApp users as improper and then screened by the company’s artificial intelligence systems.” A deep-dive investigation by ProPublica’s  Peter Elkind, Jack Gillum and Craig Silverman documents the gap between Facebook’s assurances of ironclad privacy for users of its popular WhatsApp messaging app and the actual reality of the company’s assiduous sifting of those messages for objectionable or illegal content – or for the company’s own purposes.
  • A preprint of a paper by Tzu-Kun Hsiao and Jodi Schneider, currently accepted for publication by the journal Quantitative Science Studies, presents findings that show biomedical papers that have been retracted continue to be cited for some time, although such citations eventually tail off. Moreover, they also find that even when retracted papers are cited, the fact of their retraction is only rarely noted by the citing authors.
  • For those keeping track, it is currently the year 2021, and yet: “An Arkansas doctor under investigation for prescribing an anti-parasite drug called ivermectin to jail detainees with COVID-19, even though federal health officials specifically warn against it, has said that those patients took the drug willingly. But several inmates at the Washington County jail say that is not the case — that they were given the pills with no indication of what they really were.” CBS News’ Li Cohen has the story.
  • A woman sitting against a dark background with hand to face, her face illuminated by what appears to be the glow of a phone or computer screen.
    Image credit: Niklas Hamann/Unsplash

    “That meeting has now led to the formation of an ambitious new anti-aging company called Altos Labs, according to people familiar with the plans. Altos is pursuing biological reprogramming technology, a way to rejuvenate cells in the lab that some scientists think could be extended to revitalize entire animal bodies, ultimately prolonging human life.” At MIT Technology Review, Antonio Regalado fills in some background on what appears to be the debut of a new Silicon Valley-funded company with the ultimate goal of creating effective therapeutics for reversing the aging process.

  • “…tech use has not been an unmitigated boon for everyone. ‘Zoom fatigue’ was widely speculated to be a problem in the pandemic, and some Americans report related experiences in the new survey: 40% of those who have ever talked with others via video calls since the beginning of the pandemic say they have felt worn out or fatigued often or sometimes by the time they spend on them. Moreover, changes in screen time occurred for Americans generally and for parents of young children.” Recently released Pew Research survey data documents how Americans have been using the internet during the COVID pandemic, and finds that while most have relied on internet options to continue working and learning, these technologies have not been uniformly beneficial, or even accessible, for everyone.
  • Retraction Watch documents an ongoing wrangle over the retraction (amid legal threats) of a paper critical of the inclusion of purportedly “predatory” journals in the Scopus journal database – a wrangle that includes strong pushback from the authors, who objected to the retraction, and by an editorial board member of the retracting journal (Scientometrics), who may resign in protest.
  • “Even though this project is still being tested—with mixed reviews—Starlink is getting a lot of attention in Washington at a moment when the government is willing to spend taxpayer dollars on infrastructure and take chances on new broadband deployment methods.” At Scientific American, a brief video documentary by Jacob Templin traces ongoing efforts to position SpaceX’s Starlink satellite network as a provider of broadband internet connectivity in rural and remote locations.

POLICY

  • “The sheer amount of health apps available to consumers certainly raises problems. For example, it would be desirable for consumers to know whether these apps work as promised and whether their privacy is adequately protected. Some kind of review would definitely be welcomed. However, the FDA has limited resources, and the review of all health apps is unlikely to be feasible.” A blog post by Sara Gerke and Chloe Reichel at the Harvard Petrie-Flom School’s Bill of Health blog examines the question of how direct-to-consumer health apps should be regulated, and whether a “one-size-fits-all” approach is appropriate.
  • Slide5.PNG
    Image credit: Mat Napo/Unsplash

    “Such tragedies arise from a quiet public health failure made no less deadly because it’s so easily overlooked: policies, protocols, and practices mandating that first responders transport stroke emergencies to the nearest hospital regardless of the capacity and proficiency of the receiving hospital. These practices quietly cost human lives, particularly in big-city America, where the nearest hospital may be only marginally closer than one better equipped to provide proper treatment.” In a new post at Health Affairs blog, authors Harold A. Pollack, Shyam Prabhakaran, Ali Mansour argue for revisiting standards for emergency transportation of stroke patients to ensure that patients arrive soonest at a facility that’s best suited to treat them.

  • “If Section 2713 were repealed, insurers would have the freedom to reimpose patient cost-sharing for preventive care. In the short run, this could increase the financial strain that patients face when seeking preventive care and discourage them from doing so. In the long run, this could result in increased rates of preventable and expensive-to-treat chronic conditions. And because Section 2713 is what allows free COVID-19 vaccines for those with private health insurance, some patients may have to pay for their vaccines and future boosters if the provision is axed.” At The Conversation, Paul Shafer and Alex Hoagland describe the latest in an ongoing series of legal challenges to the Affordable Care Act, and walk readers through the likely implications for coverage for preventive care should the lawsuit prevail.
  • “To say that we’re on the brink of disaster offers hope that the people in charge can take steps to keep us from plunging toward an abyss. It suggests that the situation is at least temporarily sustainable, that maybe you can keep hunkering down and doing what you’ve been doing, and everything will be fine. But it is not sustainable, and it is not fine.” In a bleak and bracing article for Slate, Brown University physician Vishal Khetpal argues that the US healthcare system is failing in critical ways.
  • “It is well within the FDA’s purview to assess devices, including spectrum-use BCIs [brain-computer interfaces], strictly along the lines of what is safe and effective. However, implicit in considerations of safety and efficacy is also an assessment of risk and benefit. This assessment involves a value judgment not only about a device’s absolute level of risk and benefit but also about what is the right balance between the two.” At the AMA Journal of Ethics, a viewpoint article by Binkley and colleagues probes the knotty question of how, and be whom, implanted brain-computer interfaces (BCIs) should be regulated (H/T @scuethics).
  • “…police officers are trained to use strategies that include violence and harm, forcing their subjects to acquiesce. While not a new phenomenon, police violence among racial and ethnic minority groups has been catapulted to a subject of considerable attention within the national public discourse, capturing the interest of individuals, communities, and policy makers across the US over the past year.” A viewpoint article by Heard-Garris, Johnson, and Hardeman published in JAMA Pediatrics examines the health toll inflicted on racial and ethnic minorities by current policing methods.
  • Following up on a story included in a previous Roundup: an Ohio judge’s decision requiring a hospital to administer ivermectin to an intubated COVID patient has been reversed, per the New York Times: “The judge, Michael A. Oster Jr., wrote that ‘there can be no doubt that the medical and scientific communities do not support the use of ivermectin as a treatment for Covid-19’ and that the plaintiff had failed to provide convincing evidence to show that it was effective.”