It could be useful in a trial setting as a screening tool for study population enrichment. A big problem with clinical trials investigating treatments for dementia (and in particular, Alzheimer's) is that either (1) the intent-to-treat population contains too many participants who don't have the disease; or (2) the patients who have the disease are at a stage of disease progression that reduces intervention efficacy.

We've conducted trials in the apolipoprotein-e4 population, for example, and more recently, have used PET imaging results, but a blood-based biomarker with even moderate predictive power would be a tremendous help, because the ability to detect these patients at an early stage of their disease could very well translate to clinical evidence of a treatment's ability to stop symptomatic disease.

But I agree that without extremely high sensitivity and specificity (like, > 99% and > 95%), and without treatment options, such a biomarker would not be useful diagnostic information for the general population. I have hope that we are getting closer to nailing down Alzheimer's pathology.

Independent parallel programming might have caught the error. Just have two individuals create independent implementations of the same calculations. Neither can see the other's work. A third party views the outputs and checks that they match. If not, they have to find the error. I don't think that would cost 10x the amount.

There's no such thing as 100% accuracy. Errors always happen. It's just a matter of how much error one is willing to accept, and if the stakes are high enough, you do a cost-benefit analysis to see whether the effort to bring the risk down outweighs the cost.

In this case, it seems like there was a single point of failure, which as we all know, is not good practice.

They created Zelle and then relentlessly pushed it to their account holders. I've lost count of how many times they've sent me advertising to set up Zelle, how easy it is to transfer money, blah blah blah. They also deliberately avoided implementing adequate security measures to prevent scams, because doing so would impact ease of use, and they wanted to put the responsibility solely on the account holder for avoiding them.

I never wanted Zelle and I will never use it. They keep trying to force me to use it. If you offer a service that is so easily misused, you can't be absolved of liability when people get tricked, because you haven't done your due diligence in educating your account holders on how to use it safely. You created this security hole, so you have to pay.

I would argue that $2500 doesn't even begin to scratch the surface of the damages for a first-time malware intrusion, because the hardware isn't the most valuable property; it is the data stored on or accessed with the hardware.

Malware installed on a device through a malicious advertisement could be ransomware that irretrievably destroys unique data (and their backups, if not detected in time); it could also be a password sniffer, which could compromise otherwise secure banking credentials. On a work computer, it could result in exfiltration of corporate trade secrets. There is literally no upper bound on the extent of damages that could apply. The machine itself is the least of one's concerns.

The idea that a company in the business of serving ads to users could avoid performing due diligence that such data would not be malicious, yet unilaterally claim that they are not legally liable for the outcome, is absolutely repugnant. Oh, I'm sure they do some level of checking, but malware has been served through ads in the past, and there is absolutely no reason to believe that it will never happen again. Yet Google not only absolves itself of any and all responsibility, but has the arrogance to circumvent measures put in place by the user to prevent such a security threat. They absolutely need to be punished extremely severely for this.

For anyone who says "if you don't want to see their ads, then don't use their services"--what cave have you been living in for the past 30 years? Google and Facebook have their tentacles in every fucking website on the planet. Blocking is the only way to reliably access the web, because you cannot know in advance whether or not a given site accesses their code and ads. I don't use Facebook at all, I don't have an account, and yet I'm forced to block all of their invasive scripts on the client side in order to avoid being tracked by them. Google is even worse. Even if you try to get away, you can't.

All I did was look up the wikipedia article for tritium https://en.wikipedia.org/wiki/... and it seems that given the short half-life (so that it will not persist in the environment and remain a problem for centuries, like cesium-137), the lack of accumulation in biological tissues (so that it does not become concentrated in the food chain, unlike mercury), the rapid dilution of volume into the ocean, and the apparent release by other countries (most notably, France) of tritium into bodies of water at least an order of magnitude higher and on an annual basis (see the table under Fission), the Fukushima wastewater release really does not seem to pose a threat to the environment or to public health.

That last point is something I think needs emphasis: I do not claim to be an expert, but if my understanding of that table is correct, and the amount of tritium being stored from Fukushima is correct, I would like someone with more subject matter expertise to explain why there is so much controversy over this decision when France dumps 10x more tritium into the English Channel from their nuclear plants every year. I suspect the objections (domestically and from neighboring Asian nations such as China) are either based in ignorance or they are a political ploy to gain some kind of leverage.

I did not say they did not need to enroll ANY unvaccinated people. Safety is highly unlikely to be worse in the unvaccinated population than in the previously vaccinated. And if it were, then anyone who is unvaccinated can get the old vaccines first, then get the new one--not that any of them would.

What you say is true, but it's not what the original issue was about, which was recruitment of unvaccinated people for large scale confirmatory trials. Your point, that the sample size needs to be substantially larger to measure a smaller effect size due to a population that already has some degree of immunity, may or may not be an actual problem. The reason why is because during the time the major vaccine trials were being conducted, the US was under lockdown, distancing was enforced, and there were no variants that strongly evaded immunity. If the attack rate is higher with the current variants and people are no longer masking or staying home, it could end up being the case that efficacy can be established with fewer subjects or less observation time than those previous trials. For instance, just looking at the reported case rates of the omicron variant, it is easily spreading at a rate far exceeding that observed during mid 2020. This suggests that omicron is able to spread largely unimpeded through a twice-vaccinated population.

Certainly, the ideal case is to enroll roughly equal numbers of volunteers from the unvaccinated and vaccinated populations, but we know this is not going to happen. My earlier comment is simply saying that the estimation of vaccine efficacy in the unvaccinated population is not necessarily the primary endpoint.

I don't think an appropriate trial design should compare the novel vaccine candidate against a booster of approved vaccine. While one can certainly study this comparison, it should not be the basis for an EUA, which may be granted if the candidate is reasonably safe and effective, not necessarily superior to an approved product. Full approval depends on the overall risk/benefit profile. As such the design I have in mind for a Phase III study to support EUA would be a placebo-controlled trial. Even a single arm open-label study with historical control would be preferable to a design with active comparator, which can come later, since there is now just so much data about the existing vaccines. One could even do unequal randomization and use a limited placebo arm as a bridge to match against real-world data, further gaining efficiency in sample size.

Also keep in mind that those previous studies were Phase I/II/III seamless designs; they were also powered for a VE of at least 50%, whereas the actual VE observed was around 95%. I don't know how the Army plans to conduct their clinical development program. It could be slower if they use the traditional approach, or it could be faster if they can leverage resources that private industry doesn't have.

Your observation brings up an interesting point that people working in drug development (specifically, clinical trial design) deal with quite often, which is, how do we determine efficacy in the indicated population if there already exists a standard of care treatment? In this case, as you noted, a large portion of the US population has already received vaccines, and the ones who have not are unlikely to volunteer for a clinical trial to determine the efficacy and safety profile of a new one.

But in fact, this does not pose a significant problem: precisely because most people have been vaccinated, it is no longer of primary relevance to measure the efficacy in a vaccine-naive population. What you actually want to measure is how well this new vaccine performs compared to people who may have been previously vaccinated, but do not get this new vaccine--in other words, how much additional protection does this vaccine confer? The answer to that question is what the FDA will be looking at because if over half the US population has already gotten some kind of COVID vaccine, the way this new vaccine behaves (both in terms of efficacy and safety) in this population is just as important as how it might behave in a vaccine-naive population.

A similar principle applies in general to many drugs with existing treatments. For example, the initial management of Type 2 diabetes often involves metformin, a widely used drug with an excellent safety profile. When new antidiabetic medications are researched, the trial sponsor isn't going to exclude patients who are metformin-naive for the sake of determining efficacy in a population with no previous antidiabetic therapy, because in real-world practice, patients already commonly have it prescribed to them.

FIH (first in human) trials are a bit of a different story, because the goal is to establish that the drug is safe for the purposes of enrolling patients who have the condition the drug is supposed to treat. So these trials generally involve dose-escalation designs on a very small number of healthy subjects. For a vaccine, this would require enrollment of both vaccine-naive and vaccinated volunteers, because such a trial needs to establish that it is safe to enroll both of these groups in a larger Phase II trial.

If you go to the manufacturer's website and read the prescribing information https://www.rxabbvie.com/pdf/vuity_pi.pdf%20%20, you can see on page 5, Section 14, Clinical Studies, that the efficacy is actually very modest, 31% vs. 8% in the first trial, and 26% vs. 11% in the second (vuity vs. vehicle; "vehicle" here means placebo drops). While statistically significant, in practice what this means is that maybe at most 1/3 of people who try these drops would expect to receive substantial clinical benefit that would translate to real-world improvements in near vision. That's not exactly impressive, especially when you consider the limited duration and the risk of infections from constantly using eye drops.

Like many drugs, there is a use case for these drops, which is ultimately why they got approval, but since the effect is likely to be modest and short-lived, it probably won't make sense for most people with presbyopia. Your eye doctor could let you try it but it's by no means a life-changing treatment.

Sorry, but that's a myopic and naive point of view. People take matters into their own hands because of a lack of accountability or justice. If the "rule of law" were actually followed through, countless victims would recover what they lost, and the perpetrators would be in jail. If you want people to stop baiting scammers, then do your fucking job and shut down the scammers. If these scambaiters can find the thieves this easily but "official channels" somehow can't, that means they lack either the ability or will, and any sense of credibility to be lecturing others on the rule of law.

You have no idea how glad I am to see you post this. I've been saying the same thing since last year and the message keeps falling on deaf ears.

We have had numerous well-documented cases of superspreading events since early 2020 in which no other causal mechanism for the transmission of the virus could be plausible. There was the Korean call center, in which people who were sitting far away from the index case got infected; the choir singing at a church and infecting people not sitting nearby; the restaurant at which the indoor HVAC system was suspected of blowing aerosolized viral particles through the room.

Yet public health agencies and businesses are still operating under a long-since discredited model in which droplet spread is the primary mode of transmission. We have corporations requiring return to in-person work and setting up partitions and distancing in enclosed spaces, and restaurants with cramped indoor dining separated by plexiglass.

That said, the delta variant and other more virulent lineages that have evolved certainly are more pathogenic and more easily spread--the evidence for that is as plain as the fact that these have supplanted earlier strains--but the reason is not so simple as to say that they are somehow "more airborne." That is revisionist history and an attempt to deflect from or cover up the truth that the public health agencies have mischaracterized the nature of coronavirus transmission for well over a year, and the consequences of which have directly lead to the preventable deaths of many people.

Someone who does not understand the risk/benefit profile of a vaccine that has overwhelming real-world evidence of efficacy and safety is, I would argue, by definition less educated than those who do understand it. I would not claim that such idiocy is the sole domain of right-wing conservatives, but the data on vaccination rates by political affiliation would suggest it is a huge proportion of them. Those who are supposedly progressive-minded but do not accept the evidence are just as incapable of critical thinking. The point I am making is that distrust of science and education is a narrative perpetuated almost entirely by the conservative end of the political spectrum, but the effect of which is seen in both liberals and conservatives in the form of entire generations of Americans who do not know how to reason.

There has always been a distrust of science in the less educated segments of American society, a sort of arrogant suspicion of evidence-based reasoning that is supplanted by a mode of thought motivated by fear, superstition, dogma, and spite. The use of specific terms and catchphrases meant to denigrate rational thought, such as "(liberal) elites," is a reflection of a deep-seated cultural hatred and resentment of science that is perceived as something that is force-fed to the American public. Anything that dares to contradict their worldview is demonized, yet their small and menial lives rely upon and are constantly enriched and empowered by thousands of years of hard-fought scientific and technological advancement.

These same people who deny vaccines and poison themselves with veterinary antihelminthics solely because they have been told that this "left-wing" segment of society has shown vaccines work, yet will rush to the hospital begging to be saved from their own stupidity, are the predictable and inevitable consequence of generations of deliberate educational sabotage and right-wing corporate/political media propaganda specifically designed to engender distrust of science and prevent teaching of critical thinking, so that the wealthy ruling class can have access to cheap, abundant, unquestioning slave labor, and the politicians they buy can more easily manipulate and disenfranchise the public.

COVID is not the disease that is rotting American society. The pandemic, at least as it has unfolded within US borders, is merely a symptom of the underlying cancer of scientific illiteracy and the failure to teach scientific skepticism, as Carl Sagan had explained in his book "The Demon-Haunted World: Science As a Candle In the Dark." Each successive generation of Americans has become increasingly less able to reason logically; to the point now that large numbers of people are eating horse paste. And the internet, which delivers information but not wisdom, only serves to increase the rate at which disinformation spreads.

Even if COVID goes away, there will be another catastrophe. Maybe it is climate change, or maybe another disease. If Americans refuse to learn, they will suffer the consequences.

The scientific and therapeutic landscape of Alzheimer's is littered with the data from countless failed trials and compounds, and a big part of the reason why has to do with a lack of understanding of AD pathophysiology. But it is not the only reason. In fact, there are other considerations in the context of Alzheimer's disease that are in some sense precursors to why we have not affirmatively established the disease mechanism.

Foremost is the fact that AD is not the only type of dementia, and that until the advent of relatively recent noninvasive imaging modalities, identifying the afflicted has been largely retrospective (histopathology from autopsy), or based on symptomatology (i.e., clinically identifiable loss of cognitive function). Therefore, by the time patients are suspected of having Alzheimer's, they are either dead or they have progressed to the point where the damage is significant and likely irreversible. This impedes the investigation of a complete picture of the course of disease. Blood and CSF biomarkers are of course too indirect--and it is perhaps for the reason that amyloid levels are measurable that the science has focused so much on the notion that abnormal elevations are a causative agent for AD, and the reason why the drug development industry has chosen to advance compounds targeting the reduction of amyloid.

That said, I don't think that any AD researcher worth their credentials is wedded to the amyloid hypothesis as the only possible explanation for the disease. Most, if not all, are well aware of the complexity of the picture. The real problem in the Alzheimer's space as it pertains to regulatory approval is the lack of meaningful and reliable efficacy measures for assessing the impact of treatment on disease modification. While reduction in amyloid may be used as a surrogate endpoint that could justify accelerated approval, it has yet to be validated as an endpoint for traditional approval--the FDA still considers tests of cognitive function and activities of daily living to be the standard by which efficacy should be measured, because these are things that matter to the patients and their caregivers, and not how much amyloid burden is present. The challenge with such measures of efficacy, however, is that they are largely based on questionnaires and are subject to numerous statistical and interventional biases, as well as high noise.

That said, I am highly skeptical of Biogen's filing, and the original advisory committee meeting earlier this year brought to light some glaring issues with study conduct and the statistical integrity of their submission. This of course has nothing to do with whether I believe amyloid is the correct target, but it says volumes about how (un)reliable are the efficacy claims that were made.

In summary, there are so many problems with developing effective therapies for Alzheimer's: we don't really know what aspect of the disease mechanism we need to target (this is what you have characterized); but we also don't know how to accurately select the patient population (as you'd need to detect them reliably and early enough to be able to prevent rather than repair cognitive decline), and we don't know how best to measure effectiveness in a trial setting. If by some miracle we discover how and why certain individuals develop AD (beyond, say, genetic predisposition such as apolipoprotein-epsilon 4 genotype), and the disease mechanism is well elucidated, then the other issues become much easier to address; but one might as well wish for unicorns because the notion that there is some clear explanation for this insidious disease despite many decades and untold billions of research spent, is just not realistic. AD is much more likely to have a multifactorial and complex etiology, or could even represent a constellation of pathologies, sort of like how "cancer" is an umbrella term for numerous distinct diseases and different causes.

Indeed. The thing about incandescent is that it's basically an ideal black body radiator, with a smooth spectral power distribution, as opposed to LEDs which emit radiation at specific peaks. Even the white LEDs, which emit in the UV band which is converted to visible by a phosphor coating (and thus work on a principle that is actually closer to CFL than incandescent), the color rendering index is still far from ideal for color-critical applications. There are "full-spectrum" LEDs such as those found in grow lamps, but just looking at the SPD curves tells the truth--they are just "cool" and "warm" white LEDs assembled in an array.

Sunlight (specifically, daylight) is notoriously difficult to reproduce accurately. We live in an RGB-ified world. I love LEDs for their energy efficiency, small size, and versatility for home illumination, but they are not suitable for artists.