I came to leave a snarky joke comment about "in mice?", and then I remembered that mice don't actually need parachutes because their terminal velocity is roughly the same as a human using a parachute [1]. If someone actually ran the study with mice, they'd find that parachutes aren't necessary!
> You can drop a mouse down a thousand-yard mine shaft; and, on arriving at the bottom, it gets a slight shock and walks away, provided that the ground is fairly soft. A rat is killed, a man is broken, a horse splashes.
"A horse splashes". That's some mental imagery right there.
Makes me wonder if a bunch of the world's smallest flightless mammal (a fraction of the house mouse's mass) were released at altitude, would they survive? (Not proposing that this be done but thought experiments and/or simulations):
At altitude it's very cold, the shrews don't work well in the cold due to the same thing that makes them suitable for this experiment - they are very small, thus they lose heat to the environment easily. So, it's possible they freeze to death first. But yes, I imagine that if they survive the cold and aren't falling onto a hard surface (e.g. they land in grass, not a car park) they would survive.
For them other than the cold a drop from 1000 metres is the same as a drop from 10 metres.
Nah, do the same thing they did for one of the Venus landers--fasten them with something that melts. Once they have reached an altitude where the temperature won't kill them the tether melts and the blanket falls away. (On Venus they did it to cut a chute because it would have made the lander fall so slowly it would have died before landing.)
However, shrew blankets change the whole scenario to irrelevance--a shrew in a blanket sufficient to keep it safe has very different aerodynamic behavior than a naked shrew.
I think if we're going to give shrews blankets, we should just give the shrews blankets, not conditioned on volunteering for this potentially fatal research project.
The J. B. S. Haldane essay he’s referring to is On Being the Right Size from 1926. It’s part of the anthology On Being the Right Size and Other Essays, which you can still find, and is still worth reading.
The fact that mice don't appear to deliberately jump off high things as a way to get down tells me that they probably do get hurt...
Just because they don't fall super fast doesn't mean they don't get injured because they don't have as much distance to "drop and roll" like a human does.
> The fact that mice don't appear to deliberately jump off high things as a way to get down tells me that they probably do get hurt...
They explicitly added the caveat about the soft ground. There'd be very little opportunity for mouses to become comfortable with jumping from high areas as there'd be very few occasions that satisfy the requirement. If the mouse jumped into a cave and can't get out, it'd starve to death. If it jumped off a cliff and fell into water, it would drown. If it fell off a skyscraper, there'd be Asphalt at the ground etc. Off the top of my head, there are no natural places in which mouses could safely get comfortable with jumping. Thus, they'd only do so if they're in mortal danger. And they absolutely do jump off high places if they're terrified
Mice around me fairly frequently climb trees. They know whats at the bottom of the tree (hard soil). Yet I never see them jumping off - they always climb back down the way they came up.
It could be to avoid getting disorientated though, or to be able to see a predator and reverse course...
Where I live in northern Europe almost all ground in the forests are soft. There is moss, small plants, grass or leaves everywhere.
I can't imagine why the mice would climb trees, but if they did and fall down, they probably gonna be fine most of the time.
In many forests there is lots of large rocks and steep slope/cliffs/precipice in the range 0-5 m where mice possibly could shorten their path by jumping. But I'd suppose they rather use a path they can go in both directions.
Well me too but if i could just jump off of stuff if be skipping the stairs a lot more often. I agree with the other person, seems logical to me that for a prey animal they'd be careful about changing their threat landscape too fast. Once you let gravity take over there's no telling where you'll land. You could roll right into a ditch or a cats paws. And maybe they actually do jump sometimes, but only when there's certainly no threat nearby, which would involve no human observers, or maybe seeing other mice where they plan to land.
And then there's the actual vision. Are mice known for being able to see long distance? One could posit, if not then most jumps would be leaps of faith into vague amorphous blobs, full of potential danger and impossible to triangulate or assess.
I tend to wear jeans with skinny legs that don't seem conducive to taking across-the-stairs (I'm thinking 3 to 5 meters in the horizontal direction) jumps.
Why do I wear impractical clothing? I never even thought of it as impractical.
Cats live warm--what is a perfect thermal balance for them (where their body needs not expend any effort for heating or cooling) is unpleasantly warm for a human (who will be sweating.) Typical water will be chilling for a cat. Going swimming in most environments will pose a hypothermia risk at times and thus fearing it is a logical evolutionary result.
The grey squirrels in my UK garden routinely make what we call the 'leap of death' between two trees whose upper branches are a few feet apart, at different heights. I've seen two falls - one because the target branch snapped when the squirrel landed, the other during a high-speed chase where one of the squirrels apparently mis-jumped. In both cases the squirrels fell into lower level bushes and disappeared (no subsequent bodies). It's not unusual to see a squirrel with one one useless leg but still managing to feed and get around, so I can quite believe that they can recover from a broken limb given sufficient time.
* * *
We also had a pheasant who routinely visited the garden and one day showed up with a broken leg, that survived for ages before eventually disappearing. Initially he sat quietly in the undergrowth but started hopping around and doing short flights (with clumsy landings). I put out bird seed for him and he incredibly quickly learned to hoot and flap outside my back door when he was hungry. He once followed me into my shed when he was expecting feeding and had to be shoo'd out.
Flying squirrels show an intermediate stage of such an evolution process. Evolution always involves intermediate stages. Even if a large discrete change could occur, fine tuning it is an extended incremental process.
If you go to the Cornell Ornithology Lab feeder cam and scan through the feed at night, you will find flying squirrels at the feeder. They show signs of nocturnal adaptation, another connection to bats.
> mice don't appear to deliberately jump off high things
It could also be because they don't like to change their location too fast. You never know where you could end up. When you fall you might fall next to a cat for instance.
Note that mouse survival depends on the landing. It's quite reasonable for a mouse to fear heights even if in the *typical* situation it won't be hurt. If a mouse freely jumped down and was impaled 1 time in 1,000 that would be a very bad choice for the mouse. Also, a mouse who jumps down might find itself in a situation where it can't get home.
> Parachute use did not significantly reduce death or major injury
I really recommend reading. It is a very funny thing until you realize you've read a few dozen serious articles, well received by a large community (some times even the academic community) with nearly the same kind of arguments.
Also correlation is not causation. Just because some people jumping from planes have parachutes doesn't mean that all people jumping from planes have parachutes.
Yup, *highly* skewed sample, we can't draw conclusions. The number of people who have willingly jumped from planes without desiring death is tiny. The number who have survived is even lower. (Yes, people jump without desiring death. Death by splat is better than death by fire.)
In the first few days of their life, wood ducklings jump out of their nest and fall to the ground. If their nest is in man-made bird-box, that's probably a 4 foot drop into the water. If it's a natural nest in a tree, that might be a 50 or 60 foot drop onto the ground.
I think people misunderstand what evidence based medicine is[0]. There are many types of evidence ranging from lab studies and expert opinion at the bottom to systematic reviews/meta-analyses at the top.
The point being that we should strive for RCTs when possible, and that RCTs carry more weight than expert opinion or case-control. Many (I don’t have an exact number) clinical practice guidelines are based on expert opinion or observational studies, and are completely valid/evidence-based (it’s just weak evidence).
The point of EBM is not “RCT or bust” it’s really more about the hierarchy, that medicine should not practice in defiance of quality evidence and that we should strive to obtain the highest quality evidence possible.
> Many (I don’t have an exact number) clinical practice guidelines are based on expert opinion or observational studies, and are completely valid/evidence-based (it’s just weak evidence).
A big problem is that humans have a preference for action, and a dislike of waiting for information.
Many flawed guidelines are based on a desire to act even in the absence of evidence of benefit.
"Primum non nocere" is probably the most accepted but least respected tenant of good medical practice.
> Many flawed guidelines are based on a desire to act even in the absence of evidence of benefit.
Very true, I mostly intended to illustrate that many "good" or accepted guidelines are not necessarily based on randomized controlled trials as there seems to be a misconception (based on recent trending posts) that these are the only forms of "evidence" in medicine.
Some guidelines are borderline pseudoscience but pertinently guidelines are not binding. You point out the correct approach - one is entirely allowed to deviate from guidelines for good reason (i.e. no/very weak evidence, questionable benefit).
That's also the beauty of the EBM, you'll find grade & strength of recommendation in any modern guideline to help interpret them. We also have quality assessment tools like QUADAS-2 and the STARD reporting guidelines to help the reader assess quality.
> "Primum non nocere" is probably the most accepted but least respected tenant of good medical practice.
Also agree but I think EBM addresses this a lot more than the old days of science-based medicine where we did things because "they made sense" without looking at various outcome measures that sometimes reveal missed harm.
> A big problem is that humans have a preference for action, and a dislike of waiting for information.
Part of this is also by necessity due to the nature of patient encounters which have a need for decision-making.
Where the guideline is clearly poor (e.g. adrenal nodules, homeopathic follow-up at best) I disregard them but sometimes you have to err on the side of caution.
As an example I recommend pancreatic cyst follow-up per the ACR guidelines even though my (abdominal radiologist)/institutional expert opinion is they're overkill and unnecessary based on very weak and dated observational evidence. But what's the alternative, everyone just does what they want and a patient's follow-up will be variable depending on the reporting physician?
Apologies, bad medicine habit to use ambiguous acronyms.
I meant "randomized controlled trial" in this case (as the post). "Randomized clinical trial" is often used interchangeably but does not necessarily require a control group (e.g. a non-inferiority study comparing two different treatments).
Background (2003): A systematic review of the literature found no evidence to support the routine use of parachutes to prevent death from gravitational challenge and concludes with "We think that everyone might benefit if the most radical protagonists of evidence based medicine organised and participated in a double blind, randomised, placebo controlled, crossover trial of the parachute." [https://www.bmj.com/content/327/7429/1459]
2018: Protagonists of EBM both organized and participated in such a study and found that parachutes don't make any difference in survival when jumping from a plane.
> Advocates of evidence based medicine have criticised the adoption of interventions evaluated by using only observational data. We think that everyone might benefit if the most radical protagonists of evidence based medicine organised and participated in a double blind, randomised, placebo controlled, crossover trial of the parachute
> “Indeed, in seminal work published in the BMJ in 2003, a systematic search by Smith and Pell for randomized clinical trials evaluating the efficacy of parachutes during gravitational challenge yielded no published studies.1 In part, our study was designed as a response to their call to (broken) arms in order to address this critical knowledge gap.”
> individuals were screened for inclusion in the
PArticipation in RAndomized trials Compromised by
widely Held beliefs aboUt lack of Treatment Equipoise
(PARACHUTE) trial.
I learned a new word, "equipoise". That is, there an ethical dilemma when conducting a controlled trial: it's not ethical to withhold treatment from the control group, if the treatment is already known to work. You'd be basically sacrificing peoples' health for the sake of trial results - to get data that's not worth much (since the treatment is already known to work). This seems to be a big issue in medical trials.
> Only two options exist. The first is that we accept that, under exceptional circumstances, common sense might be applied when considering the potential risks and benefits of interventions. The second is that we continue our quest for the holy grail of exclusively evidence based interventions and preclude parachute use outside the context of a properly conducted trial.
Both options may be true for different cases. It may be simple to decide on parachutes. But think of bureaucrats who were patting themselves on the head for being “flexible enough and humble enough” to issue new guidance on masks instead of saying “We don’t know” in the first place.
“The sovereign is he who selects the null hypothesis.”
If I'm a passenger in an airliner with no parachute, I definitely don't want my pilot bailing out.
Also, insurance rates on aircraft with airframe parachutes are higher than for aircraft without them - because pilots often panic and pull the chute, which guarantees a hard landing and totals the aircraft, rather than attempting a landing (with, say, a 90% success rate or whatever). The same might apply to individuals: jumping out of a plane is quite dangerous and it might be "statistically" safer to stay onboard and attempt a landing.
> insurance rates on aircraft with airframe parachutes are higher than for aircraft without them
Are we sure this isn't because aircraft iwth airframe parachutes are wildly expensive and/or heavily-customized vehicles? I'm not sure the correlation has been established as causative here.
It’s like the difference between civil law and criminal law. In civil law the balance of probabilities is used. In criminal law the accusation should be proved beyond reasonable doubt.
Option one means treating something like civil law cases. Option two means treating it like criminal law cases.
At first I was confused about how this would work, but thinking a little about it, it should be fairly easy. Give some researchers (randomly selected!) results from studies that are not RCTs ("control") and others RCT-based results ("treatment") to replicate. See which group manages to replicate more results.
Expensive, yes. Would I like to see it? Heck yes, if nothing else as a way to trick the establishment into doing more replication studies.
If the meta-RCT shows that RCTs are efficacious, then RCT's are effective. But, if the meta-RCT shows that RCTs are not efficacious, that means the meta-RCT, which is also an RCT, cannot be trusted, which means that RCTs are still effective.
> Is there any RCT done on evaluating the efficacy of RCT?
What is the set of possible results from the experiment you propose? What function are you going to test: randomness in general, some special RCT event, maybe some doctor who loves to break some double blind experiments?
Did they even consider the speed of the plane? I suspect this is multi factorial and not just a simple case of using a parachute. More funding is needed. And please, don't call me an altitude denier, I firmly believe height has a role to play in this.
Yes.
All participants (parachute and control) jumped from vehicles with 0 speed and 0.6 meters altitude ;-)
I'm not joking. This is the actual study.
It shows how studies can be valid statistically and shows nothing usefull for the real world.
I see this all the time when researching various non-pharmasutical treatments. Vitamin D supplementation has no benefit for people without a deficiency for example. It's annoying how much of this is out there.
Yup, this sort of factor can make it hard to find the answer.
Do some patients benefit from D supplements? Certainly. Yet you can find a group that gains no benefit. In this case we obviously can divide them up but in many cases the division is unknown.
That headline reminds me of the most recent XKCD [1] where an astronomy research group realize how they can churn out findings.
The joke is they’re ruling out the obvious. The headline makes you believe they are stating the obvious (but actually the study is a comment on randomized trials and what kind of research is possible)
All this means is that statistical significance is not enough to reject anything. All statistical significance means is that the effect did not come from sampling error. It can come from any number of other confounders still.
Statistical significance is a hint that it could be worth running more replication studies, and vary the conditions in other studies. Significance alone does not indicate the effect as real.
This para, buried deep in the paper, is the point:
A minor caveat to our findings is that the rate of
the primary outcome was substantially lower in this
study than was anticipated at the time of its conception
and design, which potentially underpowered our
ability to detect clinically meaningful differences, as
well as important interactions. Although randomized
participants had similar characteristics compared with
those who were screened but did not enroll, they could
have been at lower risk of death or major trauma because
they jumped from an average altitude of 0.6 m (SD 0.1) on
aircraft moving at an average of 0 km/h (SD 0). Clinicians
will need to consider this information when extrapolating
to their own settings of parachute use
Real studies have an equivalent para far too often.
The point is that the Cochrane review is probably wrong, and face masks do protect, just like parachutes and condoms. And the more general point is to criticize the whole process, that has made half of the Western medical doctors mask skeptics. The point is to say that there are protective equipment that obviously, measurably works, but for which it is difficult to present evidence from RCT-type studies.
It's also parodying the the problem in some RCT to use unrealistic test cases which are supposed to be representative of actual clinical uses (hence testing parachutes as protection for jumping out of landed aircraft).
That is a good point, but I think their point was to bring attention to the flawed philosophy, so that mistakes could be avoided. And surprisingly soon after their 2018 paper, the flawed philosophy lead to such mistakes in a surprisingly relevant, important and timely matter.
Linking their general points from 2018 to specific examples from the recent pandemic, is of course post hoc. But it demonstrates that their notion was not merely theoretical, but important in practice.
Studies, including RCTs, are often affected by selection biases introduced (intentionally or unintentionally) by how participants are selected and/or self-select. Even the gold standard RCTs are thus not guaranteed to produce valid (or rather, useful/non-misleading) results.
They thoroughly researched the effectiveness of parachutes on a pool of candidates at various speeds and altitudes. While the initial pool of candidates covered a variety of situations, for some reason, the pool of participants willing to actually participate in the study was heavily biased towards people in an aircraft standing still on the airport, with about 60 cm from the door to the ground.
The bias may have been further increased by the selection criteria where participants deemed mentally unfit by the investigator were excluded, and willingness to jump from an airplane at altitude with a 50% chance of an empty backpack increased the likelihood of exclusion.
The bottom half of the right column on page 4 explains how this is relevant for real world studies.
They state the point at the end, and it's serious point: if you were inclined to use "science" to argue against a certain intervention that some group believes in, you could do a randomized trial of that intervention, but manipulate the conditions so that the intervention is unlikely to help. The study would look right and sound right to some people. It would use all the best practices. And yet it would be a bullshit study.
I suppose an example would be comparing the death rates of people who'd been on a statin drug for a few weeks to those who weren't on the drug at all. You'd probably find no different in morbidity, which might suggest that statins don't do anything.
"Best practices" is always bullshit, man. It's a marketing term, not an engineering or a science term. Any practice is bad or good based on the context and purpose for which it is applied.
That's the spirit in which I used the term. In plain English: you can correctly apply certain practices that are famously powerful and acceptable, while subtly undermining them in a way people might not notice, or would happily pretend not to notice if the study fits with their political position on the matter.
Then your supposedly serious point resolves to a tautology "people can bs if they do bs". Duh! This conveys nothing re. any scientific technique or using that in arguments against some intervention
I am explaining the point of the paper. If you don’t think a parody based on that point is useful, maybe take it up with the take it up with the authors, Duh!
They do a Randomized Controlled Trial on parachutes, but due to ethical reasons the plane is only a few feet off the ground and there's no effect (the parachutes don't have time to deploy).
I took it as a commentary on relying on scientific methodology when deciding the importance of a paper. Often the fact that there is a RCT study or whatever, a peer reviewed article in a prestigious journal, this is seen alone as sufficient reason to believe a finding. Scientists make this assumption, laypeople make this assumption, pop-sci media certainly do their thing; but this is essentially a counterexample, a scientific article that ostensibly checks all the boxes of rigorous science while being pretty obviously nonsense.
Sort of this: we asked 100 people sitting on a plane to jump out, with or without a parachute. All those who accepted the challenge were sitting on a fake plane standing still at 0.6 metres from the ground. We saw that the use of the parachute made no difference.
>The PARACHUTE trial satirically highlights some
of the limitations of randomized controlled trials.
Nevertheless, we believe that such trials remain the gold
standard for the evaluation of most new treatments.
The PARACHUTE trial does suggest, however, that their
accurate interpretation requires more than a cursory
reading of the abstract.
It's a satirical response to a well-known prior paper critiquing Evidence-Based Medicine (EBM) that essentially said "we have no evidence to support the use of parachutes, so we should re-evaluate our use of parachutes", and concluded with a suggestion that people demanding evidence for simple things should themselves participate in a randomized controlled trial of the parachute.
It is clear that it is satire, but it does not really work because it asks the wrong question. When you ask the right question (about travelling in an airplane, not jumping out of one) the fact that commercial aviation is conducted without parachutes (though with seat restraints) makes sense.
Surely this work's result is indicative for commercial air travel. The vast, vast majority of passengers will leave the aeroplane when it is stationary and they will drop far less than half a meter when disembarking, circumstances in which the parachute was ineffective. So that's not the wrong question at all.
The ideal satire ought to bring up what we should do, ridiculing it. For example, the famous "A Modest Proposal" says that we shouldn't consider taxing the wealthy, particularly the people who make most money from Ireland while not even living there. It insists you couldn't even begin to tackle the problem with taxation or other measures, whereas eating babies would solve the problem entirely.
Is it typical for clinical trials to completely miss the point like this? I’d like to read a followup meta-analysis that focuses on the actual physiology — “Parachute use to prevent death … when hitting the ground”.
I don't think it's missing the point, but you do want to be aware of the context a little bit. Randomized controlled trials are often held up as the gold standard of evidence, and they are good when it is possible to carry them out, but the point of the linked study is that the kinds of treatments it is possible and ethical to perform RCTs for limit their applicability. In this case, it would be unethical and criminal to get volunteers to jump off airplanes at 10,000ft with placebo parachutes, and the kind of RCT it is possible to carry out (this study) does not show parachutes to be effective. Thus no RCT showing that parachutes are effective is possible.
I had Hodgkins Lymphoma when I was younger. Because we've been successfully treating Hodgkins for so long we don't have RCTs for the basic idea of the treatment. We can observe that people who choose not to be treated tend to die (but that might be for some other reason, they weren't randomised) and we have modern studies comparing one therapy to another as it has advanced, e.g. I got ABVD, I'm not sure what the current gold standard is - but it would be unethical to run the basic RCT now. "We're going to randomly kill some people just to prove that we were right".
Perhaps it was intended as part of the satire. A whole new set of trials will be needed, from scratch, to investigate deaths caused by base jumping from cliff tops!
.3 seconds after jumping from a high altitude plane there was no colleration to parachute usage wrt injuries, ramdomized trial confirms. (Double blind study awaiting approval)
To the best of my knowledge that hasn't been tried. Among other things, fatal crashes typically occur at speeds over 2x those of car crashes. There's no meaningful ability to emergency brake before a crash, and many small airplanes don't have the weight budget to get the things that help you survive a crash, like crumple zones.
An airbag won't do much if you're still going to be squeezed between the engine and tail of the plane at 180 mph.
Well, if you are in the middle of a 737 you have quite some crumple zone. And you can probably see the crash coming in advance, so plenty of time to inflate your airbag suit/chamber.
Hey, I wear a parachute to work every day and it's kept me alive this far! Granted, I work in a ground-floor office, but you can't prove that the parachute isn't what saved me.
That doesn't prove anything. What it does clearly prove, though, is that the Earth is flat. Think about it: have you ever successfully landed with a parachute on a spherical object like a soccer ball or large marble? See?
I think the paper is a satire on "evidence-based medicine" - a framework which insists upon randomized controlled trials as the primary basis for medical decisions.
Notably this is exactly such a trial, while also absolutely irrelevant to the question of needing a parachute or not (the trial planes were all on the ground at the time of the jumps).
It's a follow-up to a previous satire, which questioned whether or not we need RCTs for time-tested, common-sense interventions. The original paper suggested that the most die-hard EBM fans organize and participate in a RCT of the parachute. Several years later, tongue planted firmly in cheek, they did.
funny enough I actually use this article and others in a project when I teach stats.
This article is part of what is known as the 'BMJ Christmas Edition' which gone on for quite a while[0]. it is meant to be at the same time scientifically serious and practically silly [1]. Its...basically satire, but satire as specific scientific commentary. In this case, about the details and clarity of scientific reporting and the claims made from data.
There have been a couple notable problems...One is that satire and human don't always generalize culturally. The second problem though became much more insidious with the electronic shift in publishing. If you access a Christmas edition article electronically, nothing clearly indicates that anything is different about these articles. Nothing indicates 'Christmas edition'. Nothing indicates that they are meant to be light hearted scientific commentary. So they get cited...both as satire and as not satire. Either for cultural reasons or because people aren't being rigorous.
1: https://johnmjennings.com/can-a-mouse-survive-a-fall-from-a-...