Thanks, agree totally. The reason engineering (and most natural sciences) do not use RCTs routinely is our ability to directly measure what we’re studying. An RCT is a very poor method in any framing where the study parameters are quantifiable.
Like Loeb 2022, this study also paid no attention to the ventilation status of the facility. Even with an RCT the question of detection thresholds is critical to establish. If you have 6+ ACH through HVAC, the level of transmission may already be lowered below the detection threshold of the study method. This too is a basic criteria in any engineering study - if your signal is undetectable because it’s below the noise floor, your study only tests the sensitivity of detection.
Thank you for this. I've seen a lot of awful COVID minimizers sharing this single study as authoritative while blatantly ignoring or misrepresenting its actual content and limitations. It's great to see public pushback on this from people who actually care and know what they're talking about.
This looks like a debate about intention to treat analysis and per-protocol analysis. Both provide information but for what to invest in financially in a real world scenario the intention to treat analysis is most helpful. Why invest in interventions to stop illnesses if they are not going to used because they aren't tolerated?
People don’t buy and use air filters to reduce pathogens per cubic cm. They do so due to a desire to reduce actual infections. In the course of their actual lives.
Look up the definition of “surrogate outcome” cuz that’s what you’re worshipping here.
If your scenario for showing the benefit of hepa filters involves locking nursing home residents in a room 24/7, that really ought to tell you all you need to know. And you should look up “external validity” while ur at it.
I would like to suggest you are viewing the testing of this technology through the wrong lens, which is something I note several people of engineering backgrounds have done
For the proposed expected effect and implementation of these technologies, RCTs are absolutely the right way to test them
I have written a more comprehensive explanation below
Thanks. We all want to know if these things prevent sickness in real world settings but RCTs can’t tell us that. That’s the point. It’s interesting that we’ve never conducted RCTs on refrigeration. We simply measure bacterial growth rates under different temperature regimes and extrapolate a safe food storage temperature from laboratory studies and with the benefit of experience. It works.
Thanks, but I find your thinking really jumbled here. In many ways you actually appear to be agreeing with me. There are two fundamental issues concerning RCTs in this context;
1) The ethical issues (which you have raised)
2) The type of intervention – eg are we dealing with a drug treating individuals or engineered devices/processes such as air cleaning technologies, fridges, water treatment etc
Conducting an RCT on smoking is unethical, as you say, but so is an RCT on HEPA filters, because it exposes the control cohort to known harms. It looks like we can agree on this.
On the second point, HEPA filters are a device that treats a space, not an individual, so an RCT is impractical. Unlike a drug trial, where the dose and duration of the treatment is known and fixed, unless the intervention cohort remains static within the HEPA treated zone, we have no idea of the dose and duration of intervention. We don't know the source of infection. So RCTs for HEPA filtration are unethical, as for smoking, but also completely impractical.
I would suggest you read through my previous articles and comments carefully as these points have been addressed.
Thanks. Your answer demonstrates the struggle with logical reasoning that the medical establishment has got into on this topic. Engineers understand these issues with greater clarity, and in due course I expect the medical and public health establishments will catch up. Let’s agree to disagree and we’ll see where we are with air cleaning technology in five years time.
Thanks, agree totally. The reason engineering (and most natural sciences) do not use RCTs routinely is our ability to directly measure what we’re studying. An RCT is a very poor method in any framing where the study parameters are quantifiable.
Like Loeb 2022, this study also paid no attention to the ventilation status of the facility. Even with an RCT the question of detection thresholds is critical to establish. If you have 6+ ACH through HVAC, the level of transmission may already be lowered below the detection threshold of the study method. This too is a basic criteria in any engineering study - if your signal is undetectable because it’s below the noise floor, your study only tests the sensitivity of detection.
Thank you for this. I've seen a lot of awful COVID minimizers sharing this single study as authoritative while blatantly ignoring or misrepresenting its actual content and limitations. It's great to see public pushback on this from people who actually care and know what they're talking about.
This looks like a debate about intention to treat analysis and per-protocol analysis. Both provide information but for what to invest in financially in a real world scenario the intention to treat analysis is most helpful. Why invest in interventions to stop illnesses if they are not going to used because they aren't tolerated?
This is the correct analysis 👍
Wow you completely miss the point.
People don’t buy and use air filters to reduce pathogens per cubic cm. They do so due to a desire to reduce actual infections. In the course of their actual lives.
Look up the definition of “surrogate outcome” cuz that’s what you’re worshipping here.
If your scenario for showing the benefit of hepa filters involves locking nursing home residents in a room 24/7, that really ought to tell you all you need to know. And you should look up “external validity” while ur at it.
I would like to suggest you are viewing the testing of this technology through the wrong lens, which is something I note several people of engineering backgrounds have done
For the proposed expected effect and implementation of these technologies, RCTs are absolutely the right way to test them
I have written a more comprehensive explanation below
https://alasdairmunro.substack.com/p/engineers-vs-rcts?r=1fhhmw&utm_medium=ios
9/10
Missed opportunity to cite the Rothman Dahly Evidence Pyramid https://zenodo.org/records/18084434
Thanks. We all want to know if these things prevent sickness in real world settings but RCTs can’t tell us that. That’s the point. It’s interesting that we’ve never conducted RCTs on refrigeration. We simply measure bacterial growth rates under different temperature regimes and extrapolate a safe food storage temperature from laboratory studies and with the benefit of experience. It works.
Thanks, but I find your thinking really jumbled here. In many ways you actually appear to be agreeing with me. There are two fundamental issues concerning RCTs in this context;
1) The ethical issues (which you have raised)
2) The type of intervention – eg are we dealing with a drug treating individuals or engineered devices/processes such as air cleaning technologies, fridges, water treatment etc
Conducting an RCT on smoking is unethical, as you say, but so is an RCT on HEPA filters, because it exposes the control cohort to known harms. It looks like we can agree on this.
On the second point, HEPA filters are a device that treats a space, not an individual, so an RCT is impractical. Unlike a drug trial, where the dose and duration of the treatment is known and fixed, unless the intervention cohort remains static within the HEPA treated zone, we have no idea of the dose and duration of intervention. We don't know the source of infection. So RCTs for HEPA filtration are unethical, as for smoking, but also completely impractical.
I would suggest you read through my previous articles and comments carefully as these points have been addressed.
Thanks. Your answer demonstrates the struggle with logical reasoning that the medical establishment has got into on this topic. Engineers understand these issues with greater clarity, and in due course I expect the medical and public health establishments will catch up. Let’s agree to disagree and we’ll see where we are with air cleaning technology in five years time.