4 ms·

I've been remarking (to myself, mostly) about the touted detection rate using these test kits... The process is rife with opportunities to bungle the test. At

I've been remarking (to myself, mostly) about the touted detection rate using these test kits... The process is rife with opportunities to bungle the test. At least in Canada: 1. The buffer must be added to the test vial, or exactly 10 drops, depending on which kit you get. Both sets of instructions are printed, and it's not apparent which one you should follow. 2. Swab. Two ways are indicated. If you don't go deep enough (and who would?), maybe the test doesn't work? 3. Place swab into buffer. Wait 2 minutes. Try not to forget it when you get distracted and make a cuppa or something. 4. Drop exactly three drops into the test well. Don't miss. It's a wonder positives come back at all.

gus_massa 5y ago
Boolean tests are not very sensitive to these details, specially if they are designed to be used by non professionals. > 4. Drop exactly three drops into the test well. Don't miss. That probably means between 1 and 5. But 1 is too low, and there is a risk of messing of adding a bubble. Probably 2 is the smaller sensible value. > 3. Place swab into buffer. Wait 2 minutes. Try not to forget it when you get distracted and make a cuppa or something. Probably between 1 and 5 minutes are fine. Perhaps 5 is better, but the risk of distraction is too high. > 2. Swab. Two ways are indicated. If you don't go deep enough (and who would?), maybe the test doesn't work? In discussions with my friends, we got very different stories from the nurses administering the test. Some said that it must be very deep until it hurt, some said that it's not necessary. I think that too shallow is bad, but some people is crazy and go to the other extreme. 1. The buffer must be added to the test vial, or exactly 10 drops, depending on which kit you get. Both sets of instructions are printed, and it's not apparent which one you should follow. Also, probably anything between 5 and 20 is fine. The test is probably sensitive to other factors like temperature. But let's assume something between 10°C and 30°. Sunlight? Probably direct sunlight is bad. The standers instructions are an oversimplification. Some foolproof recipe that any moron can follow and ensure that the test is correct most of the time. Try to follow the indications anyway, but they have some room for mistakes. It's common in many applications were there are a lot of unknown factors. For example transistors amplify the current like x200 times. But the process to build them is not very reliable and in the same lot you get some with x100 and other with x400 and any number in between. So the trick is to put them in a circuit that only makes a x10 amplification and will work with any multiplication of the transistor that is greater than x100. You are not using all the amplification potencial of the transistor, but you get a reliable design. Another case are the wires for elevators. You calculate the weight it must carry, and then buy a wire that can lift x4 or x8 time that weight. And you put 4 wires like that, not only one. This is mandated by law here in Argentina and there are similar rules in other countries. [Perhaps I'm misremembering some number, check he official regulation before installing an elevator.] The sistem is not as efficient as possible, but you are covered in case the wires have some defect, or someone overload the elevator and other stuff.
GuB-42 5y ago
I think that these tests are way more tolerant that instructions make it. It is just that there is not much downside in being overly precise. Even if the tests can work if you add between 1 and 100 drops, if you are told to add between 1 and 100 drops, what are you going to do? 1? 10? 50? 100? open the vial and eyeball it? It creates unnecessary confusion. So you just say "add 10 drops", not because you need exactly that many, but because it is the instruction that will make it the least likely to screw up. It is the same for cooking recipes. The two ways to not screw up is to follow the recipe to the letter, or to understand the underlying idea. The latter requires some experience in cooking. Covid self tests are intended for people with no experience, so, they make you follow the recipe to the letter.

gus_massa 5y ago

Boolean tests are not very sensitive to these details, specially if they are designed to be used by non professionals. > 4. Drop exactly three drops into the test well. Don't miss. That probably means between 1 and 5. But 1 is too low, and there is a risk of messing of adding a bubble. Probably 2 is the smaller sensible value. > 3. Place swab into buffer. Wait 2 minutes. Try not to forget it when you get distracted and make a cuppa or something. Probably between 1 and 5 minutes are fine. Perhaps 5 is better, but the risk of distraction is too high. > 2. Swab. Two ways are indicated. If you don't go deep enough (and who would?), maybe the test doesn't work? In discussions with my friends, we got very different stories from the nurses administering the test. Some said that it must be very deep until it hurt, some said that it's not necessary. I think that too shallow is bad, but some people is crazy and go to the other extreme. 1. The buffer must be added to the test vial, or exactly 10 drops, depending on which kit you get. Both sets of instructions are printed, and it's not apparent which one you should follow. Also, probably anything between 5 and 20 is fine. The test is probably sensitive to other factors like temperature. But let's assume something between 10°C and 30°. Sunlight? Probably direct sunlight is bad. The standers instructions are an oversimplification. Some foolproof recipe that any moron can follow and ensure that the test is correct most of the time. Try to follow the indications anyway, but they have some room for mistakes. It's common in many applications were there are a lot of unknown factors. For example transistors amplify the current like x200 times. But the process to build them is not very reliable and in the same lot you get some with x100 and other with x400 and any number in between. So the trick is to put them in a circuit that only makes a x10 amplification and will work with any multiplication of the transistor that is greater than x100. You are not using all the amplification potencial of the transistor, but you get a reliable design. Another case are the wires for elevators. You calculate the weight it must carry, and then buy a wire that can lift x4 or x8 time that weight. And you put 4 wires like that, not only one. This is mandated by law here in Argentina and there are similar rules in other countries. [Perhaps I'm misremembering some number, check he official regulation before installing an elevator.] The sistem is not as efficient as possible, but you are covered in case the wires have some defect, or someone overload the elevator and other stuff.

GuB-42 5y ago

I think that these tests are way more tolerant that instructions make it. It is just that there is not much downside in being overly precise. Even if the tests can work if you add between 1 and 100 drops, if you are told to add between 1 and 100 drops, what are you going to do? 1? 10? 50? 100? open the vial and eyeball it? It creates unnecessary confusion. So you just say "add 10 drops", not because you need exactly that many, but because it is the instruction that will make it the least likely to screw up. It is the same for cooking recipes. The two ways to not screw up is to follow the recipe to the letter, or to understand the underlying idea. The latter requires some experience in cooking. Covid self tests are intended for people with no experience, so, they make you follow the recipe to the letter.