I'm not too sure why this is supposed to be surprising. When chemical reactions are involved, you often need chemicals.
A significant portion of digital meters used for servicing swimming pools require servicing/replacement of strips/chemicals (reagents) within the device.
Yes, there are also digital meters that work without chemicals per se, and have special electrodes that measure electric current. However, this approach quite simply doesn't work for all chemical reactions.
It's surprising because the electronics are essentially useless. Most people would expect the electronic version would have some type of advantage, because those electronics cost extra money to produce.
On a regular stripe test you are responsible to conduct the experiment and interpret the outcome which can be intimidating.
"Is there a line? I think I see a line? But it's fainter from the last time? What does it mean?"
This is replaced by an algorithm that's written by people who know what are they doing to save you from analysing a chemical reaction and have a straight answer.
edit: Just Google for "Is this pregnancy test positive" to see how many people are having trouble interpreting the results.
Ironically, on this example the commenters claim that the inaccurate one is the digital one and the tester confirms. Anyway, the point stands: It is hard to interpret the outcome of chemical reactions.
Your link actually shows you why there is no UX advantage.
OP is only two cycles into their pregnancy. A paper test can pick up on this tiny HCG measurement and show something. A digital test cannot because it's going to pick a higher threshold if it only wants to give you a binary result. A digital test will tell you that you're pregnant when you're far enough along for a paper test to be obvious.
There are so many "can someone interpret this paper test?" questions not because paper is inferior but because they are trying to interpret results 0-2 months into their pregnancy and paper can actually give you a faint reading here.
A digital test could support that by showing "possibly pregnant" which is back to the same UX, and showing "not pregnant" during this time isn't much of a win either.
This is correct except the 0-2 months. Even if you accept dating a pregnancy as starting at the start of your previous cycle, these tests will be positive before your next cycle would have started. The typical “is this a positive test” is around 8-10 days post ovulation and a few days before their period is due. They’d get a definitive test 1-3 days later. So the period of “it’s positive but I can’t be sure” is something like 0.9-0.97 cycles into a pregnancy (again, with the weird dating from your last period’s start and complicated by the fact that most people don’t know ovulation dates & cycle lengths frequently vary).
A digital test can be tuned to a better sensitivity.
The UX advantage comes from giving a definitive answer. The answer can change at later times when the body starts giving stronger signals. You can simply indicate that you need to wait X number of days to detect the pregnancy and problem solved.
This is much better UX than learning biology and chemistry to be able to read a test result. What does a faint line means? Probably means you need 4 years degree in a related field to have an educated guess. To have a definitive answer you need a to have a career in it.
Realistically though, it means you will need to ask someone who already have the degree, which means what good is the test for?
The digital test does guide you how to get to the right answer. Yes, if the threshold is too high and you test too early it can give you a false negative but this is a simple problem that can be solved by recommending a second test in few days.
It should just show the value and have the definition of them written on the box, UI should not be a concern at all if you have to have electronics in this case.
Yeah, sure if you know what you are doing observing chemical reaction can be much better. It's like the difference between consumer tools and industrial tools.
Full manual espresso machines are superior from the perspective of a coffee nerd or professional. Nespresso machines with capsules are superior from the perspective of the person who wants the coffee but doesn't fancy learning making one.
It simply means that the paper is not as good as digital from UX perspective. This doesn’t mean that the paper is unusable but obviously creates some confusion that the digital doesn’t.
The UX in this case is bad. Unnecessary and expensive fluff that obfuscates reality and gives a false sense of certainty to the user.
> On a regular stripe test you are responsible to conduct the experiment and interpret the outcome
The part of conducting the experiment is identical, I think, regardless of the type of test. You pee on it.
Interpreting the results is not an issue. You simply need to observe with your eyes (a product of billions of years of evolution) if you have || or |+.
> save you from analysing a chemical reaction and have a straight answer
Again, you don't need to analyze a chemical reaction any more than the device does. The conclusion of the analysis is summed up in [1]. The whole point of the digital part is to read the lines on the paper. Whatever heuristics went into deciding how much reflected light should be interpreted as a | or a + are just as likely to be mistaken as a human eye reading would be.
Well...if you’re 20, not at all ready to become a parent, you made a mistake and now you see a faint line-ish line is it really a line is it not just the lighting huh? Yeah then I can imagine being intimidated.
This sounds more like imaginary pregnancy test then real one. The real ones dont do faint line-ish.
Also, for christ people, 16 years old were administrating these tests for years now, without there being massive social problem of confused woman. They just somehow managed to do them, even if their school results were very bad. So trying to claim they are somehow difficult to interpret is absurd.
It looks to me that the digital test just makes the situation worse by potentially giving a false negative due to the sensitivity threshold. In many cases being pregnant and not knowing it is worse than the converse. So the false positive may be better than a false negative.
The "analog" version is capable of displaying a value marginally above 0 and lets your eyes and brain decide or at least raise a flag. The digital test replaces your eyes with crappy photodiodes, and your (even 16 year old) brain with a crappy controller that just looks at a high threshold and decides "below=0".
I see people here saying that a digital is better because it reduces confusion. But it does so by discarding data. I'd take the debatable test every time, especially when for the money a digital one costs you can get multiple paper tests to do repeated checks.
And the majority of those are inconclusive results which should be reported by digital as ... inconclusive. Removal of doubt here can cost you precious time.
I don’t have enough close female friends to hear these worries from, but I’ve seen it as a trope in movies and TV shows many times where the woman tries to interpret the strip, gets it wrong, etc
But they also sell for extra money. It's less about building a better device and more about marketing. It's playing on the fact that people assume something electronic and more expensive will be more accurate.
People will make claims about UX et al but they're more a side effect from companies marketing a pseudo-digital device to gullible customers rather than the original goal of the device. And I don't mean "gullible" in a negative way because I've fallen for this trick many times in the past too. If you don't know the hardware (and most wont) then you can't be blamed for trusting the marketing.
Ultimately though, if a digital readout reassures potential mums-to-be more than a paper strip then at least this device does serve a useful purpose beyond tricking people for a few extra dollars.
It's not surprising to me, but it is something very interesting to learn about. And it makes so much sense once you think about it. Why go through the trouble of finding a legitimate technical solution when something super simple like this works and is easier + cheaper to implement. It actually seems like a perfect fit for "hacker spirit."
But then, why couldn't you simply use the chemicals without the electronic parts?
I could imagine some reasons for pool monitoring systems - e.g., you might want to have the device perform continuous measurements and alert you if the results indicate problems - but that is not a case for a pregnancy test: People were perfectly capable to use chemicals-only tests.
This kind of makes sense in a strange way. Inexpensive red LED's have relatively narrow spectral line half width (~20nm) and you can set detection thresholds against measured values. This is much better than "I think I saw something in a dimly lit bathroom". This means that in the readout part (UI) it is hard to get wrong results ("Is it one or two?").
Then again it is a wasteful and overpriced product that can leave wrong impression of superiority to the customer.
Going over the conversation here I'm a bit surprised by how many people seem to prefer an unnecessarily complicated and obfuscating solution to the straight-forward one.
The Twitter thread clearly points out that the electronics serve strictly as an intermediary layer between the paper test and the user. They just binarize the paper reading. This in turn gives a false psychological sense of security ("it's the computer that says it, not me"), without any credible benefit.
Why should I trust the people who calibrated the electronic device more than my own eyes?
More generally, why are people so eager to discount the reliability of their own senses? You have eyes and a mind which are the product of billions of years of evolutionary refinement. If you're not sure whether it's || or |+, pee again on a different test and compare; or wait a few days. If the lighting in your bathroom is bad, take the test in natural light or a better lit room.
My personal instinct is to rely first and foremost on my own mind and senses. It's of course important to defer to expert knowledge, but we need to use our critical thinking when deciding if something is the result of expert knowledge, or simply a marketing gimmick.
I actually saw this last night and initially had the same reaction but had some time to think on it. Remember, for a lot of people taking a pregnancy test is incredibly stressful. Reducing the complexity of reading by outputting a less ambiguous result has a lot of value to some people.
This exactly. I had to convince my wife not to doubt herself. “Yes that’s a line. Instructions say that if anything appears even if faint then you’re pregnant. No we don’t need to double check. Wahoo you’re pregnant!!”
This is one of best, perhaps the singularly the best, times in life to smile and go along with it. (She has a PhD in neuroscience; entering some kind of Statistics Olympics wasn't going to improve anything in my life at that moment.)
Yes, someone who’s made it as far in a relationship as intentional pregnancy should “save” their relationship by suggesting remedial education to their partner...
Society wants -- through its observed behavior -- that very much. Is it good from an intellectual perspective? Likely not, though this is a very low impact scenario compared to mass plastic convenience packaging for near daily use items. YMMV.
Not only comfort. A lot of factors early on in pregnancy affect the health of the unborn child. Knowing when your pregnant will help you stop consumption of e.g. alcohol.
It could ultimately generate less waste, if it prevents enough misreads. The cost incurred by someone misreading a positive result as negative and having to abort later on is much higher than the premium of the digital test. I'm not sure if the numbers work out, though.
> Why should I trust the people who calibrated the electronic device more than my own eyes?
Do you as a layperson know how sensitive the test strip is, or how blue it needs to turn to have a reasonable guarantee against false positives? It seems like the engineers of the device would have a much better perspective on those things.
I know that whoever made the test, if it's a reasonably competent enterprise, made it with human capacities in mind. I don't think the threshold between positive and negative is easier to tell for an engineer than for anybody else with reasonable eyesight.
I might change my mind if it would surface that the paper tests used in the electronic devices were somehow different from those available to the general public and they were tailor made for the sensors and diodes in order to make the results more reliable. I doubt this is the case, however.
> I know that whoever made the test, if it's a reasonably competent enterprise, made it with human capacities in mind.
Are you sure they weren't just constrained by technological limitations when they designed the paper test? Maybe the reason the digital test exists is that a competent enterprise wanted to adapt the paper technology to better fit human capacities?
Maybe, although I doubt it. It would mean that they put in extra research to, for example, plot an ROC curve and see what the optimal threshold for the test would be.
Even if that was the case, you can still frame the question in terms of cost/benefits. The electronic test is, let's say, 10 times more expensive than a basic one. Is it more reliable than taking 10 basic tests (maybe at different times) and averaging the results? Probably not.
> Even if that was the case, you can still frame the question in terms of cost/benefits. The electronic test is, let's say, 10 times more expensive than a basic one. Is it more reliable than taking 10 basic tests (maybe at different times) and averaging the results? Probably not.
Food that costs ten times more usually doesn't taste ten times as good. Perhaps twice as good.
This is the beauty of market options. We can both have our cake and eat it too. No need to judge others and their product choices, just buy what you want.
If you want, you can buy 10 cheaper strips, Try different lighting and times of day, then take the numeric average. I can buy a more expensive digital strip simply because I like the box's color.
The levels of hCG are two orders of magnitude larger post conception and rise to five orders of magnitude a month post conception. The antibodies the test uses are extremely selective.
>I'm a bit surprised by how many people seem to prefer an unnecessarily complicated and obfuscating solution to the straight-forward one.
This makes many presumptions about the users and whether to them this design is unnecessarily complicated or obfuscating. Beyond whether it actually is for them or not, advertising works or people wouldn't do it...the old tech has been labeled and messaged and reinforced for years as confusing to the point that it can become a normative perception whether or not it is true.
I'm a kind of color blind where I only see 150,000 colors. I have trouble distinguishing where purple ends and blue starts. Same goes for a bunch of other color pairs. I didn't figure it out for decades, and before I did I remember some heated arguments where it was me vs everyone else on "this shelf is purple" vs "that shelf is blue, not purple".
The day I figured it out those memories started coming back and it all clicked. I stopped trusting myself after that. I knew I didn't trust myself anymore when my 3 year old son would reference the bus I used to catch by it's color and I didn't know whether he was wrong or I was, so I just kept my mouth shut cos there was a better chance he was right.
> Going over the conversation here I'm a bit surprised by how many people seem to prefer an unnecessarily complicated and obfuscating solution to the straight-forward one.
Welcome to modern IT, where complexity fetishests have ruined everything.
I saw the beginnings of this on Twitter last night; not surprised it wound up here. If anyone here isn’t already following @foone, I highly recommend it. Their teardowns are fascinating.
One of the more interesting comments:
“This thing [the chip in the device] is probably faster at number crunching and basic I/O than the CPU used in the original IBM PC, and this one is in something you pee on and throw away.”
(Unfortunately, “...there's no reprogramming this chip to run other programs. We can't run Doom on this pregnancy tester, I'm sorry.”)
Pregnancy/Motherhood groups are swarmed with the same "can you read this, I think I see a faint line" + photo question every single day because of the cheap tests. The strip results are NOT easy to interpret, because you may still see a faint line on the paper when not pregnant, and it's kind of an important emotional question as some couples have been trying to get pregnant for years. Your OB does you a mercy if they tell you "just buy the digital one".
I think you're right here. When you're emotionally invested in the result (you want to be pregnant, or you don't want to be), then a less than certain result can force you to agonise over whether you're doing it right, or trust your own senses.
A digital readout (assuming is pretty accurate) can remove this ambiguity.
The other benefit is for anyone who has difficulty with colour perception, or is in a poorly lit environment. I assume the device can read the stip in pure blackness leaving little enivornmental light to cause confusion.
But when it's all said and one, I bet this was created by the marketing department.
Well yes but the digital test does not give your better info. They translate a non binary result to a binary one.
What I did when "we" were doing these tests (got 2 types both paper stips, one high sensitivity and one of the cheapest one) simply compare a unused strip and the one with pee.
It's not hard and the manual in the ones I got were really good. Like color pictures and lots of examples.
I’d think these things have minimum standards they must meet and so the cheapest test is just as good as the more marketed (and more expensive) one. Am I missing something?
kind of an important emotional question as some couples have been trying to get pregnant for years
And for some just the opposite... like 19 year old me decades ago -- ambiguous test result, and no 24 hour drugstores at the time so led to a sleepless night for both of us until we could get a second test the next day. It was negative. Would have much rather had an unambiguous POS/NEG result.
If it was ambiguous, these digital tests may have told you it was negative even when it would be positive a day or two later. You might have slept better but it wouldn’t have been based on accurate data.
The second test had more clear instructions about "read results after X minutes, ignore any changes after that time". The electronic test wouldn't have had a visible line after 10 minutes that scares young college students when they think they can see the shadow of a second line if they hold it to the light the right way.
Perhaps there should be more examples of positive and negative readings in the paper test’s leaflet? Including photos like those that get posted on forums. I know this wouldn’t solve all cases but seems like it could be helpful for some.
Or given that a photo of a piece of paper is hard to interpret too, possibly the literature could include an actual piece of test strip to compare... possibly an untreated one, unless that changes how it looks.
While the waste is lamentable, perhaps it's also worth considering this as a UI/UX problem. How many people mis-read the paper test vs how much extra error % does the circuitry introduce?
Not for this test. It's a lateral flow assay [1]. The urine is absorbed by the strip and then moves along with capillary action. If there is hCG (human chorionic gonadotropin) in the urine, then the it binds to linked-antibodies (often linked to gold atoms) and drags them along, then those antibodies bind to the actual assay and a control assay (antibodies to the antibodies).
I supposed you could make the line words, but they'd be pretty small.
"The word "not" is faintly showing, what does that mean?"
It it was trivial to show a single clear indicator word, they could just show a clear indicator line, then there'd be no ambiguity. It's not the second line that confuses people, it's figuring out if that line is really visible enough.
Do you? The entire point of the machine is to read the test to the limits of what the test strip can detect.
It the digital reader can't detect the line, that should be a limitation of the test strip, not the automation. And it should do better than an untrained human.
I suspect it is difficult to get the 'raw' output from a cell-phone camera so you are dependent on the phone's processing (contrast enhancement etc), which might vary between models. Not impossible (e.g. having control lines) but needs thinking about.
"Pregnancy tests are 99% accurate in the lab, 75% accurate in the wild due to misreads- mistakes which are highly dependent on education and socioeconomic status. No, it is not stupid or wasteful to use a hardware interface to help women with this."
That thread (by someone injecting racism) says nothing about the accuracy of machine tests in the wild, nor whether the errors are false positive or false negative.
So It's basically like the blood sugar testers diabetics use, except they want you to buy a new reader with every strip.
There is now an opportunity for one of the manufacturers of pregnancy tests to make a reusable reader. The add could be like "We know you're going to test more than once, we won't force you to buy a reader every time." or "instead of asking the strangers if they see a line (cut to forum posts of tests)
, let {product name} be your second opinion" Or even "Don't wait until you've missing your period, test every week for an entire year for less than the cost of 3 of the competitors digital tests"
Pregnancy tests and blood sugar tests have pretty different usage patterns. Blood sugar tests will (should) be used every day by diabetics; the cost per test matters a lot. Women on the other hand go through pregnancy tests much less frequently, and care much more about accuracy than cost. The impact of a false positive or negative from a pregnancy test can be massive, and even the perception of increased accuracy is good marketing.
The issue here is that paper tests require the user to read and properly understand instructions. Teenagers in a panic can and will misread and misunderstand, as people often do when under duress.
It may seem "pointless", but making the tests more user friendly and less prone to user error is super important.
Here is a paper showing how these tests have 99% sensitivity in the lab, but this drops to 91% when the test is read in the field by a volunteer, and down to 75% when the test is read by the subject themselves:
Moreover, if it's such that even a trace of a line indicates a high confidence of pregnancy, users may misinterpret a faint result as somehow meaning "kinda" pregnant.
I wouldn't say super important. I'll probably get down voted for my opinion, but if we had proper sex education for everyone and valued women's health care than this wouldn't be an issue.
I don't think that sex education and accessible abortions stops people from being emotionally invested in finding out if they are pregnant. This can be a stressful question whether you want a child, don't want a child, or don't know!
Teenagers were specifically mentioned in the comment. Good sex education helps because reproduction is now not some mysterious process. They will then be able to make good choices when they decide to sex. Having access to abortion services helps because having sex shouldn't ruin someones life.
I don't think that anyone was arguing those points. simply that they wont eliminate the stress associated with test taking, which is the purpose of this device.
And then it sits in a landfill for eternity! Imagine how many millions of tiny computers (and batteries - oof) are just sitting in dumps right now. That's the most egregious part for me: this test is exactly the same as the analog one, but it generates eWaste.
Not a native speaker; I think I understand your point, though I fail to see a distinction between these two words: if the output is a number to the nth decimal, but it is not "accurate" (ie. the numbers are actually wrong), can it still be said to be "precise" ?
Precise means lots of fine-grained information; accurate means close to the truth. If the weather is 20° exactly and one thermometer reads 20.2° and another 20.562° the first can be described as more accurate and the latter as more precise. Hence the expression, “precisely wrong”.
My 3rd world country has pregnancy tests[1] available in almost all vending machines, but unfortunately they're the paper kind. Obviously we managed to get them evenly distributed, but utterly failed to provide the future.
[1] For the convenience of both those hoping for positive results and those hoping for negative. I'm guessing there are more of the latter than the former, because they're invariably located in the slot next to the condoms.
And it was selling at a loss as well; the device was ridiculously overbuilt, see this teardown by everyone's favorite canadian dad: https://www.youtube.com/watch?v=_Cp-BGQfpHQ
I just checked on cvs.com, a 2 pack of CVS branded "blue line" style pregnancy tests cost $13.99, a 2 pack of "clear blue" digital tests costs $17.79.
You can buy bulk strips for around 50 cents each, but many people would rather pay for the convenience of peeing on a stick than peeing in a cup and them dipping a test strip in it.
Many people are squeamish enough about their own urine to not want to hold a 2" long test strip in the stream, a 6" plastic device gives more clearance.
But in any case, test strips vary in how they work - some require a couple drops of urine in a specific spot on the strip, some say to dip in the urine cup until a dye mark appears. Is holding the test strip in the stream going to give the same result? I don't know.
I wonder if there is an opportunity to make a mobile app that interpreted the results from these tests in a similar manner and sell tests that are app enabled.
You'd have to make sure the diodes were exactly lined up with the strip in order to avoid false negatives. That doesn't seem like a risk worth taking to save a few bucks. Maybe some sort of cartridge system could be developed.
And are people supposed to tear it down and dispose of the battery separately? (the battery is hazardous and has markings indicating it is not to be thrown away).
I guarantee literally nobody removes the batteries before throwing these out. It's a shame, really. I'm sure there are bulkier sources of eWaste, but still. Someone pees on a computer for 30 seconds and then it gets thrown away forever.
No one is digging through trash looking for tiny batteries embedded in disposable electronics. The dump truck gets unloaded straight on top of a landfill and bulldozers spread it around in the majority of the US.
I've noticed companies that build medical devices have followed the lead of big tech companies like Apple. First it was devices with non-replaceable batteries, and now: subscription services.
I didn't find it mentioned here or in the Twitter thread, but the digital test also keeps track of previous results and alerts you when the scores are ramping up. Of course you could do this manually by keeping all the strips and comparing them but the device makes it a bit easier.
A significant portion of digital meters used for servicing swimming pools require servicing/replacement of strips/chemicals (reagents) within the device.
Yes, there are also digital meters that work without chemicals per se, and have special electrodes that measure electric current. However, this approach quite simply doesn't work for all chemical reactions.