4 ms·

I hope I'm not becoming "that guy," but I feel like everyone is still missing the point on Theranos. Even if they hadn't defrauded investors, the company would

I hope I'm not becoming "that guy," but I feel like everyone is still missing the point on Theranos. Even if they hadn't defrauded investors, the company would never have had the impact they claim, because Theranos' entire finger-stick idea just doesn't fit into the lab industry. I spent a bunch of years in labs, working for the tech companies that made the "state of the art machines" that did the actual testing. Theranos wasn't only competing with LabCorp and Quest, they were also competing with the testing machine companies like Beckman-Coulter, Roche, Siemens, Bio-Rad, and many others. So many tests get ordered for a CMP (comprehensive metabolic panel, the most common battery of tests), that you would need like 30 finger sticks to get enough blood. Theranos never clarified if they could do 30 tests on 1 drop, or 1 test per drop. That little misdirect actually matters a lot to the patient who might have to get stuck 20+ times for all the things the doctor needs. I'm quite familiar with the way the state of the art systems work. A modern system is already testing microliters of plasma (your blood goes through a centrifuge) per test. Comparing size of blood draws is misleading. The several tubes that get drawn are for logistical reasons that don't go away if you switch to finger-sticks. For example, the law requires that a lab keep some blood on hand for emergency re-testing if it turns out you have a critical level. Maybe a tech drops your sample. Maybe the machine has a problem right when it's doing your test (happens all the time, no reason it wouldn't happen with the Theranos instrument, had it existed). How many finger sticks are required to keep some extra on hand? Theranos was also misleading about the test time, but that doesn't get mentioned either. They claimed 4-hour test time, but compared it to the 24-hour turn around time of a major lab like LabCorp. In reality, modern systems take 8-15 minutes for most tests, depending on what it is. Large molecules (hormones, etc) and whole-blood tests can take longer, sometimes up to 1.5 hours on older machines. Nothing approaching 4 hours, though. A 4 hours test time is atrociously long. And after all that, you still need a doctor to look at the results. Let's say the dream happened, and we all started going to Walgreens to get Theranos tested. Even if we got 30+ tests on a single finger-stick drop, we'd still need to wait 24-72 hours for the doctor to look at the results and call us. Maybe I'm beating a dead horse at this point, but I think from the very beginning, Theranos didn't have a serious business model. I also would like anyone on HN reading this, who's interested in the medlab space, to understand that there's a lot more going on than just testing on fingerpricks and fraud here. If any of you out there actually do have a revolutionary one-drop testing device, don't try to partner with Walgreens and open your own labs. Just sell/lease it directly to doctors and hospitals.

mirimir 8y ago
Could it have worked with standalone systems, running in hospitals, doctors' offices, etc? That would be impossible with current (or vaguely conceivable) technology, of course. Some years ago, I was seeing a doctor who did urinalysis in house. I figured that he was money grubbing, and maybe insurance companies eventually decided the same.
- CptFribble 8y ago
  It depends. There are Point-Of-Care systems already, usually for things like urinalysis as you mentioned. If they did have a small, accurate, fast (~10 minutes or less per test, definitely not 4 hours) new system that could do multiple tests on a single drop, then yes, putting those directly in doctor's offices and hospitals would be the right idea, as long as you can make the financials make sense for the doctors. Really, there are already small systems for just about any test you might want to do, it's just that economies of scale for a big lab like LabCorp usually make it too comparatively expensive to run your own system, because you need to pay for the system and hire a licensed technologist to operate it. That's not just bureaucracy, either. Most current systems can have lots of weird problems that affect the results, and you need someone who understands both the medical relevance of the test (and normal/abnormal results), as well as some statistics, in order to know when the machine is operating correctly and when it's giving you subtly biased results and might need re-calibration or something. In big cities like NYC, it's more expensive to ship the blood out of the city to LabCorp/Quest's big regional facilities, and the "big" labs inside the city are way smaller (so they cost more). So doctors are more likely to be able to swing operating their own system. Also, some doctors charge way more than normal for their services, particularly IVF/fertility clinics. Those places always have their own machines (their clientele are also more demanding, and would expect results faster than you can get by sending out to LC/Quest).
- thaumasiotes 8y ago
  No, it could not have worked as desired with any system, because it is fundamentally impossible to use a drop of blood to detect chemicals which are present in the circulatory system from which it was drawn, but are not present in the specific drop being tested. Blood is not a homogeneous substance. In order to be satisfied that you can detect substances that are present in the person, you need to extract more than a drop of blood.

mirimir 8y ago

Could it have worked with standalone systems, running in hospitals, doctors' offices, etc? That would be impossible with current (or vaguely conceivable) technology, of course. Some years ago, I was seeing a doctor who did urinalysis in house. I figured that he was money grubbing, and maybe insurance companies eventually decided the same.

CptFribble 8y ago
It depends. There are Point-Of-Care systems already, usually for things like urinalysis as you mentioned. If they did have a small, accurate, fast (~10 minutes or less per test, definitely not 4 hours) new system that could do multiple tests on a single drop, then yes, putting those directly in doctor's offices and hospitals would be the right idea, as long as you can make the financials make sense for the doctors. Really, there are already small systems for just about any test you might want to do, it's just that economies of scale for a big lab like LabCorp usually make it too comparatively expensive to run your own system, because you need to pay for the system and hire a licensed technologist to operate it. That's not just bureaucracy, either. Most current systems can have lots of weird problems that affect the results, and you need someone who understands both the medical relevance of the test (and normal/abnormal results), as well as some statistics, in order to know when the machine is operating correctly and when it's giving you subtly biased results and might need re-calibration or something. In big cities like NYC, it's more expensive to ship the blood out of the city to LabCorp/Quest's big regional facilities, and the "big" labs inside the city are way smaller (so they cost more). So doctors are more likely to be able to swing operating their own system. Also, some doctors charge way more than normal for their services, particularly IVF/fertility clinics. Those places always have their own machines (their clientele are also more demanding, and would expect results faster than you can get by sending out to LC/Quest).
thaumasiotes 8y ago
No, it could not have worked as desired with any system, because it is fundamentally impossible to use a drop of blood to detect chemicals which are present in the circulatory system from which it was drawn, but are not present in the specific drop being tested. Blood is not a homogeneous substance. In order to be satisfied that you can detect substances that are present in the person, you need to extract more than a drop of blood.