4 ms·

Theranos technology is more like typical startup tech. How many bugs did Facebook's feed algorithm suffer before it finally got stable? How many pages of sub-o

Theranos technology is more like typical startup tech. How many bugs did Facebook's feed algorithm suffer before it finally got stable? How many pages of sub-optimal search results has Google returned? How many major security bugs have been found in Windows, Chrome, Linux, etc? Healthcare is one area where many people (particularly Americans) have the absurd idea that everything has to work perfectly. Think of the level of indignation people feel when a surgeon who spent 15 years beyond college doing advanced training (for little pay) makes a small mistake that costs someone's life. Or an OB misjudges something and ends up in a major lawsuit. But for other jobs, mistakes are a part of life. In software we are proud to decrease our bugs through testing and improved techniques and processes, but nobody expects perfection. Even mass market products like TurboTax ship with tons of bugs. Lawyers routinely let innocent people wind up in jail out of pure incompetence. We need to get over it. Theranos is doing something very smart and while it needs some additional refinement, it is pretty clear that doing large scale blood testing/monitoring is the future of diagnostic medicine. The irrationality we have about healthcare being perfect is a) false (mistakes get swept under the rug all the time) and b) counterproductive because it gives us a sense of mastery over our destiny that we simply do not (yet) have. Let's also not forget that medicine is plagued by a very irrational risk aversion, which is why many useful treatments are available around the world but not in the US.

minimaxir 10y ago
> Healthcare is one area where many people (particularly Americans) have the absurd idea that everything has to work perfectly. There's a line between not working perfectly and not working at all.
dimdimdim 10y ago
Not when your bug can kill a person or damage his organs due to wrong medication give because of a wrong test result. http://www.wsj.com/article_email/theranos-voids-two-years-of-edison-blood-test-results-1463616976-lMyQjAxMTI2NTEzOTcxODkxWj http://www.wsj.com/article_email/theranos-voids-two-years-of... There is a reason Healthcare is regulated and required multiple degrees of testing before it can be released to the general public -- imagine every new "health startup" pushing new drugs and then saying they'll "fix a bug" when a couple of folks die. The comparison is immature.
- grandalf 10y ago
  > The comparison is immature Quite the opposite. Health regulations suffer from a number of perverse incentives that economists have studied extensively: Drugs are needlessly delayed in getting to market. Doctors and patients are perfectly capable of assessing the risk/reward of an unproven treatment by looking at scientific studies. The FDA adds a layer of false-security... in spite of its sluggish process, the FDA is often quite wrong and drugs that were approved get taken off the market. For other drugs, they are delayed for several years and much harm is caused. Economists have studied this and found that regulators cause much needless suffering and death by being excessively risk-averse. Consider too illegal drugs. Our regulators criminalize many substances, resulting in black market behavior and significant crime and other negative effects. Here too, backward regulations cause much unnecessary suffering, death (and crime too). For people who do not understand healthcare, the world appears quite black and white. In reality no treatment works all of the time, and no test is without its share of inaccurate results. Doctors are equipped with enough Bayesian reasoning ability to use the available information wisely. Also, a physician should not base any decisions on a single blood test result. That is not the proper way to use such tests. If it were, doctors and medical training would not be necessary. Most tests have a significant probability of false positive and false negative results, so docs are trained to consider the results only in the context of other diagnostic clues.
jeffmould 10y ago
> Healthcare is one area where many people (particularly Americans) have the absurd idea that everything has to work perfectly. Yes, healthcare tech should at work at near perfect. There is an enormous difference between getting poor search results or not seeing photos of your drunk friend from the weekend and determining if you have a life-threatening illness. Healthcare is not an area where you want to be "bug testing" and the outcome can be life or death. I agree that the tech should be explored and probably could be refined over time, but healthcare is an industry where you can't "fake it till you make it."
- grandalf 10y ago
  Well, all tests have some probability of returning a false negative or a false positive. The key is for the physician to know the probability of each and to reason appropriately using the diagnostic methods at his/her disposal.
  - joezydeco 10y ago
    Which means your chemistry is within, and stays within, the variances and probabilities that the FDA reviewed and approved. When you start to slide wildly of those bounds? That's when you get shut down. Where was Theranos?
dimdimdim 10y ago
On a more humorous note - I'd like to see you volunteer as a Beta Tester for Theranos' next release :) :)
- grandalf 10y ago
  I would happily do so. I do not have any health conditions that require testing, however, so I would not be exposing myself to risk by doing so.
  - legostormtroopr 10y ago
    Except your test results could come back falsely positive causing you to get sick from treatment to cure a disease you don't have. See the problem?
    - grandalf 10y ago
      I'd expect my physician to interpret them with the proper expectation. All tests have some probability of a false negative or false positive, and medicine is not so simple as reading a test result and ordering treatment.
      - baudehlo 10y ago
        It shouldn't be, but it is.
DanBC 10y ago
People know that clinicians making their very best efforts will sometimes kill patients. Hopefully there's some process in place for that organisation to learn from the error. Compare that to Theranos, where non-clinicians were deceiving customers with a tech that could not work in order to make money. That kind of behaviour is rightly criticised.
- grandalf 10y ago
  > That kind of behaviour is rightly criticised. Indeed it is. I think there is a big difference between a promising company whose technology has a few bugs and whose salespeople are overzealous and unethical, and a company that is an outright fraud.

minimaxir 10y ago

> Healthcare is one area where many people (particularly Americans) have the absurd idea that everything has to work perfectly. There's a line between not working perfectly and not working at all.

dimdimdim 10y ago

Not when your bug can kill a person or damage his organs due to wrong medication give because of a wrong test result. http://www.wsj.com/article_email/theranos-voids-two-years-of-edison-blood-test-results-1463616976-lMyQjAxMTI2NTEzOTcxODkxWj http://www.wsj.com/article_email/theranos-voids-two-years-of... There is a reason Healthcare is regulated and required multiple degrees of testing before it can be released to the general public -- imagine every new "health startup" pushing new drugs and then saying they'll "fix a bug" when a couple of folks die. The comparison is immature.

grandalf 10y ago
> The comparison is immature Quite the opposite. Health regulations suffer from a number of perverse incentives that economists have studied extensively: Drugs are needlessly delayed in getting to market. Doctors and patients are perfectly capable of assessing the risk/reward of an unproven treatment by looking at scientific studies. The FDA adds a layer of false-security... in spite of its sluggish process, the FDA is often quite wrong and drugs that were approved get taken off the market. For other drugs, they are delayed for several years and much harm is caused. Economists have studied this and found that regulators cause much needless suffering and death by being excessively risk-averse. Consider too illegal drugs. Our regulators criminalize many substances, resulting in black market behavior and significant crime and other negative effects. Here too, backward regulations cause much unnecessary suffering, death (and crime too). For people who do not understand healthcare, the world appears quite black and white. In reality no treatment works all of the time, and no test is without its share of inaccurate results. Doctors are equipped with enough Bayesian reasoning ability to use the available information wisely. Also, a physician should not base any decisions on a single blood test result. That is not the proper way to use such tests. If it were, doctors and medical training would not be necessary. Most tests have a significant probability of false positive and false negative results, so docs are trained to consider the results only in the context of other diagnostic clues.

jeffmould 10y ago

> Healthcare is one area where many people (particularly Americans) have the absurd idea that everything has to work perfectly. Yes, healthcare tech should at work at near perfect. There is an enormous difference between getting poor search results or not seeing photos of your drunk friend from the weekend and determining if you have a life-threatening illness. Healthcare is not an area where you want to be "bug testing" and the outcome can be life or death. I agree that the tech should be explored and probably could be refined over time, but healthcare is an industry where you can't "fake it till you make it."

grandalf 10y ago
Well, all tests have some probability of returning a false negative or a false positive. The key is for the physician to know the probability of each and to reason appropriately using the diagnostic methods at his/her disposal.
- joezydeco 10y ago
  Which means your chemistry is within, and stays within, the variances and probabilities that the FDA reviewed and approved. When you start to slide wildly of those bounds? That's when you get shut down. Where was Theranos?

dimdimdim 10y ago

On a more humorous note - I'd like to see you volunteer as a Beta Tester for Theranos' next release :) :)

grandalf 10y ago
I would happily do so. I do not have any health conditions that require testing, however, so I would not be exposing myself to risk by doing so.
- legostormtroopr 10y ago
  Except your test results could come back falsely positive causing you to get sick from treatment to cure a disease you don't have. See the problem?
  - grandalf 10y ago
    I'd expect my physician to interpret them with the proper expectation. All tests have some probability of a false negative or false positive, and medicine is not so simple as reading a test result and ordering treatment.
    - baudehlo 10y ago
      It shouldn't be, but it is.

DanBC 10y ago

People know that clinicians making their very best efforts will sometimes kill patients. Hopefully there's some process in place for that organisation to learn from the error. Compare that to Theranos, where non-clinicians were deceiving customers with a tech that could not work in order to make money. That kind of behaviour is rightly criticised.

grandalf 10y ago
> That kind of behaviour is rightly criticised. Indeed it is. I think there is a big difference between a promising company whose technology has a few bugs and whose salespeople are overzealous and unethical, and a company that is an outright fraud.