Hacker News Reader

Comment by rdlecler1 | Hacker News Reader

The key questions that needs to be addressed, is how homogenous or heterogenous are blood markers at the level they want to measure at? The machines may be reporting results accurately, but a pin-prick-sized droplet may simply not be a large enough sample size for the tests they want to run.

View on HN

There was a recent publication on this subject in the American Journal of Clinical Pathology (http://news.rice.edu/2015/11/17/blood-test-results-vary-from...):

> Bond and Richards-Kortum found that averaging the results of the droplet tests could produce results that were on par with venous blood tests, but tests on six to nine drops blood were needed to achieve consistent results.

>prick-sized droplet may simply not be a large enough sample size for the tests they want to run

This was always the main criticism of Theranos. Finger-stick blood tests have never been considered reliable for clinical diagnostic tests because you don't get enough blood and the blood you do get can be contaminated.

How does the blood become contaminated?

Stuff on your fingertips.

For one thing, it's almost impossible to get your skin perfectly clean. Any test that uses a single drop of blood will be very sensitive to even the slightest amount of soil or contamination on the fingertip where the drop is extracted.

Couldn't this be alleviated with a finger prick method that prevents the droplet from sitting on the surface of the skin before it is collected? (i.e. a tiny needle that reaches below the epidermis to extract a drop of blood)

The needle breaks the skin and moves through any contaminants that were on the skin's surface--potentially mixing the contaminants with the sample regardless of whether the blood sits on the surface of the skin or not.

Isn't it common practice to wipe the skin with a disenfectant to prevent this very thing? A 60-70% alcohol-based solution will kill the vast majority of germs on the skin.

Contaminants don't have to be alive to confuse results. Also many contaminants we're talking about were never alive in the first place.

I wonder if you could apply some substance on the skin or needle so that after the prick it doesn't heal up instantly? Keeps the blood flowing for a few seconds.

Reply on news.ycombinator.com