Theranos is the ultimate fake-it-till-you-make-it story. The company and its founder went from being under the spotlights of silicon valley to being charged with fraud. Investors have been pouring billions of dollars into the company, but it turned out that the whole thing is actually a scam.

It's very surprising to me that it took so long to find out that it's a huge scam, especially knowing that laboratory diagnostic professionals were involved. As a lab person myself, I would like to provide some technical details to explain why the initial goal was not achievable.

Theranos product, the Edison machine, was supposed to have 2 main characteristics: Be able to perform analysis of hundreds of blood parameters on one single machine, small enough to be placed on a desk, and with a sample as small as a blood drop. Each one of these constraints taken separately, is a huge challenge, and the combination of the 2 the whole thing impossible, and I will try to explain why.

1. Let's start by the size of the machine. When you do many parameters, you need many different techniques for the analysis. Regular laboratory analyzers are in most cases dedicated to one or two analytical methods, and it's already very high technology, so bringing the size down will inevitably lead to compromising on quality. In addition, when you analyze many samples from many different people, you need an efficient washing system, because you need to eliminate all the residues from the previous reaction so that you can start a new reaction on clean hardware. Efficient washing systems need vacuum pumps, valves and other hydraulic systems, and this takes some space, and quite a lot of it. You may be able to obtain quality results using a single small machine but for just one parameter, maybe few parameters, but certainly not hundreds.

2. The size of the sample: Probably the most important limitation. In the lab testing we talk a lot about the importance of the preanalytical phase, because if the sample is not processed correctly, the result is already wrong, even before the analysis is even performed. The quality of the sample is the main prerequisite for the quality of the result. And as it's said in the industry: no result is better than a wrong result. A blood drop has two problems that make it extremely unreliable: Easy evaporation, and easy contamination. Regular blood tests require a sample of about 4ml of blood, which is already a relatively small volume, and the evaporation and contamination problems are already challenging to manage at this volume. With a blood drop, it's a nightmare.

I think that the whole project would have never been started in the first place if the founder and her associates had taken some time to study the current technologies used in the clinical laboratories. A lot of improvements have been made over the last decades, however sample quality still remains a challenge today.

Sticking to the idea of replacing existing technology was a mistake, but when it comes to one or just a few specific parameters, this can actually be very helpful. Creating small machines that can for example give some kind of scores in an emergency context can be very useful.