Holmes insisted on using just a prick of blood. Problem: unreproducible results. A miniaturized catridge made storing and using the reagents hard.
Costs. Each cartridge cost +$200 to make and was single use. Holmes even purchased a $2m automated packaging line.
Control of Information. The biochemists and the engineers reported up to Holmes and weren’t encouraged to communicate with each other. Holmes was frustrated by the engineering team’s slow progress.
Holmes convinced Pfizer to try out Theranos 1.0 in Tennessee on terminal cancer patients. Results were sent to Theranos, who analyzed them and forwarded them to Pfizer. The data wasn’t for adjusting treatments, but for Pfizer to assess Theranos’s tech.
Chapter 3 implies that Holmes didn’t make this clear to her employees. Anna, a product lead, suggested ironing out kinks, but Holmes trudged on. In Chapter 6, Pfizer ended this collaboration after underwhelming results.
3 ex-Theranos folk formed Avidnostics (for veterinarians), believing to have solved Theranos 1.0’s microfluidic problems. Well, Theranos sued them for IP theft, and involved the FBI.
Holmes hired a new (rival) engineering team. The team pivoted to automating steps done by chemists while testing blood. Thus The Edison (internally ridiculed as GlueBot) came to be. Used 50μl instead of 10μl, but it was still just a drop. The previous team was let go.
You can’t Steve Jobs your way through medical devices. But you can’t change the world by being cynical.