The World Health Organization estimates that drug-resistant diseases will cause 10 million deaths in 2050 — but biotech startup PhageLab thinks it has the solution.
According to their pitch deck (published by TechCrunch), PhageLab’s technology develops new antibiotics in just 45 days, and in field trials those antibiotics have been up to 100% effective in eliminating salmonella.
This is, as TechCrunch says, a hell of a promise. And before I eat that PhageLab chicken, I’d like to know the bottom end of their product’s effectiveness than the top end. Because that’s what “up to” means: It’s the maximum possible number. It’s not an average or a median, it’s a best case scenario.
Whenever someone claims their product or service or sale is “up to” something, you can be pretty sure they’re the ones who are up to something:
Huckberry’s April sale includes items for “up to” 60% off. Including this shirt, which is only 23% off, and this jacket, which is a paltry 10% off.
The State of California will give “up to” $150,000 to first-time homebuyers. But that means some buyers will receive $50,000 or $15,000 or nothing at all.
A mommy blogger was sentenced to “up to” 60 years in prison. But on average, prisoners only serve 44% of their sentences before they’re released.
So while PhageLab’s best field trials showed they could eliminate 100% of salmonella presence, they’ve formulated this claim in a way that hides the results of their worst field trials, and even their average trials. I think I’ll pass on that chicken wing.
And sure enough, when you dig into PhageLab’s deck you find this isn’t the only questionable data. They misquote that WHO estimate (to their benefit, of course), and if you believe their claims then the standard antibiotic drug requires more than 120 years of R&D.'
So beware of those who claim “up to” — and if this language sneaks into your own data copy, find a better way to explain your data.
Thanks for reading. What’s the most dubious data claim you’ve seen? Post it in the comments below or on LinkedIn. And if you want good data delivered to your inbox, subscribe here.