3 ms·

"The company says the agreed cut-off point for data was 17 February, by which time there had been 141 cases and efficacy was 79%. Within 48 hours, AstraZeneca h

by 0-_-0 6y ago

"The company says the agreed cut-off point for data was 17 February, by which time there had been 141 cases and efficacy was 79%. Within 48 hours, AstraZeneca had added more recent data. On a total of 190 cases, there was 76% efficacy (and as always 100% against severe illness and death). It’s not much of a difference." [0] [0]: https://www.theguardian.com/business/2021/mar/26/how-the-astrazeneca-vaccine-became-a-political-football-and-a-pr-disaster https://www.theguardian.com/business/2021/mar/26/how-the-ast...

ethanbond 6y ago
Is this supposed to be some mitigation? A pharma company has a responsibility to publish accurate and up to date data. Full stop.
makomk 6y ago
That baffled me. They got a huge amount of stick for measuring up to the pre-determined cutoff point, which as I understand it is the correct way to do a scientific study if you want your results to actually be valid...
- refurb 6y ago
  From what gathered, AZ said they had final data, but hadn’t actually finished the analysis, but preliminary numbers looked similar to interim so they just used those.
- mensetmanusman 6y ago
  This. The reason it is risky to measure data after a predetermined stop point is that it lets human bias play a role in making conclusions. It is very subtle, but this human behavior has been a contributor to the replication crisis: https://en.m.wikipedia.org/wiki/Replication_crisis https://en.m.wikipedia.org/wiki/Replication_crisis
hannob 6y ago
"It's not much of a difference" is a terrible excuse. This isn't about the percentage points they were off, this is about how such a thing looks in the public. When you get a warning from a gov agency that you shared outdated data that just looks terrible. Just don't do anything in a press release that could even remotely look like you're trying to cheat with your data. It's a terrible idea in a situation like this.

by 0-_-0 6y ago

ethanbond 6y ago

Is this supposed to be some mitigation? A pharma company has a responsibility to publish accurate and up to date data. Full stop.

makomk 6y ago

That baffled me. They got a huge amount of stick for measuring up to the pre-determined cutoff point, which as I understand it is the correct way to do a scientific study if you want your results to actually be valid...

refurb 6y ago
From what gathered, AZ said they had final data, but hadn’t actually finished the analysis, but preliminary numbers looked similar to interim so they just used those.
mensetmanusman 6y ago
This. The reason it is risky to measure data after a predetermined stop point is that it lets human bias play a role in making conclusions. It is very subtle, but this human behavior has been a contributor to the replication crisis: https://en.m.wikipedia.org/wiki/Replication_crisis https://en.m.wikipedia.org/wiki/Replication_crisis

hannob 6y ago

"It's not much of a difference" is a terrible excuse. This isn't about the percentage points they were off, this is about how such a thing looks in the public. When you get a warning from a gov agency that you shared outdated data that just looks terrible. Just don't do anything in a press release that could even remotely look like you're trying to cheat with your data. It's a terrible idea in a situation like this.