5 ms·

It's worse than that: Ben Goldacre is reporting ( https://twitter.com/bengoldacre/status/440475049880195073 https://twitter.com/bengoldacre/status/4404750498801

by pja 13y ago

It's worse than that: Ben Goldacre is reporting ( https://twitter.com/bengoldacre/status/440475049880195073 https://twitter.com/bengoldacre/status/440475049880195073 ) that the data was made publicly available. This is beyond parody.

arethuza 13y ago
To add insult to injury they were probably getting paid vast amounts of taxpayers' money to be this incompetent.
chillax 13y ago
another update: "No individuals directly named records online. But a massive breach of the most basic information security policies to prevent jigsaw." https://twitter.com/bengoldacre/status/440488463008550912 https://twitter.com/bengoldacre/status/440488463008550912
- flurpitude 13y ago
  We're going to need something longer than a tweet to explain what he's actually talking about.
  - sentenza 13y ago
    I'd also assume the worst, even if there are no patient names out. There are too many examples of de-anonymization of purposefully anonymized data out there to warrant any belief that missing real world names alone should constitute much of a privacy blanket.
    - vidarh 13y ago
      The dataset in question supposedly contains date of birth, gender and post code. For my family, that uniquely identifies every member, given the size of UK post codes.
      - Silhouette 13y ago
        If that means full postcodes, then that information will easily be sufficient to identify every member of the UK population, aside from outliers like twins or the occasional statistical fluke.
Sniperfish 13y ago
recent (20m): "I’m sorry to say I’m also now aware of more @HSCIC stories breaking shortly. What a mess. They need to come clean asap, clean stables." "Wait for details on story i’ve been tweeting today: twitter is first draft. Story is: small number rule breach, I believe, which is bad." https://twitter.com/bengoldacre https://twitter.com/bengoldacre
pja 13y ago
It transpires that the publicly downloadable data was a mock dataset. How it took them 24 hours to work that out I've no idea, but I imagine there's a lot of headless chicken impressions going on behind the scenes at the moment. https://twitter.com/bengoldacre/status/440576479248662528 https://twitter.com/bengoldacre/status/440576479248662528 It's still the case that real NHS data was uploaded to Google's servers.
lern_too_spel 13y ago
It's beyond parody because it's just plain false. Par for the course for The Guardian over the last few months. http://www.earthware.co.uk/default.aspx http://www.earthware.co.uk/default.aspx
- pja 13y ago
  It had nothing to do with the Guardian. What on earth are you talking about?
  - lern_too_spel 13y ago
    Ben Goldacre writes for The Guardian. What else would I be talking about?

by pja 13y ago

arethuza 13y ago

To add insult to injury they were probably getting paid vast amounts of taxpayers' money to be this incompetent.

chillax 13y ago

another update: "No individuals directly named records online. But a massive breach of the most basic information security policies to prevent jigsaw." https://twitter.com/bengoldacre/status/440488463008550912 https://twitter.com/bengoldacre/status/440488463008550912

flurpitude 13y ago
We're going to need something longer than a tweet to explain what he's actually talking about.
- sentenza 13y ago
  I'd also assume the worst, even if there are no patient names out. There are too many examples of de-anonymization of purposefully anonymized data out there to warrant any belief that missing real world names alone should constitute much of a privacy blanket.
  - vidarh 13y ago
    The dataset in question supposedly contains date of birth, gender and post code. For my family, that uniquely identifies every member, given the size of UK post codes.
    - Silhouette 13y ago
      If that means full postcodes, then that information will easily be sufficient to identify every member of the UK population, aside from outliers like twins or the occasional statistical fluke.

Sniperfish 13y ago

recent (20m): "I’m sorry to say I’m also now aware of more @HSCIC stories breaking shortly. What a mess. They need to come clean asap, clean stables." "Wait for details on story i’ve been tweeting today: twitter is first draft. Story is: small number rule breach, I believe, which is bad." https://twitter.com/bengoldacre https://twitter.com/bengoldacre

pja 13y ago

It transpires that the publicly downloadable data was a mock dataset. How it took them 24 hours to work that out I've no idea, but I imagine there's a lot of headless chicken impressions going on behind the scenes at the moment. https://twitter.com/bengoldacre/status/440576479248662528 https://twitter.com/bengoldacre/status/440576479248662528 It's still the case that real NHS data was uploaded to Google's servers.

lern_too_spel 13y ago

It's beyond parody because it's just plain false. Par for the course for The Guardian over the last few months. http://www.earthware.co.uk/default.aspx http://www.earthware.co.uk/default.aspx

pja 13y ago
It had nothing to do with the Guardian. What on earth are you talking about?
- lern_too_spel 13y ago
  Ben Goldacre writes for The Guardian. What else would I be talking about?