4 ms·

Just wait until you find out all the trendy SaaS platforms your employer subscribes to are also having their corporate data sold.

specialist 6y ago
Of course. I should have thought of that. Ironically, now that corporations are people too, they'd also benefit from my solution. Grandpa story time: mid aughts, the company (Quest Diagnostics) that acquired us was trying to figure out how to monetize all the awesome new patient data our product was aggregating. Labs, scripts, allergies, etc. Negotiations with MS, Google, pharmas. Our Zuck and Thiel equivalents were electrified by the prospects. Us nerds were horrified. I now forget which specific ruling killed the idea. But it now occurs to me that I should have leaned in. At the time I was still trying to find, invent, steal a technical solution and had not yet figured out that's impossible.
- altdatathrow 6y ago
  Haha, I was not suggesting corporations should enjoy the same inherent rights as people but it’s still pretty bad. Amounts to corporate information theft sold to anyone willing to pay. There’s healthcare providers who sell patient data today still. It’s “anonymized” of course but if you want to know the history a given medication or procedure is performed on a per-patient basis, that’s all available. It’s just an anonymous identifier and somewhere someone checked a box that grants permission to resell that info, so it’s unfortunately legally acceptable under current law.
  - specialist 6y ago
    Damn. I didn't know "anonymized" patient data was being monetized. I've been out of healthcare IT since 2009. For two reasons. #1 Since I'm a slow learner, it took me a long time figure out that big data will always win in the anon/deanon arms race. I don't really grok "differential privacy", so I just say it quantifies how much one has to fuzz the data to protect any single individual, given the data's bounds. But even then it requires the "honor system" to prevent others from using additional data to deanon fuzzed data. It really clicked for me during a demo of Seisent (now owned by LexisNexus). Given enough data sources, the errors cancel out, so there's no way to hide or be an imposter. You'd stick out. Their sales team gave some examples of law enforcement using their tool to close cold cases. Not in so many words, because people would freak out, but basically: instead of finding the needle in the haystack, they rule out everyone who couldn't be the perp, and any one remaining is a suspect, and one of them has to be the bad guy. So then they work backwards, just like working around tainted evidence. (I've since had this theory confirmed by a professional bounty hunter. He was very surprised a layperson figured this out.) I've talked to a lot (20+) healthcare startups, and attended confs and meetups, since. As of two years ago, exactly none of them understood the problem space or patient privacy. And I just don't have the goodwill or energy to refight those battles with another crop of noobs. #2 Most of us in our startup wanted to "fix" healthcare. Our contribution was going to be portable medical records. Specific initial use cases were to prevent mistakes due to missing patient history. Over time we got really good at showing our clients their own data, which became the prime selling point. It eventually became clear that only a single payer system would have the juice to force all the competitors to play nicely. Good so far. The Bush Admin intended to allocate $17b federal grants to make it happen. Alas. Their mortgage default dumpster fire, causing the 2008 economic meltdown, and all federal juice went poof. Sadly, the Obama Admin didn't pick up that mantle. (They had other ideas about data sharing, which seemed naive at the time, but I haven't gone back to see who was more right.) This isn't just a sob story. Ironically, I came to understand that one viable strategy to protecting patient privacy requires a top-down approach. Like a Real ID for medical data and services. Until that, or something better, happens, I just can't muster any interest.

specialist 6y ago

Of course. I should have thought of that. Ironically, now that corporations are people too, they'd also benefit from my solution. Grandpa story time: mid aughts, the company (Quest Diagnostics) that acquired us was trying to figure out how to monetize all the awesome new patient data our product was aggregating. Labs, scripts, allergies, etc. Negotiations with MS, Google, pharmas. Our Zuck and Thiel equivalents were electrified by the prospects. Us nerds were horrified. I now forget which specific ruling killed the idea. But it now occurs to me that I should have leaned in. At the time I was still trying to find, invent, steal a technical solution and had not yet figured out that's impossible.

altdatathrow 6y ago
Haha, I was not suggesting corporations should enjoy the same inherent rights as people but it’s still pretty bad. Amounts to corporate information theft sold to anyone willing to pay. There’s healthcare providers who sell patient data today still. It’s “anonymized” of course but if you want to know the history a given medication or procedure is performed on a per-patient basis, that’s all available. It’s just an anonymous identifier and somewhere someone checked a box that grants permission to resell that info, so it’s unfortunately legally acceptable under current law.
- specialist 6y ago
  Damn. I didn't know "anonymized" patient data was being monetized. I've been out of healthcare IT since 2009. For two reasons. #1 Since I'm a slow learner, it took me a long time figure out that big data will always win in the anon/deanon arms race. I don't really grok "differential privacy", so I just say it quantifies how much one has to fuzz the data to protect any single individual, given the data's bounds. But even then it requires the "honor system" to prevent others from using additional data to deanon fuzzed data. It really clicked for me during a demo of Seisent (now owned by LexisNexus). Given enough data sources, the errors cancel out, so there's no way to hide or be an imposter. You'd stick out. Their sales team gave some examples of law enforcement using their tool to close cold cases. Not in so many words, because people would freak out, but basically: instead of finding the needle in the haystack, they rule out everyone who couldn't be the perp, and any one remaining is a suspect, and one of them has to be the bad guy. So then they work backwards, just like working around tainted evidence. (I've since had this theory confirmed by a professional bounty hunter. He was very surprised a layperson figured this out.) I've talked to a lot (20+) healthcare startups, and attended confs and meetups, since. As of two years ago, exactly none of them understood the problem space or patient privacy. And I just don't have the goodwill or energy to refight those battles with another crop of noobs. #2 Most of us in our startup wanted to "fix" healthcare. Our contribution was going to be portable medical records. Specific initial use cases were to prevent mistakes due to missing patient history. Over time we got really good at showing our clients their own data, which became the prime selling point. It eventually became clear that only a single payer system would have the juice to force all the competitors to play nicely. Good so far. The Bush Admin intended to allocate $17b federal grants to make it happen. Alas. Their mortgage default dumpster fire, causing the 2008 economic meltdown, and all federal juice went poof. Sadly, the Obama Admin didn't pick up that mantle. (They had other ideas about data sharing, which seemed naive at the time, but I haven't gone back to see who was more right.) This isn't just a sob story. Ironically, I came to understand that one viable strategy to protecting patient privacy requires a top-down approach. Like a Real ID for medical data and services. Until that, or something better, happens, I just can't muster any interest.