38 replies »

1. This might be too simplistic, but I’m not sure that anyone has the rights to my data in any form. That would be pre, post or in the act of. It’s not your ‘right’ to have anyone else’s data at all. If so, then what you should do is pay a recurring fee for such data since it will be used for years to come!

2. This thread really IS about Matthew Holt and everyone that frames privacy as a compromise. Privacy vs. Research. Privacy vs. Progress. Privacy vs. Open Data. Privacy vs. Terrorism. Privacy vs. Freedom. In my experience, all such framing says more about the proponent than it does about the subject.

   Almost always, the framing of privacy as a compromise is a cover for lack of transparency and an economic foundation that benefits from externalities. The best-known example in health care is people being denied access to our own health records “because of HIPAA”. The clearest example of hiding costly externalities is the lack of accounting for disclosures by declaring that a pseudonym is equivalent to de-identification and thereby eliminating the right of patient consent. (My point in the comment at 12:54 PM).

   I will be amazed and delighted if you and THCB can actually get some of the data brokers and other hidden data businesses to join this thread and explain how they get to have personal data about us that we ourselves have a hard time getting.

3. Adrian–this thread isn’t about me (or you for that matter) advocating anything. It’s about whether the HIPAA CEs are in fact sharing IDENTIFIED data with the data brokers they do business with, as Deborah claims. Or at last as I think she claims–cant be too careful when I surmise her position lest TCHB editors get another email from her!

   I have no idea how the “is that data you are sharing really deidentified” position is enforced, other than as you say both sides saying “trust us” so I will ask some more informed people and revert back.

4. The bipartisan Coalition for Patient Privacy worked with Congress to ban health data sales in HITECH/ARRA. But it was eliminated during the regulatory process. The regulations issued by HHS allow the sale of PHI to continue, all the current uses were deemed to be exceptions. Congressman Joe Barton of Texas championed the ban on sales of PHI.

5. Matthew: The current HIPAA situation is illogical:

   – a service provider can claim they are sharing de-identified data
   – this exempts the service provider form both consent and accounting for disclosures
   ,
   What recourse does society or a patient have to seek transparency with this kind of regulation? In the limit, this makes all HIPAA CEs self-policed with only whistleblowers to fear. The HIPAA CE doesn’t even need a published privacy policy the way non HIPAA regulated services do. In the limit, Health 2.0 companies will need to buy the patient data from the HIPAA CE.

   Is this what you’re advocating?

6. The only use of patient data obtained during the course of treatment shoul be tor research by the staff of the hospital. Sales of said data should be banned.

7. Adrian–I completely agree. But that’s not what I or Deborah is talking about.

8. Matthew – US healthcare is wasting over $1,000 Billion per year partly through fraud but mostly through price-fixing. That level of waste requires obfuscation of cost and quality on a scale that only our current system of private data brokers and employer-based health insurance can provide.

   The harms of lack of patient and physician access to our own information is that $3 Trillion is being spent in ways that do not effectively reflect the wants and needs of patients. Without effective transparency, our privatized health system appears perfectly designed to give the results we’re getting.

9. Suggested reading: JASON Report, Section 6.3, pp 51-54

   “JASON finds that de-ID is not a viable approach for ensuring patient privacy going forward.”
   …. and other illuminating insights…..

10. Indeed. However, until my own data is accessible and controllable BY ME, I’m just gon’ be a one-note song.

    I’m not thrilled about 3rd parties – brokers or others – having access to my data, de-IDed or not. It’s another example of how money drowns out any logic, or humanity, in the US healthcare system. But I’m a realist, and know that paradigm won’t shift until we start a big enough fire/argument in the PATIENT part of the forest.

11. Amen Casey. I want full access to all their data for all patients. BUT that’s from the transactional data systems that currently have those records. This Argument is about what happens to that data when it gets passed onto data brokers (and whether it is de-identified or not when it is passed on)

12. Deborah–the entire HIT industry is about $30-40 billion in revenue. How do you get to ” industry profits are in the tens to hundreds of billions of dollars/year from the sale, trade, and exchange of our health information.”

    All I’m asking is some level of proof to back up your assertions.

13. Adrian–You’re arguing about 2 different things. fm is saying that the data brokers do NOT use PHI in the creation of their longitudinal data sets which is absolutely not what Deborah says. I think fm is correct (but perhaps Deborah & you know different)

    You are saying that there shouldn’t be a commercial data broking industry but instead should be open data sets for the betterment of scient and society. Which is hard to argue with and seems to be what is happening in the UK.

    But these are different arguments

14. It’s called “surveillant anxiety.”

    http://thenewinquiry.com/essays/the-anxieties-of-big-data/

15. Practice Fusion is not a physician. Name some MDs who are allegedly knowlingly and directly selling ePHI.

16. This he said, she said stuff is rather unproductive.

    Some feel that we have a privacy and PHI issue, some don’t. As an average joe, I am concerned about my information being sold, shared, released, whatever you want to call it. I know it’s done in this and other markets, that still does not make me comfortable.

17. Holt and Peel should wager in an electronic futures market and put some skin in. “Will PHI be sold in a one billion dollar a year illegal marketplace in the US within five years time?”

    If we can control centrifuges in Iran we can find your MRI diagnosis. Alas,
    Stuxnet has now been succeeded by a better product, “Regin”. .

18. http://latanyasweeney.org/work/identifiability.html

    Prof Latanya Sweeney weighs in on these issues.

    There needs to be a hard stop on vendors selling the data that their devices gather when the intended purposes for their devices are for mitigating disease.

19. This was in Science Jan.18, 2013.

    Identifying Personal Genomes by Surname Inference

    Melissa Gymrek1,2,3,4,
    Amy L. McGuire5,
    David Golan6,
    Eran Halperin7,8,9,
    Yaniv Erlich1,*

    + Author Affiliations

    1Whitehead Institute for Biomedical Research, 9 Cambridge Center, Cambridge, MA 02142, USA.
    2Harvard–Massachusetts Institute of Technology (MIT) Division of Health Sciences and Technology, MIT, Cambridge, MA 02139, USA.
    3Program in Medical and Population Genetics, Broad Institute of MIT and Harvard, Cambridge, MA 02142, USA.
    4Department of Molecular Biology and Diabetes Unit, Massachusetts General Hospital, Boston, MA 02114, USA.
    5Center for Medical Ethics and Health Policy, Baylor College of Medicine, Houston, TX 77030, USA.
    6Department of Statistics and Operations Research, Tel Aviv University, Tel Aviv 69978, Israel.
    7School of Computer Science, Tel Aviv University, Tel Aviv 69978, Israel.
    8Department of Molecular Microbiology and Biotechnology, Tel-Aviv University, Tel Aviv 69978, Israel.
    9The International Computer Science Institute, Berkeley, CA 94704, USA.

    ↵*To whom correspondence should be addressed. E-mail: yaniv@wi.mit.edu

    Abstract

    Sharing sequencing data sets without identifiers has become a common practice in genomics. Here, we report that surnames can be recovered from personal genomes by profiling short tandem repeats on the Y chromosome (Y-STRs) and querying recreational genetic genealogy databases. We show that a combination of a surname with other types of metadata, such as age and state, can be used to triangulate the identity of the target. A key feature of this technique is that it entirely relies on free, publicly accessible Internet resources. We quantitatively analyze the probability of identification for U.S. males. We further demonstrate the feasibility of this technique by tracing back with high probability the identities of multiple participants in public sequencing projects.

20. I probably should have added that, yes, of course, I know that there are also numerous federal and state criminal laws that might come into play regarding specific privacy violations. Just that they are statutory in nature, not Constitutional.

    The 4th Amendment is pretty much on life support, in any event.

    http://bgladd.blogspot.com/2013/06/clapp-trap.html

21. Ms. Peel, @11:01,

    “So, we know, for example, that physicians and their EHRs sell the [ePHI] data…”

    http://www.youtube.com/watch?v=f1JPjLCxPFQ

    Nice conjunctive conflation. Name some physicians who are selling ePHI?

22. “Patient Privacy Rights is a small non-profit 501c3 organization”
    __

    This is true. I’ve examined your 990’s. At least YOU, commendably, are not making bank in the high six figures or more. Unlike a lot of other ePHI “non-profit” organizations, like HL7.

23. Hi Mighty Casey:

    PPR and I derive no profit from fighting for “patient-controlled digital data exchange”. I am a FT volunteer. Patient Privacy Rights is a small non-profit 501c3 organization I founded in 2004 because there was no national organization fighting up for the privacy rights of the people of this nation–and 95% of the US public wants to control who can see, use, and sell their most sensitive personal information: data about their minds and bodies.

    On the other hand, industry profits are in the tens to hundreds of billions of dollars/year from the sale, trade, and exchange of our health information.

24. Ah, Brandeis. Keyword “dissent.”

    From my 1998 grad thesis:

    As I have alluded to elsewhere in this thesis, one of the most durably contentious of American Constitutional claims involves the right to privacy. If we are to establish a case for privacy as a fundamental ethical principle that the law ought reflect and administer with vigor, we ought examine a bit of its legal, sociological, and philosophical evolution. In Chapter 4 we began by examining the historical evolution of search-and-seizure restraints that ultimately found their way into our Fourth Amendment, and we ended with a review of the convoluted, often contradictory U.S. Supreme Court case law history and constitutional interpretation theory that undergirds our current legal and political confusion over the role of privacy as it pertains to drug policy. We begin here with some general sociopolitical and legal theory reflections that serve as foundation for and transition into the larger philosophical concepts bearing on privacy discussed in the latter part of this chapter.

    Some regard privacy as an inseparable aspect of personal autonomy requisite for the very notion of liberty we ostensibly revere as a cardinal element our social and legal order. Critics, on the other hand, either dismiss the notion of a general right to privacy out of hand, or assert that it is a relatively recent, weak, and “derivative” declaration, one inherently inimical to and necessarily deferential to society’s “right-to-know” in the interest of commercial efficiency, public safety, and criminal prosecution. Those holding this latter position view the quest for privacy as a reaction to increasing urbanization and advances in information processing technologies, that the inhabitants of earlier eras and non-industrial cultures had and have little concern with our notions of “privacy.” Critics of the former persuasion who disavow the very notion of a general right to privacy under federal law find the concept adequately accounted for principally in terms of property rights. Libertarian advocate Murray N. Rothbard, for example, argues in The Ethics of Liberty that “there is no such thing as a right to privacy except the right to protect one’s property from invasion.” Rothbard holds that what some regard as an invasion of privacy is more correctly seen as a misappropriation of property, “not some vague and woolly invasion of a “right to privacy.”’”

    http://bgladd.com/drugwar/chapter5.htm

    See also Chapter 4. e.g.,

    In his 1991 American Business Law Journal article Workplace Privacy and the Fourth Amendment: An End to Reasonable Expectations?(Vol. 29), legal scholar Don Mayer argues that the major drug testing cases have served a central role in the devaluation the original meaning of “privacy” under the Fourth Amendment. Mayer notes that in Fourth Amendment criminal cases generally, “the Court has generally found individual expectations unreasonable, and in civil cases has generally discarded both probable cause and warrant requirements” and that “ the Court’s contractarian thinking encourages the notion that not only statutory but constitutional rights can be explicitly or implicitly waived by the demands of employers or the ‘operational realities of the workplace’.” (p. 631)

    Mayer observes that what was once a constitutional “right” has been downgraded to a mere “interest” in competition with other, more powerful interests: an individual interest whose worth must be calculated on the basis of “reasonable expectations.” What are “reasonable expectations”? Well, whatever society is prepared to accept as “reasonable,” a notion with a charmingly bootstrap quality where drug testing cases are concerned. For example, Mayer cites Willner v. Thornburgh (928 F.2d 1185, D.C. Cir. 1991), wherein the majority wrote that

    [m]ore than 85 percent of employers with drug-testing programs tested job applicants . . . Some of the nation’s largest employers, including American Telephone & Telegraph, DuPont, Exxon, Federal Express, Trans World Airlines, and United Airlines . . . What is occurring generally outside government is some indication of what expectations of privacy “society is prepared to accept as reasonable” when the government engages in the hiring process. (Mayer, p. 650)

    So, the very fact that private sector corporations, “unguarded” by the Fourth Amendment (recall Ginsberg earlier in Chandler), screen for illegal drugs willy-nilly without cause becomes itself the reference standard for “reasonable expectations” under the Fourth Amendment. How conveniently circular.

    In essence, on this line of reasoning the scope of the Fourth Amendment is determined by Disney and its corporate brethren, not by the courts. Whatever is required to meet the “operational realities of the workplace” trumps, by virtue of its mere declaration and enactment, an employee’s puny privacy “interest.”
    __

    Again, you have no explicit, uncontested 4th Amendment right to informational privacy. SCOTUS and myriad inferior courts have been all over the map on the issue. Moreover, the ePHI privacy issue PALES in comparison to what the data miners are doing with all of the other digital crumbs of our lives.

    None of which is to imply that I’m OK with any of this. Far from it.

25. Practice Fusion does sell patient data–in the earlier thread I posted quotes from the CEO:
    “Every healthcare vendor is selling data. Everyone has this data, but we’ll have more of it and it will be real-time and aggregated,” Howard said. ”

    URL: http://www.healthcareitnews.com/news/practice-fusion-expands-shows-signs-rapid-growth?single-page=true

    Practice Fusion’s business model of selling patient data was also explained by Chris Anderson in his book “Free, the future of a radical price” on page 104 (published by Hyperion in 2009)

26. Sorry some information was left out:

    The paragraph is on page 16 of the download = page 91 in the book itself.

27. Imma come into this swordfight with nothing but a 5 iron, in the form of being a patient/simple-human being solely interested in gaining ACCESS TO my own health data, contained hither and thither across the medical-industrial complex.

    I frankly could care less about the sale of my data. Since I can’t full access it without spending hours/days in endless rounds of begging, with robust deployment of face-palm and head-desk when I’m told, “we can’t, because HIPAA.”

    I’m very afraid that all the Philadelphia lawyer language that’s been sprayed all over the health data issue have made it impossible for anyone to understand what the rules really are re data access/exchange.

    So I’ll ask “qui bono?” [Not a lawyer, just a recovering Catholic.] I’m very afraid that the Arguing Classes have gained ascendancy, much to the detriment of ever having a civil discussion on this topic. Or EVER arriving at a resolution in the form of patient-controlled digital data exchange.

    ‘Cause the folks that have it, and who are involved in The Argument, are profiting from keeping The Argument going.

28. Hi BobbyGvegas:

    Re: Our constitutional rights to ‘health information privacy’—-Below is a more complete explanation— it’s in the free HIMSS book chapter on the page 1 (the numbers are for the citations listed at the end of the chapter):

    “Americans also have strong Constitutional rights to privacy and health information privacy. The Constitutional right to privacy grew from Justice Brandeis’ 1928 dissent in Olmsted.27 He famously wrote, “The right to be let alone is the most comprehensive of rights and the right most valued by civilized men.” The right to privacy of highly personal information is protected under the right to be free of unreasonable searches and seizures under the Fourth Amendment, and the right to liberty under the Fifth and Fourteenth Amendments to the U.S. Constitution.28 The right to privacy of personal information also has been recognized by Congress and by HHS as a fundamental constitutional right.29 The right to “informational privacy,” i.e., the right of an individual to have his personal information kept private, 30 grew from Whalen v Roe in 1977.31 Finally, the U.S. Supreme Court established a psychotherapist-patient privilege in 1996.32 “The mere possibility of disclosure may impede the confidential relationship necessary for successful treatment.”33 Failure to protect the right to health information privacy leads to less health information because communications between practitioners and patients “would surely be chilled.”34

    One more time–the chapter is FREE to download because PPR licensed it: http://patientprivacyrights.org/wp-content/uploads/2014/06/Peel-chapter-HIMSS-book.pdf

29. fm – please see my comment @10:39. Does the privatization of medicine by hidden data brokers serve either the patient or the public interest? Is there still a role for open, peer reviewed, and freely teachable research and should patient and research organizations be going to the data brokers to buy back our own health records?

30. Matthew – let’s define de-identification and sale.

    For example, consider Foundation Health or CollabRx (purveyors of cancer genome testing services) who get a copy of the patient’s de-identified health record and outcome and then sell their clinical association analysis to the next patient that has a similar genome.

    Can the HIPAA CE that sent their patient’s sample and her health record to Foundation Health or CollabRx avoid asking for patient consent by cliaming the information was de-identified?

    Is what these companies are doing by treating patient outcome results as their intellectual property and selling the clinical association to other patients not a sale?

    Is this kind of de-identification and private sale a vision of Health 2.0 where clinical data is not available for replication and peer review because it’s moved directly from HIPAA CE to unregulated data broker with no patient notice or access?

31. “What has happened since the $30 billion HIT stimulus program is that technology outside of Health IT has changed radically, and we need to get that new tech (cloud/tablet/mobile/improved user experience) into the real world of doctors hospitals and patients.”
    __

    Indeed. See “EHR Science,” Jerome Carter MD:

    http://ehrscience.com/2014/11/17/a-question-for-ehr-vendors-what-is-a-real-computer/

32. The joke is that I by and large agree with PPR that we should have more access to and control over our personal health data, so I’m happy to concede that I misunderstood every time I saw Deborah quoted, and that she’s never been opposed to health IT. I also invite Deborah to write a piece in THCB summarizing the chapter in the HIMSS book, as it would be nice for more than the 4 people who bought and read it to know what she really wants.

    I think our difference (and BTW the argument I have with Scott Silverstein) is that I start in the real world, and want to move it to a better state, whereas (it appears) Deborah et al just want their ideal state and dont want to work with the current system. But again that’s irrelevant. As my favorite defense secretary said, you go to war with the systems you had not the ones you want. What has happened since the $30 billion HIT stimulus program is that technology outside of Health IT has changed radically, and we need to get that new tech (cloud/tablet/mobile/improved user experience) into the real world of doctors hospitals and patients.

    But this thread was not supposed to be about what I said about Deborah or what she said about IT in general, it’s supposed to be about the sale of personally identified data by Health IT vendors and specific accusations Deborah made about athenahealth, IMS and Practice Fusion.. Adrian in his comment added Optum to that list.

    I agree that we should break up (as HIPAA does) the use of data for medical purposes versus marketing, and lets have a fact based explanation about what is actually happening in both those sectors, and how personal information is being used. I want vendors and covered entities to comment, but in particular Deborah has the chance to discuss the three specific examples she raised. Will she take it?

33. “our constitutional rights to health information privacy”
    __

    Seriously? There IS no such explicit right. 4th Amendment Constitutional rights, to the extent they are clarified at all and fixed over time, pertain to only a very limited subset of government intrusion. Harmful commercial trafficking in your private health information (or ANY kind of personal data, mostly occuring beyond the confines of CE’s and their BA’s, btw) falls under the domain of tort, and is mostly a crazy-quilt state-level jurisdictional matter.

34. “Optum Insight proudly advertizes that they can combine health records and insurance claims for millions of patients.”
    __

    And Uber knows if you’ve been naughty.

    https://web.archive.org/web/20140827195715/http://blog.uber.com/2012/01/09/uberdata-san-franciscomics/

35. She’s even gone so far as to recently claim — however vaguely (conveniently) — that PHYSICIANS are selling their patients’ data.

    Interestingly, on this topic, ‘Quantified Self-Incrimination?’ “When Fitbit Is the Expert Witness” see http://www.theatlantic.com/technology/archive/2014/11/when-fitbit-is-the-expert-witness/382936/

36. Matthew, I appreciate your efforts to shine a light on the data practices of our HIPAA “Covered Entities” – the institutions we go to when we’re hurting and the doctors that are manipulated into using technology that our doctors don’t purchase and don’t control. Patient data is strategic in health reform and the physician-patient relationship is collateral damage.

    Measuring the quality of health care http://www.nytimes.com/2014/11/19/opinion/how-medical-care-is-being-corrupted.html and http://www.nytimes.com/2014/11/23/upshot/how-to-arrive-at-the-best-health-policies-.html is more important than ever. Who do we want to trust with managing and analyzing all of this digital data? Patients and patient advocates? Physicians and medical societies? No, let’s have the foxes guard the hen-house. The folks that run the institutions, specify the EHRs and move the data around to set and justify the prices have the data and they will tell us what it says.

    Ten years and $30 B into a well-intentioned effort to digitize patient data, patients don’t even have visibility, much less control over our own data. The days of paper, when I could ask my doctor for a copy of my health record and get the same handed to me are gone. In the digital EHR data world, my doctor can’t do that any more. That is the point of last-week’s related thread by Deborah and me https://thehealthcareblog.com/blog/2014/11/14/an-open-letter-on-the-nate-trust-community/

    As Deborah points out, our patient data is being shared, aggregated, and analyzed at an unprecedented rate. Data brokers assemble profiles of us as patients that span time and service provider institutions. Optum Insight proudly advertizes that they can combine health records and insurance claims for millions of patients. IMS Health also creates longitudinal profiles for analysis on a massive scale.

    Here’s an opportunity for THCB readers to suggest how health data brokers assemble these patient profiles if the data they get from our HIPAA Covered Entities is de-identified.

    Here’s an opportunity for EHR vendors and their HIPAA Covered Entity customers to shine some light into how our data is sent out for analysis and why it’s too hard, in 2014, for us as patients to have that same feed visible and a copy of that data sent to us.

    Adrian

37. Matthew:

    Again you start by asserting something that is false: I have never opposed health IT. PPR and I are very pro-HIT. I have never been quoted as opposing health IT, because I don’t.

    What PPR and I do is press for HIT that complies with longstanding US law, medical ethics, and our constitutional rights to health information privacy. These consensus protections developed over the past 100+ years in state and federal law, and the courts have upheld patients’ rights of consent (with rare exceptions) before personal health information is shared or disclosed.

    PPR and I are pro-HIT, but current technology must be fixed so that patients’ rights to privacy are baked in, not eliminated.

    I asked you to please read our free chapter in a book published by HIMSS about how HIT can be fixed, so patients can trust HIT, and so that all the benefits of technology can be reaped while preventing harms to patients. It appears that you didn’t read it. Here is the link once again: http://patientprivacyrights.org/wp-content/uploads/2014/06/Peel-chapter-HIMSS-book.pdf

    It is very disheartening to read again your misrepresentations of me and PPR.

    It’s time for you and the other TCHB editors to retract the erroneous statement above that I oppose health IT on the home page and formally apologize. Again: I’ve never said or written that I oppose HIT.

    Technology is amazing, essential, and can provide incredible breakthroughs and benefits to humans and to healthcare systems. Technology is not the problem, the problem is building health technology that complies with US law, medical ethics, human and civil rights to privacy, and the expectations of the public to control who can see and use their most sensitive personal information.

    Deborah C. Peel, MD

38. Point of fact (which may be irrelevant given the tone of this debate)-

    Deborah uses the fact that IMS/SDI have longitudinal records as proof that they must have identifiable information.

    In fact they use one-way hash functions to create anonymous but linked (ie pseudonymized) records.

    patent: http://www.google.com/patents/US6397224
    nice description of science: http://datamining.anu.edu.au/talks/2008/ausdm2008-ppdl-tutorial-linkage.pdf