Showing posts with label data sharing. Show all posts
Showing posts with label data sharing. Show all posts

Tuesday, September 22, 2009

Data Sharing and Privacy: In the News

Roughly one year after the NIH and The Wellcome Trust restricted access to genome-wide association studies [see: Modifications to Genome-Wide Association Studies (GWAS) Data Access (NIH, August 28, 2008 - PDF)], data sharing is back in the news. In August, researchers in Tel Aviv and Berkeley announced a new method of protecting the privacy of individuals in genetic research studies. In the press release, one of the authors, Eran Halperin, alludes to the PLoS Genetics paper that led to the NIH's data access modifications (Homer N, et al.). As Halperin sees it, the "knee-jerk response stymied potential breakthrough genetic research." Perhaps he hopes that the new privacy approach will re-open the doors of safe data sharing. Halperin's method (published in Nature Genetics) uses a likelihood ratio (LR) test to measure the risk of exposing a single individual within a data set. Ultimately, the LR test limits the number of SNPs shared and thereby decreases the chances of privacy losses. If you would like to give this method a try, the software is available online (with registration) at SECUREGENOME.

In related news, The Toronto International Data Release Workshop (Genome Canada, May 2009) recently recommended the "rapid release of prepublication data" to speed the pace of scientific discover. The recommendation was published in a special, data sharing edition of Nature. The Workshop cites the Human Genome Project as an example of how sharing data leads to public benefits: "This experience ultimately demonstrated that the broad and early availability of sequence data greatly benefited life sciences research by leading to many new insights and discoveries, including new information on 30 disease genes published prior to the draft sequence." In a note addressing the human subjects concerns, the Workshop acknowledges the privacy risks: "Data about human subjects participating in genetic and epidemiological research require particularly careful consideration owing to privacy-protection issues and the potential harms that could arise from misuse. ... [F]or clinical and genomic data that are associated with a unique, but not directly identifiable individual, access may be restricted."

Will these developments (a proposed technical solution to the privacy barrier and a reaffirmation of the social benefits of data sharing) encourage the NIH and The Wellcome Trust to reduce some of the restrictions on data sharing? (I’d say: not yet.) While the safety of data sharing may be increased with technical solutions, it seems unlikely that all privacy risks will be eliminated. After all, if I share something "private" with you (or with anyone else), it's not so "private" any more. At the same time, if data access is restricted (for example, by using smaller, "safer" data sets), the public benefit of sharing resources declines. Clearly, we are looking for a way to eat our cake and have it too; protecting the individual's privacy while sharing for common benefits. Perhaps we should ask ourselves, what are the risks that we are willing to impose on individuals for the benefit of the common good? As an individual, exactly how private is your privacy? And at what cost?

References:

Homer N, et al. Resolving individuals contributing trace amounts of DNA to highly complex mixtures using high-density SNP genotyping microarrays. PLoS Genet. 2008 Aug 29;4(8):e1000167. PMID: 18769715

Sankararaman S, Obozinski G, Jordan MI, Halperin E. Genomic privacy and limits of individual detection in a pool. Nat Genet. 2009 Sep;41(9):965-7. Epub 2009 Aug 23. PMID: 19701190.

Toronto International Data Release Workshop Authors, et al. Prepublication data sharing. Nature. 2009 Sep 10;461(7261):168-70. PMID: 19741685.

Other Predictive Health Ethics News

Nikki Tait. Pan-European biobanking moves closer. Financial Times. September 16, 2009.

Michael Rugnetta and Whitney Kramer. Paving the Way for Personalized Medicine. Science Progress. September 14, 2009.

Daniel Vorhaus and Lawrence Moore. What happens when a personal genomics company goes bankrupt? Genetic Future. September 14, 2009.

Caroline Wright. HGC public consultation on DTC genetic testing services. PHG Foundation News. September 8, 2009.

Over Ninety Per Cent Of Pathologists Find Research Rules Too Complex. Medical News Today, September 8, 2009.

Jane E. Brody. Buyer beware of home DNA tests. The New York Times. August 31, 2009.

Brad Therrell, Harry Hannon, Don Bailey, et al. Considerations and Recommendations for a National Policy Regarding the Retention and Use of Dried Blood Spot Specimens after Newborn Screening. Genetic Alliance. August 21, 2009.

Turner Ray. Lack of Physician Education, Genetic Counseling Could Ruin Value Proposition of PGx Testing, Insurer Says. Genomeweb: Pharmacogenomics Reporter. August 12, 2009.

The Genetics and Public Policy Center. Center publishes new survey of state false advertising laws. dnapolicy.org, August 11, 2009.
- J.O.

Wednesday, July 1, 2009

Will Stronger Privacy Protections Result in Better Health Data? The Health Privacy Project Recommendations

The Health Privacy Project of the Center for Democracy & Technology (CDT) recently released a paper arguing for changes in how the HIPAA Privacy Rule protects "de-identified" health information. The recommendations grow from a one-day, CDT workshop held in September 2008. The Health Privacy Project makes the following eight recommendations:

1. Re-examine the Privacy Rule de-identification provisions (in particular, the safe harbor method for de-identification);
2. Strengthen accountability by requiring data use agreements;
3. Expand data anonymization options under the Privacy Rule;
4. Provide incentives to use less than fully identifiable data for certain purposes;
5. Provide support through “Centers of Excellence” in de-identification;
6. Require or encourage the use of limited access datasets and other technical solutions;
7. Require education and training of staff de-identifying data; and
8. Consider increasing public transparency regarding uses of de-identified data.


The Project argues that the HHS needs to re-examine the Privacy Rule "to ensure that the de-identification standard remains robust as re-identification becomes easier."

For readers struggling with the "Babel" of data privacy vocabulary (for example, what's the difference between "anonymous" and "anonymized"?), these recommendations may open the door to additional confusion, especially if #3 (above) means that additional categories of protected data are created. The Privacy Rule currently offers two categories data which are exempt from regulation: "de-identified" (presumed to be beyond the risk of re-identification and therefore not regulated) and not fully identifiable, "limited data sets" (incomplete data which includes some identifiers, for example: birth dates). While the Rule's current categories may seem simple, The Health Privacy Project notes that a "one-size-fits-all de-identification approach" does not, one the one hand, meet the diverse data needs of researchers and health providers, nor does it, on the other hand, provide sufficient protections in era of evolving data technologies.

Reference:

The Health Privacy Project, Center for Democracy & Technology. Encouraging the use of, and rethinking protections for de-identified (and “anonymized”) health data. Center for Democracy & Technology, June 2009. http://www.cdt.org/healthprivacy/20090625_deidentify.pdf

Related:

Knoppers BM, Saginur M. The Babel of genetic data terminology. Nat Biotechnol. 2005 Aug;23(8):925-7. PubMed PMID: 16082354.

Sharyl J. Nass, Laura A. Levit, and Lawrence O. Gostin, Editors; Committee on Health Research and the Privacy of Health Information: The HIPAA Privacy Rule; Institute of Medicine. Beyond the HIPAA Privacy Rule: Enhancing Privacy, Improving Health Through Research. Washington, D.C.: Institute of Medicine, The National Academies Press, 2009. http://www.nap.edu/catalog.php?record_id=12458

Other Stories in the News

Your Genes Aren’t Covered for That: One Year Later, Gaps in Genetic Discrimination Legislation Reveal the Challenges Ahead. Susannah Baruch, Science Progress. June 29, 2009.

FDA’s Current Ability to Regulate Genetic Testing Is Problematic, FDLI-AAAS Colloquium Attendees Say. Food and Drug Law Institute (FDLI) and the American Association for the Advancement of Science (AAAS) [Press Release]. June 22, 2009. http://www.fdli.org/press/pressrelease/062209.pdf

New Comparative Effectiveness Bill Enhances Dx, Genomics Focus. Matt Jones, GenomeWeb. June 18, 2009.

The GINA Law: Consumer Protection in a New Era of Genetic Testing Research Report. N. Lee Rucker, M.S.P.H., AARP Public Policy Institute, May 2009. http://www.aarp.org/research/health/prevention/fs156_gina.html

-- J.O.

Monday, March 31, 2008

Best Predictive Health Ethics Blogs - March, 2008

See February and January 2008 for prior reviews of the best predictive health ethics blogs. I have limited this month's installment to the ten posts. If, after reading these, you believe I've omitted a good post, please let me know. This month's blogs (listed chronologically) include some familiar names and a couple of new ones, including: Adventures in Ethics and Science, Synthesis, and Genetics & Health.

CF "success" story. Steve Murphy, Gene Sherpas. 1 March 2008.
Steve Murphy of Gene Sherpas seldom shies away from an ethical issue. Thus, when the NEJM reported the good news that "The number of infants born with cystic fibrosis in Massachusetts decreased by 50% from one four-year period to the next" (MedPage Today, 28 February 2008). Murphy acknowledged the upside, "children born with CF in the future may actually have less significant disease and may need to have less aggressive or less early interventions". He reminds us, however, that some doctors and patients now face heavy decisions:

The number of infants born with cystic fibrosis in Massachusetts decreased by 50% from one four-year period to the next according to this letter in the NEJM. They attribute this to the Newborn screening available in Massachusetts since 1999. … But here's what they don't say.....Preconception screening for this disease is important, but a highly personal choice. ... despite what the ACMG and ACOG say. There are some children who are doing just fine with CF. As for adults with CF like the 74 year old patient who I diagnosed a few years ago, I am certain they would not want to have never been born.

Should Researchers Share Data? Janet D. Stemwedel, Adventures in Ethics and Science. 3 March 2008.
Janet Stemwedel of Adventures in Ethics and Science comments on the widely discussed Andrew Vickers essay in The New York Times – "Cancer Data? Sorry, Can’t Have It" (22 January 2008). Stemwedel does an excellent job of highlighting the importance of data sharing for the success of translational research. She also acknowledges the research subject's contribution to this research and the exposure to risk that they have assumed. Stemwedel writes:

[R]isk is supposed to be offset by the benefits of the knowledge gained from the research. If that data sits unanalyzed, the benefits of the research are decreased and the risks undertaken by the human subjects are harder to justify. Moreover, to the extent that sitting on data instead of sharing it requires other researchers to go out and get more data of their own, this means that more human subjects are exposed to risk than might be necessary to answer the scientific questions posed in the research.

How Private is Private? Is Google a covered entity? Cheryl Lew, Women's Bioethics Blog. 4 March 2008.
After observing that HIPAA was not written with Google's emerging patient managed, personal health records in mind, Cheryl Lew of Women's Bioethics Blog wonders if additional government protections might be required. Lew writes:

[W]ill Google patients be subject to advertising spam or other intrusive advertisement adduced from their prescription lists? What guarantees that the “client” (read patient lists) won’t be sold to Pharma companies as yet another means of developing data bases about physician prescribing patterns? Who is going to regulate these issues? Google is a great search engine—I use it all the time! But I’m not sure I want to use it to manage my healthcare information.

Genomes of the Rich and Famous. Jesse Reynolds, Biopolitical Times. 5 March 2008.
Reacting, in part, to "Gene Map Becomes a Luxury Item" (Amy Harmon, The New York Times. 4 March 2008), Reynolds was the only blogger that I know about that picked up on the irony in drive towards celebrity genomics. Reynolds observes:

The prospect of biotechnology companies capitalizing on genetic information in order to develop profitable products without properly compensating the studied populations brings to mind biocolonialism, in which politically and economically vulnerable indigenous groups are exploited. But a growing convergence of genomics and information technology may lead, ironically, to the genetic "exploitation" of society's most elite.

The moral manipulation of Gattaca. Daniel MacArthur, Genetic Future. 6 March 2008.
Daniel MacArthur of Genetic Future is becoming a regular here in the "best ethics blogs" series. This month MacArthur introduces (with help from Black Belt Bayesian) us to an essay on one of the favorite films for bioethics classrooms—Gattaca:

A man is given strong medical evidence that he will die from a heart attack if exposed to the exertions of space travel, thus risking how own life and the lives of his crewmates. Ignoring this evidence, he fakes his way into astronaut training - and inexplicably, we cheer him on. How did the makers of Gattaca steer us towards this bizarre response?

The full essay, by Neven Sesardic and forthcoming in The Routledge Companion to Philosophy and Cinema, 2008, is available online: Gattaca (PDF – 209 KB).

Genetic testing - ‘recreational genomics’ or the future of diagnostics? Elaine Warburton, Genetics & Health. 18 March 2008.
Reacting to editorials in The New England Journal of Medicine and The British Medical Journal, Elaine Warburton of Genetics & Health acknowledges the limitations of the current methodologies, but notes that "introducing innovative genetic testing technology has to start somewhere". Warburton observes that 23andMe and other genomics companies are taking "huge financial and clinical risk[s] in bringing these tests to market". Warburton sees the uncertainty involved as merely a part of the ongoing cycle of research and development:

The tests are in their infancy and each of these companies are transparent in advising their customers of this fact. That said, massive scientific research continues to take place to build on the knowledge base of these tests, so that they may be refined. This process will never end.

Also see Warburton's related post: Ethical guidelines for whole genome studies. 26 March 2008.

The G.I.N.A. could be a bad thing for healthcare. William Gunn, Synthesis. 20 March 2008.
William Gunn of Synthesis was one of the few, perhaps the only, blogger brave enough to buck the current dogma on value of GINA. After reading Marcelino Fuentes's letter in Nature, "How genetic censorship would harm everyone" (PMID:18354455), Gunn agrees, but notes that: "This is all jumping the gun a little" (I hope that's not a pun) "because actual rock-solid, high-confidence correlations between a genetic feature and a disease are still rather rare, but one thing’s for sure: The better you see what’s ahead, the better you can plan for it (whether a insurance company or an individual), and having a good plan leads to better outcomes for everyone". Gunn, however, does not have a problem with acknowledging the potential inequities: "Everyone’s worried about enabling social injustice, but it can’t really be said that our current insurance system in which many are so under-served is really all that great to begin with, so let that temper your thoughts, as well".

So, what do you think? Are Fuentes and Gunn right?

Your personal health: Crowdsourcing healthcare - Pitfalls and possibilities. Deepak Singh, bbgm. 25 March 2008.
Although Deepak Singh writes more often about technology, his comments on the progress of personalized medicine and the feasibility of predictive health are usually thought provoking. In this post Deepak reacts to the many consumer privacy concerns and asserts a firm position on data ownership:

Whether it be personal genetics, or social health, our efforts must focus on consumer education, consumer privacy and perhaps most of all data ownership. Who owns the data? That is the most important question. As has been said before in these parts, we essentially need some sort of license for personal health information that establishes ownership with the person whom the data belongs to, i.e. you or I and who has access and under what circumstances. That’s where education comes in. We need access to a lot of patient information and genetic profiles for personalized medicine to be successful.

Well said, Deepak, but do you also think that patient control and "ownership" of data will improve the pace medical research?

Bringing the genome home ... but why? Sue Trinidad, Women's Bioethics Blog. March 26, 2008,
Sue Trinidad, also a regular here in the "best ethics" series, reflects on the need for oversight in the future of genetic and personalized medicine. Trinidad responds to "Genetic Testing Gets Personal" (Rick Weiss, The Washington Post. 25 March 2008.), which profiles the consumer genetics movement and the companies, like 23andMe and Knome, that are fueling it. Trinidad cautions:

Here's a fact about all these businesses: not a one of them is subject to FDA oversight. So all these claims about health benefits, etc., resulting from their services can be completely false (or at least, not based on evidence), and nothing can be done about it. If you read even a *tiny* bit of the scientific literature on genetics and "personalized medicine," you will be struck by how few experts make such claims, at least for the near term … In the meantime, though, consumers should be protected from grossly overstated or unproven claims. Whether that falls to FDA or to someone else (though the Consumer Protection Agency probably has enough work to do at the moment -- lead paint in toys, anyone?), it seems to me the Feds ought to step up on this one.

Will information on risk genes actually change behaviour? Doug MacArthur, Genetic Future. 31 March 2008.
In this second post form Genetic Future in this edition of the "best predictive health blogs", MacArthur reflects on a recent news article in Science (PMID:18369117). The article and MacArthur are among the few that have bothered to acknowledge what may be the true downfall of predictive, personalized medicine—the patient. To some extent everyone (with or without genetic information) knows and ignores health risks. MacArthur acknowledges this and observes:

[I]f it turns out that genetic information doesn't reduce risk behaviour in the real world, then the impact on public health of the hundreds of millions of dollars spent on complex disease genetics may actually be very small. I'm hopeful that this won't be the case - but it would be nice to have some actual evidence one way or the other....

- J.O.