Discuss the ethical issues of social networks and anytime,
PUBLIC HEALTH ETHICS
20. Kahn JP, Mastroianni AC, Sugarman J, eds. Ba/ond Consent: Seeking Justice in Research New York: Oxford University Press; 1998.
2 1 . Paasche-Orlow MK, Taylor HA, Brancad FL. Readability standards for informed-consent forms as compared with actual readability. N EnglJ Med. 2003,348(8):721-726.
22. Iltis AS. Timing invitations to partic- ipate in clinical research: preliminary versus informed consent / Med Philos. 2005;30(l):89-106.
23. Young DR, Hooker DT, Freeberg FE. Informed consent documents: increasing comprehension by reducing reading level./RÖ. 1990;12(3):l-5.
24. Hochhauser M. "Therapeutic mis- conception" and "recruiting doublespeak" in the informed consent process. IRB. 2002;24(1):11-12.
25. Roberts LW. Informed consent and the capacity for voluntarism. AmJ Psy- chiatry. 2002;159(5):705-712.
26. Appelbaum PS. Editorial: Missing the boat: competence and consent in psychiatric research. Am J Psychiatry. 1998:155(11):1486-1488.
27. Campbell J. "We are the evidence,'an examination of service user research in- volvement as voice. In: Wallcraft J, Schrank B, Amering M, eds. Handbook of Service User Involvement in Mental Health Research. New York: Wüey; 2009:113-137.
28. Gunsalus CK. The nanny state meets the inner lawyer: overregulating while underprotecting human participants in re- search. Ethics Behav. 2004;14(4):369-382.
29. Jeste DV, Palmer BW, Appelbaum PS, et al. A new brief Instrument for assessing dedsional capacity for clinical
research. Arch Gen Psychiatry. 2007; 64(8):966-974.
30. Agency for Healthcare Research and Quality. Researcher's certification of consent and authorization, 2010. Avail- able at: http://www.ahrq.gov/fund/ informedconsent/id'orm4.htm. Accessed June 23, 2011.
3 1 . Festinger DS, Marlowe DB, Dugosh KL, Croft JR, Arabia PL. Higher mag- nitude cash payments improve re- search follow-up rates without in- creasing drug use or perceived coercion. Drug Alcohol Depend. 2 0 0 8 : 9 6 ( l - 2 ) : 1 2 8 - 1 3 5 .
32. Del Vecdiio P, BIyler CR. Idendfying critical outcomes and setting priorities for mental health services research. In: WallCTaft J, Schrank B, Amering M, eds. Handbook of Service User Involvement
in Mental Health Research. New York:
Wiley: 2 0 0 9 : 9 2 - 1 1 1 .
33. Fisher CB. Relational ethics and re- search with vulnerable populations. In: National Bioethics Advisory Commission, ed. Research Involving Persons With Men- tal Disorders That May Affect Decision- Making Capacity. Commissioned Papers. Rockvillle, MD: National Bioethics Advi- sory Committee: 1999:29-49.
34. Israel BA, Schulz AJ, Parker EA, Becker AB. Review of community-based research: assessing partnership ap- proaches to improve public health. Annu Rev Public Health. 1998;19:173-202.
35. DuBois JM, Bailey-Burch B, Bustillos D, et al. Ethical issues in mental health research: the case for community en- gagement Curr Opin Psychiatry. 2011 : 24(3):208-214.
Ethical Issues in Health Research With Novel Online Sources Effy Vayena, PhD, Anna Mastroianni, JD, MPH, and Jeffrey Kahn, PhD, MPH
Health-related research is in- creasingly drawing on novel sources of online data, such as crowdsourced informa- tion about disease outbreaks, consumer-supplied informa- tion provided to health or well- ness Web sites, Internet search queries about personal health, and social network postings that identify health behaviors.
We offer examples of online sources and their uses, iden- tify ethical and policy issues they generate, and formulate key questions for future dis- cussion and investigation.
Further work in this area will require cross-disciplinary col- laboration to develop ethics and policy guidance for the ethical use of these novel data sour- ces in health-related research. {Am J Public Health. 2012;102; 2225-2230. doi:10.2105/AJPH. 2012.300813)
A DRAMATIC RECENT DEVEL-
opment in health-related research, and public health research in par- ticular, is the emergence in multi- ple forms of unprecedented uses of online health information. These uses indude undertaking and improving infectious disease surveillance'; understanding pat- terns of chronic disease^; probing population genetics''; assessing health behavior*; and identifying and recruiting potential partid- pants for clinical research.^ Some newer data collections rely on in- formation voluntarily provided by individuals, which may then be used in research with or without their knowledge. There are also approaches that rely on "mining" data aggregated from individuals who are likely unaware that their information is being gathered or tised for research purposes.
Examples indude data sets cre- ated from analysis of aggregate Internet search behaviors to identify illness trends (e.g., Goo- gle Trends, Google Insights for Search) and mining personal in- formation from sodal networking sites to characterize health be- haviors (e.g., Facebook, Myspace, Linkedin). Data sets are also be- ing created from Web sites that use both aggregate and individual user-provided health data, such as PatientsLikeMe.com. These public and private sour- ces of health data, used sepa- rately or in combination, create new opportunities to address health issues and will be an in- creasingly valuable tool for a wide range of health-related research.
The trend toward innovative uses of online data for health-related
research may well have started with the large amounts of genetic and genomic data collected worldwide in many separate re- search projects, some collabo- rating but others working in iso- lation. The data sets generated by resecirchers are increasingly recognized as an important re- source for so-called secondary research purposes—that is, re- seeirch purposes beyond those proposed when the information was collected. Sophisticated bio- informatics tools allow for in- creasingly larger and easier storage and combination of data sets for future analysis,® includ- ing the linkage of data to elec- tronic medical records and other sources of health information. Utilizing the growing amounts of information in such data sets is likely to aid health-related
December 2012, Vol 102, No. 12 I American Journal of Public Health Vayena et at. \ Peer Reviewed | Public Health Ethics | 2225
PUBLIC HEALTH ETHICS
Glossary of Terms
Data mining; Application of statistical methods to diveise data sets, to collect data, to identify or clarify associations or relationships within the data, and to estimate the strength of associations
or relationships'"
Crowdsourcing; Open call to large number of people (crowds) to provide ideas, information, and solutions (i.e., Wikipedia)
Interoperability of health data sets: The ability of different health data systems to exchange information accurately and to use the information that has been exchanged'*
Search volume patterns; Amount of searches that have been performed for the search term, relative to the total number of searches that have been performed in the search engine
Bioinformatics tools: Computational and information technology tools used in biology and medicine
research as well as delivery of care
in new and important ways. For
example, studies have demonstrated
such potential in genomics and
pharmacogenomics research,'''^®
and the US National Institutes of
Health Mtiltiplex Initiative currently
under way is researching how to
most effectively integrate the results
of a panel of 15 genetic tests (i.e.,
a multiplex) with the health care
delivered to large populations of
patients who are members of in-
tegrated health care systems.®
Increased use of, and access to,
the Internet is making both the
collection of online information
and access to it significantiy easier,
with both positive and negative
implications. Individuals may
search the W e b for health-related
information, may post their per-
sonal heeilth status and behaviors
via Facebook, or may report on
disease outbreaks or other indi-
vidual or public health-related
aspects of local environments
via crowdsourcing sites. Such in-
formation can be available to
those engaging in public health
surveillance as well as health-
related research.'°"'^
User-supplied information for
research raises questions of data
quality and acctiracy, but also
raises issues regarding the terms
tinder which such information can
be used in research. It is our hope
that this issue-spotting paper will
highlight the need for broader
discussion and will help stimulate
the future elaboration of principles
to guide the development of policy
and practice for the ethically ac-
ceptable sourcing and use of
online data for a wide range of
health-related research (see the
box on the next page).
HEALTH INFORMATION SHARED BY INDIVIDUAL USERS ON THE WEB
Individuals are voltmtarüy
sharing health and health-related
information about themselves
(e.g., behavior, nutrition, disease
status and treatment, genomics
data) on various Web sites. Al-
though the self-reported nature of
the information challenges the
standard approach to research
employing carefully controlled
studies, approaches relying on
"crowdsourcing"'®'^ and other
grotip-aeated collections of infor-
mation are dearly at the leading
edge of research. Some examples of
sites and how data can or have been
used for health research follow.
PatientsLikeMe.com is a Web
site that invites individuals to share
their health- and iUness-related
ejqjeriences as a way of creating user
communities. The site states that the
company's business model rests on
its ability to sell access to these data
to pharmaceutical companies and
others as a research resource.'® Data
sets collected from users have also
been used by the company's own
research teams. For example,
PatientsLikeMe published an arüde
proposing that self-reported data can
successfully be mined to assess drug
performance as well as potential ofF-
label uses.'® In another example of
research involvement by a tiser
' community, individuals with amyo-
trophic lateral sderosis registered
with PatientsLikeMe.com dedded
to take lithium to test its impact
on symptom relief and disease pro-
gression. The site provided Web
tools for data collection and a matdi-
ing algorithm to identify contiuls.
The study was published in Nature
Biotechnobgy, and refuted the daim
that lithium has an effect on disease
progression. In addition to these
findings, the authors argued that the
site can help accelerate medical dis-
covery in a range of other areas.^°
Other sites encourage users to
share their genetic information and
other health-related information to
enable community stijdies to take
place. Genomeracom's current on-
going studies indude comparisons
of dosages of vitamin supplements.
sleep patterns, and other variables
according to genetic variations.^'
OpenSNP.com also provides a plat-
form for users to upload their di-
rect-to-consumer genetic test re-
sults, share them with others, and
make them searchable by re-
searchers.^^ Direct-to-constmier ge-
netic testing companies such as
23andMe offer genome-wide
screening services directly to the
public^'' Services indude disease
risk probability, carrier status, pre-
dicted responses to certain drugs,
and ancestry information. Genome
data are maintained for future
scanning when new genetic assod-
ations are added to the testing
panel. The company uses the site to
recruit users into disease-specific
research projects via its research
arm, 23andWe. As a result,
23andMe daims a database of dose
to 100 0 0 0 genetic profiles, all of
which could potentially be used for
research. It has publidy dedared
that it is becoming a research com-
pany ,̂ ^ and has already published
a significant genomewide assoda-
tion report on Parkinson's disease,
which identified 2 new genetic as-
sodations (and replicated others).^^
In a somewhat different exam-
ple, individuals may intentionally
shcire personal health information
on the W e b yet may be unaware
of potential or actual research
2226 I Public Health Ethics | Peer Reviewed | Vayena et al. American Journal of Public Health I December 2012, Vol 102, No. 12
PUBLIC HEALTH ETHICS
Examples of Web Sites With Data That Can Be Used for Research
Google Insights: http://v(ww.google.com/insights/search
Google Trends: http://vmw.google.ch/trends
23an(ln;ie.com: http://www.23andme.com
Patientslikeme.com: http://www.patientslikeme.com
Healthmap.com: http://www.healthmap.com
Genomera: http://www.genomera.com
openSNP: http://www.opensnp.org
Free Google analysis tool that shows volume of searches over time, specific regions, and includes forecast for future
searches of the term
Free Google analysis tool that shows volume of searches for a term; also refers to a project within Google to use search
terms as a way of predicting trends in the real world
Online direct-to-consumer personal genomics service provider; owns and maintains biobank and database used for
pursuing genomewide association studies
Social networking site for patients, where individuals voluntarily share information about their illness or disease
symptoms, health behaviors, treatments, and the like; owner of the site sells access for research
Real-time surveillance of disease outbreak and public health threats that uses information provided by individuals
"on the ground" reporting local conditions; a form of crowdsourcing for public health surveillance
Web site that stores genetic profiles and other data uploaded by useis, and supports the conduct of community health studies
Web site that stores genetic profiles, which can be searched by users as well as researchers for studies
uses of their information because the primary use of the site is unrelated to research. Some sites that fall into this category are designed in part to provide tai- lored health feedback to users, but their business model is based on sharing the data they collect with commercial entities for re- search and other purposes. Al- though references to potential research uses of the collected in- formation may be included in the terms of use or service agreements and privacy policies of such com- panies, in many cases the informa- tion tends to be nonspecific. It re- mains undear whether consumers understand or are even aware that their supplied information can be used in research, whether for health-related svirveillance, re- search, or commercial purposes. One typical site, RealAge.com, claims more than 10 million mem- bers whose user-supplied health information is collected for the purpose of creating personalized health newsletters and information for delivery to tisers. Individual information is also combined with
various other data sets and shared with commercial entities, induding pharmaceutical companies, for re- search and marketing purposes.^^
PERSONAL HEALTH- RELATED INFORMATION COLLECTED FROM WEB
Other research approaches rely on collecting and analyzing existing personal information on the Web to extract potentially useful health- related information, rather than relying on users to supply such information to spedfic health- related sites. These approaches may rely on health-related infor- mation posted or shared for non- research purposes (e.g., via sodal networking sites) or information that can be gleaned from basic user activities on the Web. Some of the sites that track activity and behavior on the Web collect data primarily for targeted advertise- ments. These data can also be used for research purposes. As wül be discussed in more detail, public health data can potentially be de- rived without the direct knowledge
of partidpants (e.g., whether from search queries on health conditions or from mining information posted to sodal networks).
A collaborative group from Goo- gle and the US Centers for Disease ContiDl and Prevention (CDC) pub- lished a report in 2009 on their combined effort to use population data to better predict disease out- breaks.^^ In a project called Goo- gle Flu Trends the team applied algorithms to specific search terms related to influenza and its symp- toms, and was able to predict flu outbreak in the United States 2 weeks earlier than standard CDC modeling approaches. Subsequent analyses have examined the accu- racy of this approach and its appli- cations, but related approaches have been reported for predicting flu outbreaks as well as prevalence of other illnesses and diseases.''̂ *'̂ ° Similarly, HealthMap''' monitors pubUdy available Internet resources (and crowdsourced information) to detect disease outbreaks and to provide surveillance for public health threats; and Google Insists searches volume patterns across
location and time to determine, for example, disease inddence.'^'^^"''^
Sodal networking sites induding Facebook have aeated huge and growing virtual communities. Some of the information posted and shared by users relates to their health issues and health-related be- haviors. Research drawing on such data may indude reported risk be- haviors for infectious diseases such as HIV or illegal behaviors such as domestic 'violence, child abuse, or use of illidt drugs. Mining these sites might also gauge the effectiveness and impact of health promotion and prevention efforts, such as public health campaigns targeting smok- ers, obesity, and substance abuse (see the box on
LOOKING FORWARD
Debate concerning the use of individual health information in research has traditionally focused on balancing a variety of factors: informational privacy of individ- uals, the population-oriented goals of public health and biomédical research, the potential benefits to
December 2012, Vol 102, No. 12 | American Journal of Public Health Vayena et at. | Peer Reviewed ) Public Health Ethics | 2227
PUBLIC HEALTH ETHICS
the individuáis themselves, the central importance of individual consent in research, various as- pects of "vulnerabiüfy" in potential research partidpants, and the communify-level issues raised by research on groups. The descrip- tive account sketched previously of the types of available online sources that may be increasingly useful for health-related research and various approaches to re- search "participation" provide a springboard for addressing in a thoroughgoing fashion the ethics and policy questions generated by the use of these technology-driven and user-created data sources. We suggest research directions for addressing the ethics and policy issues raised by the research ap- proacjies previously discussed. With 2 key considerations in mind, namely protection of re- search partidpants and facilitation of high-quaüfy and ethically ac- ceptable research, we outline an agenda of ethics and policy issues that we believe should form the basis of further and future analysis.
The key safeguard for protec- tion of researcdi partidpants in the context of biomédical and public health research is informed con- sent. There are various interpre- tations of what "informed consent" means in the context of online data collection. For example, does reference to potential research uses of personal data in a Web site's terms of service constitute acceptable consent? Do checkbox agreements on Web site logins that indude permission to access users' online information represent genuine and ethically acceptable informed consent? Are "opt-in/ opt-out" models any better?
Qiaraderizing the content and implementation of informed con- sent in these novel contexts is essential for protecting partid- pants and stii-eamlining research review. Essential questions for ex- amination indude the following: (1) What criteria are important in determining whether and under what conditions consent is re- quired (e.g., is consent required from partidpants whose existing online data can be used for research in aggregated and unidentified forms)?̂ ® (2) Is fhe purpose of re- search (e.g., public health vs research for mariceting, recruiting, or other business-related motives) a factor in determining the need for consent or the form it should take? (3) Is con- sent required when patients or con- sumers initiate the research project themselves (e.g., throu^ crowd- sourcing approaches such as the PatientslikeMe.com lithium study)?
These novel sources and re- search applications also require assessment of risks to partidpant privacy and confidentiaüfy. De- spite legal protections and tech- nologies that purport to ensure online data protection, the in- creasing interoperabilify of data systems creates risks of securify breaches. Creating online systems that ensure data securify and yet allow interactive functions re- mains an ongoing challenge. In- creasing evidence points to the insuffident protection of online privacy because of both the de- velopment of informatics tools and the growth of online personal data, including data that are self- reported.''^ Two themes in par- ticular are worth exploring. (1) Has our notion of privacy, espe- dally with reference to the online
data world, changed, and, if so, how should these changes affect privacy polides and practices? (2) Do individuals who provide data online understand the issues related to data securify and are they aware of the state of online privacy and the risks to it entailed by the online environment?
In conventional biomédical and public health research, research ethics committees bear the pri- mary responsibilify for ensuring that informed consent procedures and processes are adequate, that risks are minimized and balanced against intended benefits, that partidpant selection is equitable and vulnerable populations are adequately protected from risks of harm and exploitation, and that the benefits of research are equi- tably distributed. Among the questions in need of greatest at- tention is whether researcJi uses of online health data are or should be subjeci to the various laws and polides regarding the ethical con- dud of research on human par- tidpants. Confusion in this area has already generated a plea for darification from the sdentific communify. PLoS Genetics pub- lished an editorial accompanying a research artide by the 23andMe research group that had analyzed genetic data of its customers without review by an institutional review board or other ethics re- view committee. The editorial was a call for darify and need for a standardized review approacJi.''® Thus, a key question is whether the current models of biomédical and public health research review are appropriate for health re- search involving online data sources, or is there a need for the
creation of alternative and more appropriate review processes? Furthermore, do the studies that are partidpant-initiated or participant-driven require ethics review and, if so, what sort of oversight do they require?
Although there is widespread Internet access in the developed world, the digital divide remains a concern at the global level. Sev- eral countries report Internet us- age percentages in single digits. Does health research based on such data have the inherent prob- lem of bias and questionable gen- eraüzabiüfy? This is not merely a technical issue that can be over- come by study designs that reduce or eliminate bias or focus findings on specific populations. It is also an ethical issue because particular groups are likely to be deprived of potential research benefits. Based on the general belief that health research using online information will translate into better health pol- icy or medical advances, lack of online information could exacerbate existing health inequities."*® Con- versely, if data were to be œUected fi-om countries with good Internet access but fwor track record of privacy and personal data protec- tion, questions arise regarding how such onUne data can be appropri- ately used. Attempts at international harmonization of codes and prind- ples will need to acknowledge or account for such issues.
Research outcomes may have commerdal value and hence may raise an array of issues related to intelledual properfy and patenting. Althou^ most of these issues wül be addressed by existing laws, pressing moral questions also re- main. To œlled large data sets.
2228 I Public Health Ethics | Peer Reviewed | Vayena et al. American Journal of Public Health | December 2012, Vol 102, No. 12
PUBLIC HEALTH ETHICS
many organizations appeal to partic-
ipants' altruism and sense of sodal
responsibility. Other commerdaliz-
able outcomes may result from the
use of data provided by individuals
who are unaware of their indusion in
research. In the case of outcomes
that yield intelledual property, it is
necessary to address the obligations
researdiers have to research partid-
pants. Laws related to data ownership
and intellectual property also are
implicated as the collected data and
biological samples become assets of
a company (e.g.. How are those assets
treated vviien a company files for
bankruptcy?). And, importantly, the
blurring of national and state bound-
aries in relation to Web-based activ-
ities and companies raises jurisdic-
tional issues that must be resolved.
The law wül be a vital factor in
formulating appropriate poMdes and
ensuring the success of efforts to use
online sources in an ethical manner.
As mentioned previously, the use of
online health data in research im-
plicates a wide range of privacy laws,
such as those related to health in-
formation, Internet use, and data
protection. Online sources and their
applications will also challenge tra-
ditional legal protections for re-
search partidpants' rights and wel-
fare, induding those related to
research oversi^t, partidpant re-
cruitinent, informed consent, and
risks to partidpants related to po-
tential identifiability and publication.
CONCLUSIONS
If forecasts are correct and med-
idne is increasin^y characterized
by the "P4 approach"—preventive,
partidpatory, predictive, personal-
ized—we can expect the amount of
such electronic health infonnation
to grow exponentially.''° This will
be coupled with growing interest
by consumers and patients in
having access to their health data
and in being in control of its
potential tises, induding re-
search.'"'^^ In response to these
developments, it is necessary to
adopt a proactive approach to the
important ethics and policy issues
that have not yet received ade-
quate attention. Focus on these
issues is espedaUy timely in the
United States, as regulations con-
cerning protection of humein par-
tidpants are slated for revision for
the first time in decades. The pro-
posed changes i n d u d e new and
strengthened data protection and
informational risk protection pro-
visions as well as approaches to
addressing the fact that previously
deidentified data (and all DNA
specimens and sequences) can in-
creasingly be linked to identified
individuals through sophisticated
information technologies. The
proposal also indudes strength-
ened requirements for consent in
research involving any biosped-
mens (on tiie groimds that all are
identifiable) as well as consent re-
quirements for any current or
future research use of collected
information, identifiable or not.'*''
US President Barack Obama re-
cently proposed a Consumer Pri-
vacy Bill of Rights for online data
designed to provide consumers
with control of online data uses
and i n a e a s e d data protection.
How these proposed changes
to the regulations for the protec-
tion of human partidpants and to
online data privacy protections^*
WÜ1 affect research uses of online
health-related information remains
to be seen. Under any scenario.
researchers will need to pay doser
attention to consent as well as to
data privacy and protection. Simi-
larly, the recentiy proposed revi-
sion to the European Union data
privacy regime will likely affect the
handling and protection of many
categories of digital infonnation,
induding health-related informa-
tion used in research.*^ Policy
changes in Europe and the United
States are hardly independent
from each other because online
information does not reside
within geographical borders and
both policymakers and analysts
have suggested that the sti-ongest
governmental polides will probably
dictate behavior the world over.
The unprecedented availability
of online information creates chal-
lenges that must be addressed so
that valuable health-related research
can be undertaken in ethically ap-
propriate and legally sound ways, all
within a dear, overarching policy
framework. Achieving these multi-
ple goals requires focused examina-
tion of the issues outlined here, by
experts in research ethics and policy,
privacy law, health informatics,
public health, and pharmaceutical
and biotechnology industries, as well
as by representatives of consumer
communities. There is currently no
commonly accepted ethical and
policy guidance when such health
research is conducted. Deeper
thinking on such issues will require
that experts work across disdplines
to integrate traditional research
principles with a developing body of
research and scholarship.''®^®
The policy regime for health re-
search is buflt on a foundation of
trust in oversi^t and protections, to
provide assurance that the benefits
can be realized in ways that avoid
risk to individuals and grotaps. The
use of online information for health-
related research holds out the pros-
pect of a new paradigm of research
that may also necessitate a new par-
adigm in research protections. Now
is the time to address these issues
because ignoring them can eventu-
ally lead to the undermining of the
public's very trust on which the
research enterprise is based. •
About the AuUiors E^ Vayena is with the Institute of Bio- medical Ethics, University of Zürich, Zürich, Switzerland. Anna Mastroianni is with the School of Law and Institute for Public Health Genetics, University of Washington, Seattle. Jeffrey Kahn is with the Berman Institute of Bioethics, Johns Hopkins University, Balti- more, MD.
Correspondence should be sent to Effy Vayena, PhD, Institute of Biomédical Ethics, University of Zürich, Pestalozzistrasse 24, 8032 Zürich, Switzerland (e-mail: vwjena@ ethikuzkch). Reprints can be ordered at http://www.ajphorg by clicking the 'Reprints" ¡ink.
This article was accepted March 20, 2012.
Contributors All authors contributed to the design of the article, literature review, and write up.
Acknowledgments We would like to thank the anonymous reviewers and the American Journal of Public Health editor for their insightful comments.
Human Participant Protection Human partidpant protection was not needed because research and analysis reflected in this artide are based on literature reviews.
References 1. Cook S, Conrad C, Fow!!<es AL, et al. Assessing Goog!e Flu Trends perfor- mance in the United States during the 2009 innuenza virus A (HlNl) pan- demic. PLoS ONE. 2 0 n ; 6 ( 8 ) : e 2 3 6 1 0 .
2. Weitzman ER, Adida B, Kelemen S, Mandl KS. Sharing data for public health research by members of an international
December 2012, Vo! 102, No. 12 | American Journal of Pubüc Healthi Vayena et al. \ Peer Revievned | Pubüc Hea!tb Ethics | 2229
PUBLIC HEALTH ETHICS
online diabetes sodal network. PLoS ONE. 2011:6(4):el9256.
3. McCarty CA, Qiisholm RL, Chute CG, et aL The eMERGE Network: a consortium of biorepositories linked to electronic medical records data for conducting genomic studies. BMC Med Genomics. 201^,4.13.
4. Ramo DE, ProchaskaJ. Broad readi and targeted recruitment using Facebook for an online survey of young adult substance use. J Med Internet Res. 2012:14(l):e28.
5. Allison M. Can Web 2.0 reboot clinical trials? Nat BiotechnoL 2009:27 (10):895-902.
6. Karp DR. Carlin S, Cook-Deegan R, et al. Ethical and practical issues assod- ated with aggregating databases. PLoS Mea 2008:5(9):el90. Available at: http:// www.plosmedidne.Org/artide/info:doi/ 10.1371/joumal.pmed.0050190. Ac- cessed December 15, 2 0 1 1 .
7. Wilke RA,XuH, Denny JC, et al. The emerging role of electronic medical re- cords in pharmacogenomics. Clin Phar- macol Ther. 2011:89(3):379-386.
8. Denny JC, Crawford DC, Ritdiie IVID, et al. Variants near FOXEl are assodated with hypothyroidism and other thyroid conditions: using electronic medical records for genome- and phenome-wide studies. AmJ Hum Genet 2011:89(4):529-542.
9. National Institutes of HeaMi. The MuW- plex Initiative. Available at: https://multçlex. nih.gov. Aocessed December 10, 2011.
10. Eysenbach G. Infodemiology and infoveillance: tracking online health in- formation and cyberbehavior for public health./Im/ft-ei)Med. 2011:40(5, Suppl 2):S154-S158.
11. Chew C, Eysenbach G. Pandemics in the age of Twitter: content analysis of tweets dtiring the 2 0 0 9 H l N l outbreak. PLoS ONE. 2010:5(11):el4118. Avail- able at: http://www.plosone.org/ardde/ info%3Adoi%2Fl 0.1371 %2Fjounial.pone. 0014118. Accessed December 1,2011.
12. Chuñara R, Andrews JR, Brownstein JS. Sodal and news media enable estima- tion of epidemiological patterns early in the 2010 Haitian cholera outbreak. Am J Trop Med Hyg 2012:86(l):39-45.
13. Breyer BN, Sen S, Aaronson DS, et al. Use of Google Insights for Search to track seasonal and geographic kidney stone inddence in the United States. Urology. 2011:78(2):267-271.
14. American Statistical Assodaüon. Data Mining FAQ. Available at: http://www. amstatorg. Accessed February 2 1 , 2012.
15. Heubusch K. Interoperability: what it means, why it matters./AWAM. 2006: 77(l):26-30.
16. Prainsack B. Voting with their mice: personal genome testing and the "partid- patoiy tum" in disease research. Account Äes. 2011:18(3):132-147.
17. Freifeld CC, Chuñara R, Mekaru SR, et al. Partidpatory epidemiology: use of mobile phones for community-based health reporting. PLoS Med 2010: 7(12): e l 0 0 0 3 7 6 . Available at: http://www. plosmedicine.org/artide/info%3Adoi% 2F10.1371 %2Fjoumal.pmed. 1000376. Accessed December 1, 2 0 1 1 .
18. How does PatientsLikdVle make money? Available at http://www.palientslikemecom/ help/faq/Coiporate#m_money. Accessed December 13. 2011.
19. Frost J,OkunS,Vaughan TE, Heywood J, Wicks P. Patient-reported out- comes as a source of evidence in off-label prescribing: analysis of data from Patient- sLikMe.JMed Internet Res. 2011:13(1): e6. Available at: http://www.jmir.org/ 2 0 1 1 / l / e 6 . Accessed October 10, 2 0 1 1 .
20. Wicks P, V a u ^ a n TE, Massa^ MP, Heywood J. Accelerated clinical discovery using self-reported patient data collected online and a patient-matching algorithm. Nat BiotechnoL 2011:29(5):411-414.
2 1 . Genomera. Available at: http:// genomeracom/studies. Accessed March 2 , 2 0 1 2 .
22. OpenSNP. Frequently asked questions. Available at: htlp://opensnp.org/faq# sdentists. Accessed Febmary 28, 2012.
2 3 . 23andMe genetics just got personal. Available at: http://www.23andme.com. Accessed October 10, 2 0 1 1 .
24. Timmerman L. 23andMe brings down the price of consumer genetic tests, builds up relations with big pharma Xconomy. May 24, 2 0 1 1 . Available at: http://wwwAConomy.com/san-francisco/ 201 l/05/24/23andme-moves-beyond- simple-consumer-dna-sequendng-sets-sight- on-research. Accessed October 20, 2 0 1 1 .
25. Do CB, Tung JY, Dorftnan E, et al. Web-based genome-wide assodation study identifies two novel lod and a sub- stantial genetic component for Parkin- son's disease. PLoS Genet. 2011:7(6): e l 0 0 2 1 4 1 . Available at: http://www. plosgenetics.org/ardde/info%3Adoi% 2F10.1371%2Fjoumal.pgen.l002141. Accessed December 12, 2 0 1 1 .
26. Clifford S. Online age quiz is a window for drug makeiî. The New York Times. March 25.2009. Available at: http://www.
nytimes.com/2009/03/26/technology/ intemet/26privacylitml?ref=todayspaper. Accessed October 12, 2 0 1 1 .
27. GinsbergJ, MohebbiMH, PatelRS, et al. Detecting influenza epidemics using search engine query data. Nature. 2009: 457(7232):1012-1014.
28. Valdivia A, Lopez-Alcalde J, Vicente M, et al. Monitoring influenza activity in Europe with Google Flu Trends: com- parison with the findings of sentinel phy- sidan networks—results for 2 0 0 9 - 1 0 . Euro SurvalL 2010:15(29):19621.
29. Ortiz JR,2îiouH, Shay DK, et al. Monitoring influenza activity in the United States: a comparison of traditional surveil- lance systems with Goo^e Flu Trends. PLoS ONE 201 l:6(4):el 8687. Avaflable at: http://www.plosone.org/artide/info% 3Adoi<'/o2F10.137 l%2Fjoumal.pone. 0018687s. Accessed September 6, 2 0 1 1 .
30. Dugas AF, Hsieh YH, Levin SR, et al. Goo^e Flu Trends: correlation with emer- gency department influenza rates and crowding metrics. Clin Infect Dis. 2012; 54(4):463-469.
3 1 . Healthmap.com. Available at: http:// healthmap.org/en. Accessed October 23, 2 0 1 1 .
32. Althouse BM, Ng YN, Cummings DAT. Prediction of dengue inddence using search query surveillance. PLoSNe^ Trap Dis. 2011:5(8):el 258. Available at: htlp://www.plosntds.org/artide/info:doi/ 10.1371/jounial.pntd.0001258. Ac- cessed December 4, 2 0 1 1 .
3 3 . Lee BK. Epidemiologie research and Web 2.0—the user-driven Web. Epidemi- ology. 2010,21(6)160-7 6'i.
34. Noll-HussongM,LahmannC Whiplash and Werther effect: the potential of Google Insights for Search for medical research and public health Dn German]. Fortschr Neurol Psydáaír. 2011:79(6):340-344.
35. Walcott BP, Nahed BV, Kahle KT, Redjal N, Cotmians JV. Determination of geographic variance in stroke prevalence using Internet search engine analytics. Neurosurg Focus. 201 l:30(6):el9.
36. Rothstein MA. Is deidentification suffident to protect health privacy in re- search? AmfBioetk 2010:10(9):3-l 1.
37. U F, Zou X, Liu P, et al. New threats to health data privacy. BMC Bioinfor- matics. 2011:12(Suppl 12):S7.
38. Gibson G, Copenhaver GP. Consent and Internet-enabled human genomics. PLoS Genet 2010:6(6):el000965. Avail- able at: http://www.plosgenetics.org/
artide/info:doi/l 0.1371/joumal.pgen. 1000965. Accessed December 3, 2 0 1 1 .
39. Bustamante CD, De La Vega FM, Burchard EG. Genomics for the world. Nature. 2011:475(7355):163-165.
40. Hood L, Friend SH. Predictive, personalized, preventive, partidpatory (P4) cancer medicine. Nat Rev Clin Oncol. 2011:8(3):184-187.
4 1 . Terry SF, Terry PF. Power to the people: partidpant ownership of clinical trial data. Sd Transi Med 2011:3(69):69cm3.
42. HealthDataRights.org. A dedaration of health data rights. Available at: http:// wwwliealthdatarights.org. Accessed Feb- ruary 2 1 , 2012.
4 3 . The Office of the Secretary, Depart- ment of Health and Human Services, and the Food and Drug Administration. Ad- vance notice of proposed rulemaking. Human subjects research protections: en- hancing protections for research subjects and reducing burden, delay, and ambi- guity for investigators. Federal Register. July 26, 2011:76(143):44512-44531.
44. Consumer data privacy in a networked world: a framework for protecting privacy and promoting innovation in the global digital economy. Washington, DC: The White House: Febmary 2012.
45. Proposal for a regulation of the Etiropean Parliament and the Coimcil on the Protection of Individuals with regard to the processing of personal data and on the free movement of such data (General Data Protection Regulation). Brussels, Belgium: Etiropean Commission: 2012. Available at: http://eceuropaeu/justice/ data-protection/document/review2012/ com_2012_1 l_eapdf Accessed Febmary 2 1 , 2 0 1 2 .
46. Frankd MS, Siang S. Ethical and legal aspects of human subjects research on the Intemet: a report of a workshop, June 10-11, 1999, Washington, DC. Washington, DC: American Association for the Advancement ofSdence: 1999.
47. McKee HA, Porter JE, eds. The Ethics of Internet Research: A Rhetorical Case- Based Process. New York, NY: Peter Lang Publishing; 2009.
48. Adamick J, Buchanan E, Fountain J, Goncalves MS, Profères N. Advancing ethical research across disdplines. Na- tional Sdence Foundation Directorate of Sodal, Behavioral, and Economic Sd- ences SBE 2020: white papers. Available at: http://www.nsf.gov/sbe/sbe_2020/ 2020_pdfs/AdamickJessica_l 39.pdf Accessed November 15, 2 0 1 1 .
2230 I Public Health Ethics | Peer Reviewed | Vayena et al. American Joumal of Public Health | December 2012, Vol 102, No. 12
Copyright of American Journal of Public Health is the property of American Public Health Association and its
content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for individual use.