For some of the epidemiologic challenges we are likely to face, such as anticipating the spread of new viral strains and drug-resistant bacteria, consent-based models are scientifically inappropriate for the research questions being asked. Validity depends on the characteristics of the sampling criteria used in compiling the database: if data subjects are self-selected, each epidemiologic analysis will likely require a separate analysis of the impact of the self-selection factors on the research findings. In some cases, this will mean that it is not possible to obtain a valid answer to an important research question.
Research using information collected for other purposes, such as health care delivery or health benefits administration, is critically important as we enter into the century of discoveries based on genomic science. Pinpointing differences in health or health care quality based on geography, demography, health history, or co-morbidities will become ever more important for clinical research, for public health planning, and for ensuring access to appropriate care. We do not have the luxury of time and resources to collect consents and obtain data from volunteers, and evaluate the validity of the sample for testing every unique data hypothesis.
Public health surveillance is typically conducted under specific laws authorizing or requiring the collection of certain types of data in the public interest. (See, for example, Chapter 3 this volume.) The new medical privacy regulations, for example, have explicit exemptions from the prohibitions on disclosure where the data are being collected under various public health surveillance laws.16 In enacting the mandatory reporting laws the legislature has been persuaded that the individual's interest in privacy can be achieved in other ways that are not anathema to the public interest in pharmacovigilance and other public health surveillance. They have required public health authorities and regulated entities to simultaneously protect the privacy interests of individuals while making the requisite reports and appropriately using and safeguarding the collected data.
But most epidemiologic studies do not have the legislative protections of public health surveillance. Follow-up studies of drug safety, confidentiality issues generally are covered by different laws and regulations than spontaneous reporting aspects of pharmacovigilance. The challenge in North America and the European Union is to create a system in which patient needs for confidentiality protections can be achieved while also facilitating important public health research. If society moves too far in the direction of providing absolute protection to seal off access to health information from secondary uses, such as
formal studies of drug safety, then we are at risk of eroding the information foundation that supports public health planning and health care quality, including societal judgements on the benefits and risks of medications.
Follow-up studies of drug or devise safety fall into the broad general category of "research", which is defined in the Common Rule as "a systematic investigation ... designed to develop or contribute to generalizable knowledge''.17 Instead of an exemption from the prohibitions on use or disclosure of patient data, the regulation subjects disclosure or use of information for research purposes to an entirely new patient authorization process.18 In other words, the same attention to the public interest in both privacy and research results that is seen in event reporting laws is not evident in US laws regulating data access for epidemiologic and outcomes research.
Admittedly, the "worst case'' damage to a given individual from the disclosure and misuse of personal information may be highly significant. This potential damage or risk from the non-research misuse of personal information is precisely the same risk that adheres in pharmacovigilance and public health reporting. Indeed, with respect to the types of conditions that often are the subject of mandatory public health reporting—sexually transmitted diseases (STDs), child abuse, substance abuse—the potential damage from stigmatization or prejudice arguably is at its greatest.
Was this article helpful?