PHOTOGRAPHY: JIM SHIVE ILLUSTRATION: PAUL A. BELCI
|
Patient registries have been gaining attention in recent years as health care providers, payers, and regulators have realized
they can produce real-world data that more accurately depict the safety, quality, performance, and effectiveness of drugs
and devices. There has been a growing movement to raise the quality of observational research, including patient registries,
so that it can be used for more purposes—from labeling and advertising to regulatory submissions and reimbursement decisions.
Recent efforts by several groups have brought forth important guidelines and principles to develop or evaluate observational
study data. While in clinical trials, the CONSORT statement1 criteria have set the bar and standardized practice. However, for observational research there is not a single definitive
document, but rather a series of publications and consensus efforts that build on each other. These have included publications
such as:
- Guidelines for Good Pharmacoepidemiology Practices2
- Quality of Reporting of Observational Longitudinal Research3
- Government guidance documents such as Guidance for Industry: Good Pharmacovigilance Practices and Pharmacoepidemiologic Assessment,4 Guidance for Industry: Establishing Pregnancy Exposure Registries5, and consensus statements from various working groups.
Canada, EU Exchange Safety Info
|
In May 2007, the U.S. Deptartment of Health and Human Services' Agency for Healthcare Research and Quality released a handbook
entitled Registries for Evaluating Patient Outcomes: A User's Guide,6 to help guide registry design and evaluation.
With these new principles of good practice, registries in both pre- and postmarketing areas will have more potential to provide
data to support decision making. Paralleling these efforts, so-called evidence hierarchies are also being reconsidered, and
the role of rigorous, high-quality observational data for clinical and policy decision making is clearly changing. This article
focuses on the impact of this process on unleashing the power of one type of observational program: the patient registry.
Multiple purposes
A patient registry is "an organized system that uses observational study methods to collect uniform data (clinical and other)
to evaluate specified outcomes for a population defined by a particular disease, condition or exposure, and that serves a
predetermined scientific, clinical or policy purpose(s). The registry database is the file (or files) derived from the registry."6
As this definition notes, registries are observational, meaning they collect data as patients present for care, without prescribing
treatments or tests. This is the primary difference between a registry and a clinical trial. In a registry, treatment decisions,
visit schedules, and any tests/measurements are generally left to the discretion of the provider; the registry simply observes
as patients present for care. Registries also use broad inclusion and exclusion criteria to produce data that is generalizable
to a wider population than trial data. In contrast, clinical trials are controlled experiments designed to test a focused
clinical hypothesis. They have strict inclusion and exclusion criteria that create homogenous patient groups, and treatment
decisions are generally dictated by the protocol and randomization rather than by the practitioner.
The design of clinical trials gives them strong internal validity, meaning that the data collected accurately reflects the
truth about the population under study. In contrast, registry data have strong external validity, or generalizability. Practitioners
and payers, and even some regulatory agencies, are beginning to view real-world data from registries favorably because the
limited inclusion/exclusion criteria mean that the patients are more representative of usual practice and because estimates
of the impact of treatment are more realistic.
