When clinical trials arent enough


Published on

Published in: Health & Medicine, Business
  • Be the first to comment

  • Be the first to like this

No Downloads
Total Views
On Slideshare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

When clinical trials arent enough

  1. 1. Postmarketing Studies PHOTOGRAPHY: JIM SHIVERichard Gliklich ILLUSTRATION: PAUL A. BELCIWhen Clinical TrialsAren’t EnoughEffectiveness and real-world dataunderlie the global need for morepatient registries and standards.P atient registries have been gaining attention in Outcomes: A User’s Guide,6 to help guide registr y design recent years as health care providers, payers, and and evaluation. regulators have realized they can produce With these new principles of good practice, registries in real-world data that more accurately depict the both pre- and postmarketing areas will have more potential to safety, quality, performance, and effectiveness of provide data to support decision making. Paralleling thesedrugs and devices. There has been a growing movement to efforts, so-called evidence hierarchies are also being recon-raise the quality of observational research, including patient sidered, and the role of rigorous, high-quality observationalregistries, so that it can be used for more purposes—from data for clinical and policy decision making is clearly chang-labeling and advertising to regulator y submissions and ing. This article focuses on the impact of this process onreimbursement decisions. unleashing the power of one type of observational program: Recent efforts by several groups have brought forth important the patient registry.guidelines and principles to develop or evaluate observationalstudy data. While in clinical trials, the CONSORT statement1 Multiple purposescriteria have set the bar and standardized practice. However, A patient registry is “an organized system that uses observa-for obser vational research there is not a single definitive tional study methods to collect uniform data (clinical anddocument, but rather a series of publications and consensus other) to evaluate specified outcomes for a population definedefforts that build on each other. These have included by a particular disease, condition or exposure, and that servespublications such as: a predetermined scientific, clinical or policy purpose(s). The• Guidelines for Good Pharmacoepidemiology Practices2 registry database is the file (or files) derived from the registry.”6• Quality of Reporting of Observational Longitudinal Research3 As this definition notes, registries are observational, meaning• Government guidance documents such as Guidance for they collect data as patients present for care, without prescribing Industry: Good Pharmacovigilance Practices and Pharma- treatments or tests. This is the primary difference between a coepidemiologic Assessment,4 Guidance for Industr y: registry and a clinical trial. In a registry, treatment decisions, Establishing Pregnancy Exposure Registries5, and consensus visit schedules, and any tests/measurements are generally statements from various working groups. left to the discretion of the provider; the registr y simplyIn May 2007, the U.S. Deptartment of Health and Human obser ves as patients present for care. Registries also useSer vices’ Agency for Healthcare Research and Quality broad inclusion and exclusion criteria to produce data that isreleased a handbook entitled Registries for Evaluating Patient generalizable to a wider population than trial data. In contrast,
  2. 2. Postmarketing Studiesclinical trials are controlled experiments designed to test a (and patients often have more comorbidities and concomitantfocused clinical hypothesis. They have strict inclusion and medicines than those included in the controlled setting of aexclusion criteria that create homogenous patient groups, and clinical trial). While clinical trials are the best measure of efficacy,treatment decisions are generally dictated by the protocol and registries provide strong evidence for effectiveness by observ-randomization rather than by the practitioner. ing typical practice and including a broad patient population. The design of clinical trials gives them strong internal While registries can ser ve many purposes and providevalidity, meaning that the data collected accurately reflects strong effectiveness data, decision-makers need to be able tothe truth about the population under study. In contrast, evaluate a registry in terms of the quality of the research andregistry data have strong external validity, or generalizability. evidence. The recent efforts described earlier have culmi-Practitioners and payers, and even some regulatory agencies, nated in a far better understanding of what constitutes goodare beginning to view real-world data from registries favorably registry practices, such as confidence that the design, conduct,because the limited inclusion/exclusion criteria mean that the and analysis of the registry protect against systematic errorspatients are more representative of usual practice and and errors in inference. Promoting consistent standards forbecause estimates of the impact of treatment are more realistic. registr y design and evaluation will encourage the use of Since registries are observing patients and not dictating patient registries as a way to provide evidence to support deci-treatment, they can fulfill a wide variety of purposes. They can sion making. Widespread use of patient registries as a comple-be used to conduct a prospective study of people who share a ment to clinical trials will provide benefits to patients, health care providers, companies, payers, and even regulators. Practitioners and payers, and even Patient and physician benefits some regulatory agencies, are Registries can provide data on real world treatment choices and outcomes for all types of patients. Clinical trial data may beginning to view real-world data only provide limited evidence for treating patients with multiple from registries favorably. or complex conditions or patients who are part of a sensitive subgroup, such as pregnant women, children or the elderly. Acertain characteristic, such as product use or a procedure. registry can gather data on a broad patient population andExamples include registries of patients who receive intravenous provide evidence to support treatment decisions for manythrombolytics for heart attack or particular antiplatelet types of patients. Registries often include long-term follow-uptherapies after similar events. Registries can also focus on data, which can be particularly useful for patients with chronicdiseases. They can be particularly useful for tracking long-term conditions and for insight into long-term medication use. Inprogression of chronic diseases, such as heart failure and cancer, some disease areas, particularly rare diseases, patient advo-and rare diseases like Pompe disease and cystic fibrosis. cacy groups also promote participation in a registr y as a Other registries provide a comparison of cases and controls means of increasing knowledge about the disease and promot-to understand what differentiates them. These studies can be ing greater communication between patients and providers.used to evaluate safety, characterize adherence, and describe Health care providers similarly benefit from additionaloff-label use. Finally, registries provide a means of gathering evidence on effectiveness and safety that is broadly applicabledata in areas where experimental research is impractical or to their patient populations. Registry data can provide practicalnot feasible, such as in studies of compliance and adherence; information on how to treat a wide variety of patients, includinglong-term follow-up studies; studies in special patient the sensitive subgroups rarely studied in clinical trials. Reg-subgroups (e.g., pregnant women, the elderly, and children); istries also generate data to help characterize rare diseasesand when randomization to certain treatments is not ethical. and can promote information sharing among providers treat- ing patients with such diseases. Additionally, in cases where aEfficacy and effectiveness provider has only one or two patients with a particular condi-The flexibility of registr y design and the wide number of tion, a registry can provide a way to learn more about the dis-purposes that registries can serve has led to their growing ease and connect with other providers with similar patients.popularity. Interest in registries has also increased as the focus By collecting data on treatment patterns and long-termin evidence-based medicine has shifted from efficacy to effec- outcomes, registries can also provide evidence to support thetiveness. Efficacy is the extent to which medical interventions development of treatment guidelines.achieve health improvements under ideal circumstances—essentially, does the intervention work? Effectiveness, on the A company perspectiveother hand, is the extent to which medical inter ventions Registries offer companies a way to achieve clinical andachieve health improvements in real practice settings. marketing goals, meet regulatory needs, and support other data Effectiveness adds an additional hurdle for interventions: needs. Clinical goals of company-sponsored registries includeThey must work in the real world, where providers must studying the effects of products designated for long-term use;prescribe them correctly and patients must comply with therapy examining unanticipated beneficial effects that may lead to new
  3. 3. Postmarketing Studies approved under an accelerated review process and fulfill post- The ability to provide long-term marketing commitments. Some registries are required to fur- data to support safety during ther study premarket safety signals or answer additional ques- tions from the regulatory agency. In some cases, companies real-world use is a strong benefit develop rollover safety registries as a way to continue to mon- of patient registries. itor a population from a clinical trial for any long-term safety signals. The ability to provide long-term data to support safetyindications or side effects that lead to labeling changes; and during real- world use is a strong benefit of registries and oneidentifying best practices to achieve optimal outcomes. that is encouraging some companies to implement registries A registry can support marketing goals, such as assisting in for new products as a proactive approach to risk managementmarket penetration; understanding utilization patterns and spe- in the current environment of heightened concerns.cial patient subpopulations; further documenting safety or Payers and insurers are also beginning to appreciate the ben-effectiveness; repositioning the product through different out- efits of registry data. These groups can use the practical com-comes (e.g., quality of life and cost effectiveness) and develop- parative information on clinical effectiveness and safety froming relationships with providers. Data on cost effectiveness and registries to help them make better formulary andcomparative effectiveness can be useful in advocating for reimbursement decisions, including understanding the needsnational or private payer changes in coverage determinations. of sensitive subgroups. In some cases, payers have even requ-Registries can also offer data for understanding a particular ested registry data to better evaluate the use of a new productmarket by answering questions on what kind of patients use a or procedure in populations similar to their covered population.product, what other treatments they use, how often and why The recent implementation of the coverage under evidencethey switch therapies, and what other co-morbidities they have. development7 program by the Centers for Medicare and Med- Companies also turn to registries to meet regulator y icaid Services (CMS) is an example of how registries may beneeds. Registries can collect additional data for products used to determine reimbursement policies. In the case of implantable cardioverter defibrillators (ICDs), when CMS was asked to reconsider its prior coverage decision in 2004, Canada, EU Exchange Safety Info the agency determined that the new clinical trial evidence was A new agreement between Canada and the Euro- not useful for determining the appropriateness of the proce- pean Union will allow their respective regulatory dure for an elderly population. The median age in the trial pop- experts to share confidential information, such as ulation was 60, but the median age of Medicare patients is 70 safety and authorization issues regarding products to 75. CMS also noted concerns about the expanding group of on the market or in development. The European physician specialists inserting ICDs. Commission (EC) and the European Medicines To gather data on these questions, CMS made an ICD Agency (EMEA) struck this deal with the Health registry a condition of coverage for Medicare beneficiaries Products and Food Branch of Health Canada at their and the results have been dramatic, with nearly 1500 hospitals bilateral meeting on December 7 in Brussels, Bel- participating. Full participation in a registry as a requirement gium. The confidentiality arrangements will enable for coverage for pharmaceutical and biologic products seems earlier availability of information, helping authori- unlikely in the foreseeable future. However, the interest and ties to better protect public health by improving value placed on such observational data by decision makers is the safety and performance of medicines and allow- not, and more and more data of this type will be increasingly ing speedier patient access to new treatments. requested in reimbursement determinations. Since Canada and the EU have a history of More recently, regulatory groups have begun to consider working together in the regulatory environment, registry data in safety evaluations. Registries can meet regula- this new collaboration will further previous tory requirements for risk management plans and postmar- efforts by encouraging the exchange of informa- keting commitments. Some regulatory agencies have even tion, which includes position papers and regula- tory guidance documents, scientific advice on begun including observational data in their assessments of product development, assessments of marketing requests for supplementary indication approval. The use of authorization applications, and safety informa- registry data is likely to increase as more registries can pro- tion on marketed medicines. vide data that meet the requirements for decision making. To view the letters from Health Canada and the In addition, the current debate over products with safety EC and EMEA concerning the confidentiality issues that were not detected until the product was being arrangements, visit www.emea.europa.eu/pdfs/ widely used could result in an increase in the use of patient general/direct/pr/confidentiality_canada.pdf and registries to study real-world product use. In the United www.emea.europa.eu/pdfs/general/direct/pr/ States, a September 2006 report by the Institute of Medicine confidentiality_EU_EMEA.pdf. — Samantha Etkin called for widespread changes in the way medications are monitored and marketed after approval. Recommendations
  4. 4. Postmarketing Studiesincluded changing labeling and mar- The resulting increase in registryketing rules for new products and A 2006 report by the Institute data will help patients and health carereevaluating safety and effectiveness of Medicine calls for providers make more informed treat-data of new products within five yearsof approval.8 changes in the way drugs are ment decisions, enable payers to bet- ter determine coverage policies, and Japan is already studying real monitored after approval. help companies better understandworld drug use through obser va- and market their products.tional studies called drug use investi-gations. These studies monitor real-world product use for sev- Referenceseral years after the product is approved and can be used to 1. http://www.consort-statement.org/ (accessed August 10, 2007).estimate the incidence of an adverse event or identify new 2. E.B. Andrews, F.M. Arellano, J. Avorn et al., “Guidelines for Goodsafety signals. Final drug approval is contingent upon submis- Pharmacoepidemiology Practices (GPP),” Pharmacoepidemiologysion of drug use investigation plans. As more regulatory agen- and Drug Safety, 14, 589–595 (2005).cies develop postmarketing data requirements, companies 3. L. Tooth, R. Ware, C. Bain, D.M. Purdie, A. Dobson, “Quality ofcan implement patient registries to gather long-term data Reporting of Observational Longitudinal Research,” Americanfrom a heterogeneous population to further document the Journal of Epidemiology, 161, 280–288 (2005).safety of their products. 4. U.S. Department of Health and Human Services, Guidance for Industry: Good Pharmacovigilance Practices and Pharmacoepi-Growing interest demiologic Assessment,The need for patient registries is clearly global. In many http://www.fda.gov/cder/guidance/6359OCC. htm (2005).European countries, long-term registries are being used to 5. U.S. Department of Health and Human Services, Guidance formeet both pharmacovigilance and reimbursement needs. Industry: Establishing Pregnancy Exposure Registries,Registries are being recommended or used increasingly for http://www.fda.gov/cber/gdlns/pregexp.htm (2005).risk minimization plans.9 Growth in patient registries is also 6. R.E. Gliklich and N.A. Dreyer, “Registries for Evaluating Patientoccurring from a public health perspective. In the United Registries: A User’s Guide,” Agency for Healthcare Research andKingdom, for example, in late 2006, the Cooksey Report Quality publication No. 07-EHC001 (AHRQ, Rockville, MD, 2007).specifically highlighted the need for the development of a 7. Centers for Medicare & Medicaid Services, National Coverageseries of national disease registries. The availability of inter- Determinations with Data Collection as a Condition of Coverage:national good practice principles is one enabler, as are new Coverage with Evidence Development, https://www.cms.hhs.gov/technologies that provide infrastructure for deploying multi- mcd/ncpc_view_document.asp?id=8 (accessed August 10, 2007).national programs—even with regional differences—in a 8. The Institute of Medicine, The Future of Drug Safety: Promotingcost-effective manner.10 and Protecting the Health of the Public (IOM, Washington, DC, Although clinical trials have been the gold standard for 2006).evidence for some time now, obser vational data from a 9. http://www.emea.europa.eu/pdfs/human/euleg/9626805en.pdf.well-designed study can provide strong, trustworthy evi- (accessed August 10, 2007).dence in a variety of situations.11 Observational data from 10. R.E. Gliklich, “Global Data with a Regional Flavour,” Good Clinicalregistries is gaining favor for its ability to provide generaliz- Practice Journal, 12 (1) (January 2005).able data on effectiveness, safety, and real-world treatment 11. J. Concato, N. Shah, R.I. Horwitz, “Randomized, Controlled Trials,patterns. The growing movement for good practice principles Observational Studies, and the Hierarchy of Research Designs,”for registr y data is an indication of this interest. As New England Journal of Medicine, 342, 1887–1892 (2000)researchers turn to registries to fulfill more data needs, theywill likely incorporate the handbook principles as a way to Richard Gliklich, MD, is president and chief executive officer ofensure that the evidence that they produce is strong enough Outcome Sciences, 201 Broadway, Cambridge, MA 02139,to support decision making. email: richg@outcome.com. © Reprinted from APPLIED CLINICAL TRIALS, March 2008 Printed in U.S.A. TM (617) 621-1600 • info@outcome.com • www.outcome.com