| |||||||||||||||||||||||||||||||||||||
To understand the opportunity before us, it helps to understand how little we know now. The Food and Drug Administration (FDA) often approves a drug for long-term use on the basis of minimal long-term safety data: as few as 100 people may have been followed for adverse effects in studies lasting at least one year.1 Furthermore, the FDA rarely can compel manufacturers to conduct postapproval studies. The current mainstay for drug safety is the spontaneous reporting of adverse events to the FDA's MedWatch program, which suffers from underreporting, variable data quality, and the lack of a mechanism to assess confounding risk factors.2 Spontaneous reports are a particularly poor instrument for detecting increased risks of common conditions, such as cardiac disease, that greatly affect public health.
The lack of systematic collection and analysis of post-marketing data on the use of drugs and the outcomes of treatment has delayed the discovery of some serious problems until after millions of people have been exposed. For example, cisapride (Propulsid) was withdrawn because it had been prescribed to many people with contraindications, and several coxibs are now known to increase the risk of myocardial infarction.
Two existing programs share features that can be adopted for Medicare data. First, the Vaccine Safety Datalink Project of the Centers for Disease Control and Prevention, which involves the use of linked data from health plans, is an invaluable source of data on vaccine safety.3 Second, the FDA makes some use of linked pharmacy and administrative records from health plans. These databases identify everyone who is eligible to receive a drug or vaccine, the starting and stopping dates of administration, and many outcomes and coexisting medical conditions. In addition to being useful for safety studies, these data sometimes allow assessment of the effectiveness of different therapies under conditions of actual use.
There are important limitations to these data, however. They do not identify some conditions with sufficient specificity, contain no information on actual drug ingestion, lack important information such as smoking status and body-mass index, often lack historical information, and include only drugs covered by the health plan. Therefore, it is important to review patients' full medical records in the small number of cases in which this information is critical.
Some prerequisites already exist for using Medicare data linked with data from health plans to study the use, risks, and benefits of drugs. Most important, the Centers for Medicare and Medicaid Services (CMS) has proposed linking Medicare drug claims with diagnosis and procedure claims.4 In addition, the FDA has experience in working with linked data and in collaborating with expert groups such as the Centers for Education and Research on Therapeutics (funded by the Agency for Healthcare Research and Quality and administered by that agency in collaboration with the FDA).
To take advantage of the opportunity at hand, the CMS must carry out its proposal of creating a fully linked data set and ensure that it remains current. The CMS, the FDA, and their designees will need the ability to review selected outpatient and inpatient records. These agencies will need to clarify their authority to perform these reviews and explain their purpose satisfactorily to clinicians and patients. Finally, the FDA and other agencies will need new sources of funding if they are to use this information meaningfully and to conduct research on new methods for using it optimally. Additional issues, such as a means of sharing these data with researchers and others, must also be addressed.
Medicare data will offer a great opportunity to improve our ability to understand the balance of benefits and risks of drug treatment. If we take advantage of this opportunity, we will know much more about whether drugs are used as intended, whether they have their intended effects, and how risky they are.
The views expressed in this article are those of the authors and do not necessarily represent the views of the Institute of Medicine.
Source Information
Dr. Platt is a professor in and chair of the Department of Ambulatory Care and Prevention, Harvard Medical School and Harvard Pilgrim Health Care, Boston, and principal investigator of the HMO Research Network Center for Education and Research on Therapeutics, Boston, and principal investigator of an FDA contract (HHSF 223200510010C) to perform post-marketing surveillance for drug risks and benefits. Dr. Ommaya is a senior staff member at the Institute of Medicine, Washington D.C.
An interview with Dr. Arnold Epstein about the Medicare prescription-drug benefit may be heard at www.nejm.org. Dr. Epstein is a professor of health care policy at the Harvard School of Public Health, a professor of medicine at Harvard Medical School, and an associate editor of the Journal.
References
| |||||||||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
Interview
