2004S-0170 - Medicare Prescription Drug, Improvement, and Modernization Act of 2003, Section 1013: Suggest Priority Topics for Research
FDA Comment Number : EC28
Submitter : Dr. Robert Califf Date & Time: 05/11/2004 06:05:27
Organization : CERTs Principal Investigators
Category :
Issue Areas/Comments
While there are many important research topics that could be addressed pursuant to Section 1013 of the Medicare Prescription Drug, Improvement, and Modernization Act of 2003, below are three critical topics that should be considered for the initial priority research list. Information gained through each of these three research areas could be used to improve the appropriate use of prescription drugs and thereby the quality of care that patients receive under Medicare, Medicaid and the SCHIP programs.

1. Evaluation is needed regarding the knowledge about the safety and efficacy of 'off-label' use of marketed drugs in the form of evidence reports and systematic overviews where information is available.

In current practice a high percentage prescription drug use occurs in situations that are not addressed in labeling. In addition most prescription medications used in children do not have adequate pediatric labeling. Consequently, the balance of risks and benefits in situations where drugs are used for indications not specifically addressed in labels has not been vetted through the rigorous evaluation process of the FDA. By quantifying the risk and benefit balance, patients, providers and policy-makers should be able to make better informed, evidence-based decisions about off- label use which will result in more appropriate drug and device utilization.

Given the wide-scale off-label use of medical products, aggregated information about the value and direct analysis of gaps in knowledge should improve decision making to maximize the benefit of product use, while minimizing the risk.

2. Evidence report and systematic overviews need to be done to evaluate what is known about comparative effectiveness of drugs and devices.

In most disease states at least one partially effective therapy is available, while in many high priority diseases such as HIV, heart disease and cancer, multiple therapies are in use at this time. Thus, the commonly asked question is not: 'Is this treatment effective?'. Instead the question is: 'How does the balance of risks and benefits of this treatment compare with the balance of risks of benefits of another treatment?'. In many cases, directly comparative clinical trials are not available, but even when they are, independent evaluation of the available data is frequently lacking. When directly comparative trials are not available, it is tempting to do either indirect comparisons from clinical trials or observational treatment comparisons, both of which are clouded by methodological issues. Better information about comparative effectiveness would significantly improve the prevention, treatment and cure of high priority diseases.

In situations where more than one effective drug is available, information about patients who will benefit more from one particular drug will help patients, providers and policy-makers make decisions that will result in better health outcomes for Medicare, Medicaid and SCHIP patients.

3. Methodological studies are needed on the best approaches to computerized provider order entry (CPOE), including: classifying errors and flaws
in CPOE systems, determining the most effective alerts built into CPOE systems to improve the balance of benefit and risk of CPOE, and assessing methods of combining CPOE data with clinical data to assess appropriateness of prescribing.

The Medicare prescription drug legislation mandates specifications for CPOE by 2008 and full implementation within one year after that. Despite the increasing implementation of CPOE systems, little research exists to create standards for this system. Early reports indicate that entirely new types of errors are being created from coding malfunctions and provider entry errors. In order to realize the potential of CPOE, research is needed to define best practices and clinically useful algorithms to improve prescribing.