Development of a public health information infrastructure for postmarket evidence

Abstract

Postmarket data on prescription medical product performance has historically been incomplete, underutilized, and mismanaged to inform safety and comparative clinical effectiveness. Congress has tasked the Food and Drug Administration to build a public health information infrastructure for drug safety. It also has allotted $1.1 billion dollars in new spending for comparative effectiveness research. A singular, shared, multi-purpose public health information infrastructure can be built to serve both these needs and others. It can be used by multiple public health agencies under a coordinating framework. A new independent public health authority is best positioned to manage that framework and to negotiate the security, legal, proprietary, and privacy barriers that accompany requests to access large amounts of patient data. Such a design protects privacy, avoids duplication, leverages investment, and promotes sustainability in what is truly a "greenfield" opportunity in the United States. Consequently, the policy window to influence the system design is now. Personal health data is the scarce resource needed to constitute this infrastructure. Citizens have a right and responsibility to re-examine how postmarket data is used to measure safety and comparative clinical effectiveness. A public process to establish new classification schemes that set benefit-risk targets for classes of prescription medical products is needed. Such schemes would differentiate products according to therapeutic need, expected length of treatment, expected patient population, novelty of treatment, and availability of substitutes.

(cont.) These classes would prompt different postmarket requirements according the needs and values of the affected patient population. Data collection, data analysis, risk management strategies, and reimbursement strategies would logically follow from this classification. In this paper, inadequate historical postmarket data generation mechanisms and risk management plans are reviewed. Specific attention is given to the failed use of "carrots" and "sticks" to elicit desired behavior. Next, an analysis of stakeholder interests and desired public health outcomes is performed. Policy goals for a public health information infrastructure are outlined along with strategies to achieve those goals.