The new National Patient-Centered Clinical Research Network funded by the Patient-Centered Outcomes Research Institute (PCORI) aims to provide patients and clinicians with useful information on treatment options and outcomes. By encouraging more comparative clinical effectiveness research (CER), the program also has potential to support development of new drugs and devices, explained PCORI executive director Joe Selby, in unveiling the program April 23. This prospect has generated enthusiasm for CER, along with concerns about how patient medical data could be misinterpreted and lead to erroneous conclusions.
To launch the initiative, PCORI is providing $68 million to fund eight Clinical Data Research Networks (CDRNs) formed by hospitals and health care systems for 18 months. The networks will access electronic patient health records held by members to support observational and interventional studies on large, defined populations. Additional Patient-Powered Research Networks (PPRMs) established by patient and disease groups will collect and analyze patient data on targeted and rare conditions.
The goal of this national research data infrastructure is to improve patient outcomes by accelerating patient-centered and methodological research. The system will examine ways to govern and use large clinical databases to facilitate rapid, efficient conduct of both randomized trials and observational studies. PCORI will establish a Coordinating Center for the networks, and a Steering Committee will assess policies, best practices and methods needed for an efficient and interoperable research system.
Standards for collecting and analyzing patient health data are particularly important in providing the Food and Drug Administration with “actionable information” that is reliable, pointed out Janet Woodcock, director of the Center for Drug Evaluation and Research. FDA has been developing methodological standards for its Sentinel drug safety data system that obtains drug use information from health system data banks, and the PCORI network would support that program.
John Castellani, president of the Pharmaceutical Research and Manufacturers of America (PhRMA), noted at the PCORI briefing that sponsors are spending billions of dollars on clinical trials, and that “anything done to improve and quality and efficiency of the system will benefit patients.” He expressed caution, though, that “you can’t just take the data and expect it will be useful,” noting that there is a lot of “noise in databases” and that it’s often difficult to identify “true signals” of safety problems. Woodcock explained that linking the use of a specific medicine to any particular health outcome is tricky, and that research on causal inferences requires “a level of rigor.” Studies that can fill gaps in knowledge about interactions of drugs with other drugs and with interventions is particularly important, she noted, as is long-term data on most conditions.
A main benefit of the envisioned research network is to help researchers and clinical trial sponsors identify patients to participate in studies. The data also could help sponsors design clinical trials by providing an understanding of why certain patients don’t respond to treatment and what endpoints may be important for a study, points out Robert Dubois, chief science officer at the National Pharmaceutical Council (NPC), which tracks PCORI and CER activities.
A coordinated, sustained approach to collecting and assessing outcomes data would be highly beneficial, says NPC analyst Jennifer Graff, provided it addresses a number of important issues: who has access to research data, ensuring data privacy, considering patient heterogeneity and developing standards for analyzing data. Once a drug is on the market, observes Dubois, broader patient data can help assess its performance in the real world, provided “ the analysis is right.” Graff notes that 18-months to establish these networks is a “very short time frame for data networks to come together and deal with these issues.”
Woodcock further discussed the need for standards and “appropriate analytical rigor” for outcomes research at this week’s DIA-FDA Statistics Forum in Bethesda, Md. Despite considerable enthusiasm for CER, making causal inferences “is a whole other game,” she cautioned. People need to understand limits on the reliability of health systems information, and that “getting a wrong answer doesn’t help.” She noted that the nation’s enormous investment in electronic health records and CER is generating a massive amount of health data, and urged statisticians and other experts to “help us learn how to use it.”
Many of these challenges will be addressed by PCORI’s newly appointed chief science officer, Bryan Luce, a leading expert in the medical outcomes research world. Luce has been involved in methods and policies related to evidence-based research and will be responsible for leading the development of PCORI’s CER agenda.