Co-chairs: Robert Heinssen, PhD; Joanne Severe, MS

Traditional randomized clinical trials (RCT) are considered the “gold standard” for inferring causality.  Three elements that make RCTs so valuable are:  randomization to treatments for control of systematic bias, standardization of the interventions being tested, and independence of evaluation.  But RCTs are expensive, often take a long time to yield information, and are limited in generalizability due to several factors including patient selection and the rigors of clinical trial methodology.  This session will seek to define the potential role of the practice-based patient setting in CNS intervention development and refinement.  It will explore whether the most highly valued attributes of RCTs can be achieved on practice-based networks, or if not, do the networks provide other ways to address reduction of bias and clarity of the interventions?  Can data already being collected as part of routine clinical care satisfy the standards of research data?  Are there advantages to practice-based networks such as long-term history data and the ability for longer-term follow-up that are especially valuable?  The session will also aim to help identify the circumstances where practice-based networks will not be useful so that researchers may invest in “traditional” more costly RCTs where they are likely to have the greatest return.

Speakers will address the following issues: A National Institute of Mental Health perspective on an agenda for intervention development and testing in CNS mental health from experimental targets through comparative effectiveness research; development and experience with innovative clinical trial methods utilizing practice-based networks in non-CNS research at the National Heart Lung and Blood Institute; experience from the Mental Health Research Network using a data warehouse and EMR data in setting up clinical trials; the potential role of practice-based networks in CNS development from an industry perspective; and a biostatistical and methodological view of the potential pitfalls and biases in interpretation and generalizability of clinical trials launched on such networks.