New and Improved Methodology Standards: Let Us Know What You Think
At PCORI, we know that when it comes to research, methods matter. Patients, clinicians, and other healthcare stakeholders won’t fully trust the results of PCORI-funded research if we can’t vouch for the way the projects were conducted. To help ensure that everyone can have confidence in the results of studies we fund, we have created the PCORI Methodology Standards. Our first standards were released in 2012, and we have now posted additions and revisions to those standards for public comment.
The PCORI Methodology Committee developed the initial set of standards, meeting one of the requirements of our authorizing law. As minimal requirements for scientifically valid patient-centered clinical research, these standards have been used by our awardees and many others in the field. However, knowledge about what makes for sound research is constantly growing, so the Methodology Committee's work wasn't done.
Over the past year, the committee worked on improvements to the original standards. On December 7, 2015, our Board of Governors reviewed a proposed update that included eight new standards, as well as changes to existing standards that reflect advances in methodology, clarify content, or make the standards better align with one another and other PCORI guidance. The Board approved posting of the draft standards for public comment. Comments are welcome through late March.
The original methodology standards fell into 11 categories. Of the eight new standards, three were added to these existing categories. The other five new standards make up a new category. It is called Standards for Designs Using Clusters. In creating these standards, the Methodology Committee worked with experts on this type of design. The addition of this category reflects the increasing number of clinical studies that use such protocols. These studies tend to require large numbers of patients and complex designs and analyses.
Cluster designs are usually based on the model of the randomized controlled trial (RCT). In traditional RCTs, each research participant is randomly assigned, individually, to one of the medical interventions being studied. In a cluster RCT, groups, or clusters, of participants are randomly assigned to the interventions. This design is especially appropriate for complex interventions that involve interactions between patients and clinical staff.
For example, a study design using clusters might randomly assign each participating clinic to provide one of the study’s intervention to its patients, and then the research team would examine the outcomes experienced by the individual patients attending each of those clinics.
Changes to the original standards include revised wording of 25 standards and grouping of some standards. One standard was deleted, and nine remain unchanged.
What Do You Think?
We are eager to obtain reactions to the important Methodology Standards changes from the communities that conduct and use clinical research. We request your feedback on the new and revised standards during the public comment period, which ends April 11. We also look forward to hearing from you about new areas of research methodology for which additional standards would be valuable.
The Methodology Committee will continue to provide critical guidance to us and to the larger research community on advancing high-quality patient-centered outcomes research as it fulfills our mandate to develop and improve the methods used in this research.