In the 1990s my grandmother was trying to make a number of decisions including where to live and what trips to take following her diagnosis with a chronic disease. At one point she was frustrated that her doctors could not tell her what chance she had of being able to attend my cousin’s wedding. While doctors spoke about her chances of dying, she was focused on how to live. She wanted to know about her physical, mental, and social health and what to expect so she could plan to attend the wedding. She wanted me to find research and use the findings to help her make decisions. My experience with Grandma was a powerful personal lesson in the importance of attention in research to what the patient cares about.
For decades there have been critical disadvantages to the instruments used to assess symptoms and patient-reported health in clinical research. This limitation in measurement approaches impacts every goal and strategic objective for NCCAM. It is hard to study symptoms like pain and depressed mood without using a patient-reported outcome (PRO) measure. While the need for good PROs is not unique to NCCAM, it is extra important to us and repeatedly impacts the design of studies we support.
For the last 9 years, the National Institutes of Health (NIH) Common Fund has invested in self-reported outcomes measurement development and supported the Patient-Reported Outcomes Measurement Information System (PROMIS) program. One of the main objectives of PROMIS is to compile a core set of questions or measurements to assess the most common dimensions of patient-relevant outcomes for the widest possible range of chronic disorders and diseases. PROMIS aims to provide clinicians and researchers access to efficient, precise, valid, and responsive adult- and child-reported measures of health and well-being. These are not simple checklists developed around a conference table. They are measurement instruments developed using rigorous qualitative and quantitative methods. People across many different disease groups and those who are well (after all, many of us are or want to be well!) are involved in instrument development and testing. Modern methods are used, including item response theory, to develop the scoring algorithms. The resulting item banks are used to make PROMIS instruments available, including short forms (a fixed small number of questions or items) and computer adaptive tests (CATs). Instruments are available at the Assessment Center and in part through REDCap’s Library. This is an exciting program for patients, providers, researchers, policy makers, and even math geeks like me who want to focus on developing solutions for real world problems.
While PROMIS originally focused on the development of adult and pediatric PROs for chronic disease research, we have frequent requests for use of the measures in patient care and population health settings. Patients and providers are asking that PROs be used to build and interpret the evidence base and to assist in clinical decision making and management of chronic disease. Critical to our meeting this request is the development and validation of PRO measures that are comparable across many different diseases, groups, and settings. The measures also need to be reliable, valid, flexible, and inclusive of all people.
NCCAM has played an increasing role in PROMIS over the past 10 years. Currently, NCCAM Director Josephine Briggs co-chairs the working group with NIAMS Director Stephen Katz. My colleague Barbara Stussman and I have been involved with the investigators and various committees to provide scientific insight, programmatic support, and oversight. This Common Fund program is a true team effort across the many institutes and centers of NIH; it also has involved colleagues at other agencies, including the Food and Drug Administration and the Department of Defense.
In future blog posts, we can discuss other NIH measurement initiatives such as the NIH Toolbox and Neuro-QOL. Let us know of topics in this area you are interested in reading about. If you have used PROMIS or Assessment Center, do you have any feedback for us? What do you consider when choosing to use a PRO or assessment tool in a research study? What areas do you consider prime for development or additional work?