Advancing PROMIS’s methodology: results of the Third Patient-Reported Outcomes Measurement Information System (PROMIS^®) Psychometric Summit

Abstract

In 2002, the NIH launched the ‘Roadmap for Medical Research’. The Patient-Reported Outcomes Measurement Information System (PROMIS^®) is one of the Roadmap’s key aspects. To create the next generation of patient-reported outcome measures, PROMIS utilizes item response theory (IRT) and computerized adaptive testing. In 2009, the NIH funded the second wave of PROMIS studies (PROMIS II). PROMIS II studies continue PROMIS’s agenda, but also include new features, including longitudinal analyses and more sociodemographically diverse samples. PROMIS II also includes increased emphasis on pediatric populations and evaluation of PROMIS item banks for clinical research and population science. These aspects bring new psychometric challenges. To address this, investigators associated with PROMIS gathered at the Third Psychometric Summit in September 2010 to identify, describe and discuss pressing psychometric issues and new developments in the field, as well as make analytic recommendations for PROMIS. The summit addressed five general themes: linking, differential item functioning, dimensionality, IRT models for longitudinal applications and new IRT software. In this article, we review the discussions and presentations that occurred at the Third PROMIS Psychometric Summit.

Keywords::

• computerized adaptive testing
• dimensionality
• factor analysis
• item response theory
• patient-reported outcomes
• PROMIS
• psychometrics
• structural equation modeling

Acknowledgements

Adam Carle would like to thank Tara J Carle, Lyla SB Carle and Margaret Carle, whose unending support and thoughtful comments make his work possible. Jeanne Teresi would like to thank Elayne Livote, Joseph P Eimicke, Marjorie Kleinman and Katja Ocepek-Welikson for processing and analyses of the data used in several presentations. Ron D Hays would like to thank Pam Hays for her support and encouragement during the trials and tribulations of PROMIS. He would also like to thank Karen L Spritzer for her computer programming support.

Financial & competing interests disclosure

The following grants supported this work: NINR-R15NR10631 and 3U01AR057940-02S1 (AC Carle); 1U54AR057951 (D Cella); NCI-U01-AR057971, NIMHD, P60, MD00206, NIA, P30 and AG28741 (J Teresi); R305B080016, R305D100039, NIDA-R01DA026943 and NIDA-R01DA030466 (L Cai); NIA-R01 AG 029672 (PK Crane); National Institutes of Health through the NIH Roadmap for Medical Research Grant AR052177 PROMIS Project (R Hays) and 1U01AR052181-01 and 2U01AR052181-06 (D Thissen); and SBIR contract HHSN-2612007-00013C, National Cancer Institute (D Thissen). The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.