- Posts: 60
In 2019, Huang and colleagues started with the Hays et al summary scores (Qual Life Res 27:1885-1891, 2018) and proposed an adaptation for older adults with multiple chronic conditions. See Huang, W., Rose, A.J., Bayliss, E. et al. Adapting summary scores for the PROMIS-29 v2.0 for use among older adults with multiple chronic conditions. Qual Life Res (2019) 28: 199. doi.org/10.1007/s11136-018-1988-z
Karen L. Spritzer and Ron D. Hays have produced a memo that includes SAS code for generating a PROMIS Physical Health and PROMIS Mental Health score from the PROMIS-29 v2.0 and v2.1 profile measures.
Slides from the User Conference are at:
A presentation on PROMIS-29 summary scores is available at: www.healthmeasures.net/images/LearnMore/...mmaryScores_Hays.pdf
The third slide mentions existing work to derive a single preference-based score. The work by Janel Hanmer is producing the
most robust and defensible preference-based score.
The short answer is "no".
However, there is a clear interest by many researchers to have one score for multiple domains - for example, pain interference and physical functioning following orthopaedic trauma; or one score for the Profiles in chronic conditions. These are active areas of research, especially regarding the Profiles.
The long answer to this question is that it may be possible to combine limited items across select domains in specific conditions, but not without really good psychometric input and expertise. These types of projects would build on the strengths of PROMIS, but they would not be PROMIS measures. The scores would not be on a PROMIS metric and the measure wouldn't see widespread applicability.
The psychometrics team is willing to provide consultation and analytic support for researchers with the funds and the need for these types of scores. But generally, we recommend against combining items or scores across multiple domains.
Not only is this statistically justifiable, but it is clinically reasonable. Going back to the earlier example of pain and physical functioning after orthopaedic trauma - if a single summary score was given, and the person was impaired, the clinician wouldn't know what caused the impairment. Should she send the patient to a physical therapist? Or to a pain clinic? Or both? Administering both domains and interpreting both scores provides more information regarding recommended treatments and follow-up that would not be available in a single score.
I would like to create one summary score from multiple PROMIS domains (e.g., a single score for the PROMIS-29). Can I do this?