Psychometric evaluation of the PROMIS® physical function short form 12a for use by adults with [ME/CFS], 2025, Yang et al.

[SNT Gatchaman]

Psychometric evaluation of the PROMIS® physical function short form 12a for use by adults with myalgic encephalomyelitis/chronic fatigue syndrome

Spoiler: Authors

Yang, Manshu; Keller, San; Rafiee, Parisa; Lin, Jin-Mann S

BACKGROUND
Myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) is a debilitating, long-term illness that significantly impairs physical functioning. Despite its impact, the use of modern generic instruments to assess physical function in this population remains underexplored. This study aims to assess the psychometric properties of the Patient-Reported Outcome Measurement Information System® (PROMIS) Physical Function Short Form (PF-SF) 12a for use in adults with ME/CFS.

METHODS
This study included 334 participants (173 with ME/CFS and 161 healthy controls) who took part in a Cognitive and Exercise sub-study of the Multi-Site Clinical Assessment of ME/CFS study from six clinics across the US. Data was used to examine the ceiling/floor effects, internal consistency reliability, known-groups validity, and convergent validity of the PROMIS PF-SF.

RESULTS
The mean T-score of the PROMIS PF-SF was 40.5 for participants with ME/CFS, about one standard deviation below the national norm (T-score = 50). The PROMIS PF-SF showed no substantial floor/ceiling effects and high internal consistency (standardized Cronbach’s α = 0.88 and ω = 0.92). In addition, this instrument showed good known-groups validity with medium-to-large effect sizes (η2 = 0.08–0.35). A significant, monotonic increase of the physical function score was found across ME/CFS participant groups with low, medium, and high functional impairment as defined by four different measures. Participants with ME/CFS had significantly worse physical function scores than healthy controls (η2 = 0.70). The PROMIS PF-SF also demonstrated good convergent validity with high correlations (magnitude of r = 0.47–0.55) with other relevant measures.

CONCLUSIONS
The PROMIS PF-SF 12a demonstrated satisfactory reliability and validity for use in ME/CFS research and clinical practice.

Web | PDF | Health and Quality of Life Outcomes | Open Access

 

[Utsikt]

There is a glaringly obvious problem here: they only included the relatively healthy patients.

Images from google indicates that PF-SF asks «are you able to …»

In table 2, we can see that 0 of the participants had the lowest score (presumably that they are unable to) for:

• walk a block on flat ground
• walk up and down two steps
• wash and dry your body
• get in and out of bed
• bend down and pick up clothing
• push open a heavy door
• reach and get down an object from above your head

And for every task except for «run at fast pace for two miles», a large and often close to 100 % majority had the highest score (presumably that they can do it without much difficulty).

Of course there are no floor or ceiling effects if you exclude the ones that would hit them!

The ignorance by the researchers is inexcusable.

 

[bobbler]

I've copied the Acknowledgements section because I think it is relevant to be aware that eg Klimas, Bateman, Unger are in the below

Acknowledgements​

We would like to thank the Multi-Site Clinical Assessment of ME/CFS (MCAM) study group who made the data sets available through Research Collaboration Agreement.

The group included: Division of High-Consequence, Pathogens and Pathology, Centers for Disease Control and Prevention, Atlanta, Georgia: Elizabeth R. Unger, Britany Helton, Yang Chen, and Monica Cornelius; Open Medicine Institute Consortium, Mountain View, CA: Andreas Kogelnik, Catt Phan, Joan Danver, Lucinda Bateman, Jennifer Bland, Charles Lapp, Wendy Springs, Richard Podell, Trisha Fitzpatrick, Daniel Peterson, and Marco Maynard; Institute for Neuro Immune Medicine, Miami, Florida: Nancy Klimas, Elizabeth Balbin, Precious Leaks-Gutierrez, Shuntae Parnell, Constance Sol, and Jeffry Cournoyer; and Mount Sinai Beth Israel, New York, New York: Benjamin Natelson and Diana Vu.

From this it 'seems' like those are in support of this study include many who are more serious about ME/CFS from the US.

I wanted to check for obvious reasons given that historical thing in UK where they wanted to get rid of the SF-36 physical function following it not producing a significant effect in Crawley et al (2013) even though it was done the same time as PACE in UK clinics. And the idea of introducing PROMS instead seemed to follow on from that 'issue' from certain areas.

 

[Utsikt]

I think it’s telling that they have not published any plots of the data points. Those would have shown that most of the participants are clumped towards one end.

If you compare that to the distribution of FUNCAP, it’s really telling.


The score also correlated quite badly with the other survey, only about 0.5.

 

[rvallee]

I don't really know what to think of those "validation" studies but we've seen many obviously useless questionnaires being 'validated' this way over the years, and a process that validates invalid things doesn't strike me as especially valid in itself.

Looks neat with all the buzzwords and comparisons, but if your quality control process approves broken items, the process is broken and should be considered unreliable.