Measurement properties of the Patient Health Questionnaire 15 (PHQ-15) and Somatic Symptom Disorder B-criteria scale (SSD-12)... 2026 Hybelius et al

[Andy]

Full title: Measurement properties of the Patient Health Questionnaire 15 (PHQ-15) and Somatic Symptom Disorder B-criteria scale (SSD-12), including revised 1-week versions

Abstract​

The Patient Health Questionnaire 15 (PHQ-15), which measures somatic symptom burden, and the Somatic Symptom Disorder B-criteria scale 12 (SSD-12), which measures symptom preoccupation, are widely used questionnaires in behavioral medicine. This study built on the existing literature by evaluating the PHQ-15 and SSD-12 in Swedish, and in revised 1-week versions that could facilitate repeated measurements. The questionnaires were completed online by clinical trial participants with persistent physical symptoms (n = 194), and healthy volunteers (n = 160). For both conventional and revised 1-week versions, we evaluated item distributions, factor structure, internal consistency, construct validity based on correlations with other constructs, and test-retest reliability. Regardless of timeframe, the PHQ-15 factor structure combined general and domain-specific factors. The SSD-12 showed a three-factor structure reflecting Expectation of a chronic course, Health anxiety, and Symptom focus and impairment. For both questionnaires, internal consistency and construct validity were mostly supported, but adequate test-retest reliability was only observed in the clinical data and for averages of timepoints. Test-retest reliability in the healthy volunteer data was poor. This study is cautiously supportive of wider use of revised PHQ-15 and SSS-12 versions with a 1-week focus to facilitate repeated measurements of somatic symptom burden and symptom preoccupation in clinical populations.

Open access

 

[Utsikt]

For both conventional and revised 1-week versions, we evaluated item distributions, factor structure, internal consistency, construct validity based on correlations with other constructs,

It must be nice to be able to redefine words to mean whatever you want them to mean.

 

[rvallee]

adequate test-retest reliability was only observed in the clinical data and for averages of timepoints. Test-retest reliability in the healthy volunteer data was poor

Yet again the absurd error of pretending that you can do test-retest on subjective ratings that have no firm, correct answers and are known to change over time. Might as well be doing this kind of testing for a weather model. It would occasionally find that, just as last week at the same time of day, it's sunny outside and around 20C, and other times it would, well, not. Because what is being tested changes over time, the idea of checking if it holds up another time is just ridiculous.

The Patient Health Questionnaire 15 (PHQ-15), which measures somatic symptom burden, and the Somatic Symptom Disorder B-criteria scale 12 (SSD-12), which measures symptom preoccupation, are widely used questionnaires in behavioral medicine

I have no idea what difference anyone can think there can be between symptom burden and symptom preoccupation. Certainly not in a species where one of the most common social interactions involves asking "how are you doing?", a question to which almost no one actually expects a truthful answer either way. Including in health care consultations. We are not a species that handles these things well even in the most favorable circumstances, and the circumstances here are not even close to that.

It's becoming a theme here, but, really, medicine should drop the use of such questionnaires in almost all cases, stick only to objective questions. They are making a complete mess of things while being completely incapable of seeing their own role in that failure. They always try to confirm reality with their internal models, rather than the opposite, correct way.

At least if they went with my idea of taking out calipers to measure the skull shapes of mannequins they fabricated themselves, their internal consistency and test-retest would hold up perfectly. If that's what they're after, why even bother with anything? This is basically like trying to make use of "a broken clock is right twice per day", which statistically is entirely possible as long as the experiments are rigged and the interpretation biased enough. And if anyone can do that, it's health care professionals armed with bad questionnaires.