- Posts: 23
It is important to have evidence for the reliability and validity of any measure used clinically or in research. So if the Improved HAQ is going to be used, then gathering evidence on its psychometric properties in that population would be useful. Consider using something like the HealthMeasures Scoring Service ( www.assessmentcenter.net/ac_scoringservice ) to get pattern-based IRT scores on the Improved-HAQ, as was done in the previous publications.
Hello I was the individual who had originally asked this question during a PHO webinar, and was surprised to find it listed here. Thanks for the replies.
My follow-up question is this: Is it reasonable to state that the I-HAQ/PROMIS HAQ would benefit from published studies validating some of its psychometric properties in specific patient populations?
For a scoring algorithm see
The 20 items from the original HAQ-DI were converted, with permission, into PROMIS items. This changed the number of response options and scoring rules. These converted 20 items were originally referred to as the "PROMIS HAQ" but the name was changed to the "Improved HAQ" or "Item-Improved HAQ" because it is not an officially-distributed PROMIS short form. This Improved-HAQ has been included in several studies (e.g. Fries et al. 2011; Improved responsiveness and reduced sample size requirements of PROMIS physical function scales with item response theory).
Although it is not officially distributed, interested users can create custom short forms from any of the PROMIS banks. To access the Improved HAQ, just create a custom short form from the PROMIS PF item bank with those improved items.
What relationship, if any, exists between PROMIS and the PROMIS Health Assessment Questionnaire (HAQ) which was developed by James Fries? Previous publications refer to the PROMIS HAQ, but there appears to be no mention of that measure on this website. Further confusing the issue, the PROMIS HAQ is also sometimes referred to as the Improved HAQ or the Stanford HAQ. Clarification on the history and/or affiliation would be much appreciated.