TY - JOUR
T1 - Progress in assessing physical function in arthritis
T2 - PROMIS short forms and computerized adaptive testing
AU - Fries, James F.
AU - Cella, David
AU - Rose, Matthias
AU - Krishnan, Eswar
AU - Bruce, Bonnie
PY - 2009/9
Y1 - 2009/9
N2 - Objective. Assessing self-reported physical function/disability with the Health Assessment Questionnaire Disability Index (HAQ) and other instruments has become central in arthritis research. Item response theory (IRT) and computerized adaptive testing (CAT) techniques can increase reliability and statistical power. IRT-based instruments can improve measurement precision substantially over a wider range of disease severity. These modern methods were applied and the magnitude of improvement was estimated. Methods. A 199-item physical function/disability item bank was developed by distilling 1865 items to 124, including Legacy Health Assessment Questionnaire (HAQ) and Physical Function-10 items, and improving precision through qualitative and quantitative evaluation in over 21,000 subjects, which included about 1500 patients with rheumatoid arthritis and osteoarthritis. Four new instruments, (A) Patient-Reported Outcomes Measurement Information (PROMIS) HAQ, which evolved from the original (Legacy) HAQ; (B) "best" PROMIS 10; (C) 20-item static (short) forms; and (D) simulated PROMIS CAT, which sequentially selected the most informative item, were compared with the HAQ. Results. Online and mailed administration modes yielded similar item and domain scores. The HAQ and PROMIS HAQ 20-item scales yielded greater information content versus other scales in patients with more severe disease. The "best" PROMIS 20-item scale outperformed the other 20-item static forms over a broad range of 4 standard deviations. The 10-item simulated PROMIS CAT outperformed all other forms. Conclusion. Improved items and instruments yielded better information. The PROMIS HAQ is currently available and considered validated. The new PROMIS short forms, after validation, are likely to represent further improvement. CAT-based physical function/disability assessment offers superior performance over static forms of equal length. The Journal of Rheumatology
AB - Objective. Assessing self-reported physical function/disability with the Health Assessment Questionnaire Disability Index (HAQ) and other instruments has become central in arthritis research. Item response theory (IRT) and computerized adaptive testing (CAT) techniques can increase reliability and statistical power. IRT-based instruments can improve measurement precision substantially over a wider range of disease severity. These modern methods were applied and the magnitude of improvement was estimated. Methods. A 199-item physical function/disability item bank was developed by distilling 1865 items to 124, including Legacy Health Assessment Questionnaire (HAQ) and Physical Function-10 items, and improving precision through qualitative and quantitative evaluation in over 21,000 subjects, which included about 1500 patients with rheumatoid arthritis and osteoarthritis. Four new instruments, (A) Patient-Reported Outcomes Measurement Information (PROMIS) HAQ, which evolved from the original (Legacy) HAQ; (B) "best" PROMIS 10; (C) 20-item static (short) forms; and (D) simulated PROMIS CAT, which sequentially selected the most informative item, were compared with the HAQ. Results. Online and mailed administration modes yielded similar item and domain scores. The HAQ and PROMIS HAQ 20-item scales yielded greater information content versus other scales in patients with more severe disease. The "best" PROMIS 20-item scale outperformed the other 20-item static forms over a broad range of 4 standard deviations. The 10-item simulated PROMIS CAT outperformed all other forms. Conclusion. Improved items and instruments yielded better information. The PROMIS HAQ is currently available and considered validated. The new PROMIS short forms, after validation, are likely to represent further improvement. CAT-based physical function/disability assessment offers superior performance over static forms of equal length. The Journal of Rheumatology
KW - Computerized adaptive testing
KW - Health Assessment Questionnaire Disability Index
KW - Item response theory
KW - PROMIS
UR - http://www.scopus.com/inward/record.url?scp=70349254791&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=70349254791&partnerID=8YFLogxK
U2 - 10.3899/jrheum.090358
DO - 10.3899/jrheum.090358
M3 - Article
C2 - 19738214
AN - SCOPUS:70349254791
SN - 0315-162X
VL - 36
SP - 2061
EP - 2066
JO - Journal of Rheumatology
JF - Journal of Rheumatology
IS - 9
ER -