The future of measuring patient-reported outcomes in rheumatology: Patient-Reported Outcomes Measurement Information System (PROMIS).

摘要: The National Institutes of Health (NIH) Patient-Reported Outcomes Measurement Information System (PROMIS®) Roadmap initiative (www.nihpromis.org) is a cooperative research program designed to develop, evaluate, and standardize item banks measure patient-reported outcomes (PROs) across different medical conditions as well the US population (1). goal PROMIS develop reliable valid using response theory (IRT) that can be administered in variety formats including short forms computerized adaptive tests (CAT)(1-3). IRT often referred “modern psychometric theory,” contrast “classic test or CTT. basic idea behind both CTT there some latent construct, “trait,” underlying an illness experience. This construct cannot directly measured, but indirectly measured by creating items are scaled scored. For example, “fatigue,” “pain,” “disability,” even “happiness” constructs, i.e. subjective feelings – we take picture, snap X-Ray view them, run blood check for them. However, know they exist. People experience more less these thus it helpful try translate into several levels represented scores. models associations between construct. Specifically, describe relationships respondent's level on probability particular responses. Tests developed with (such Assessment Questionnaire-Disability Index(4), Scleroderma Gastrointestinal Tract instrument(5)) require administering all items, though only appropriate persons' trait level. Some too high those low (e.g., “can you walk 100 yards” patient wheelchair) get up from chair?” runner). In contrast, methods make possible estimate person any subset pool. As such, set pool could fixed form or, greatest efficiency, CAT. CAT approach bank most informative measuring health order achieve target standard error measurement. A good will have represent range content difficulty, provide information, perform equivalently subgroups population. How does work? Without prior first typically one medium “In past 7 days I was grouchy” multi-level “never” “always.” After each response, person's associated estimated. next someone not endorsing item, “easier” item. If endorses “harder” terminated when falls below acceptable value. provides one's score minimal number questions no loss measurement precision. addition, scores studies compared common scale. scale (theta) items. All calibrated same metric independently collectively theta. Hence, estimated score. allows assessment patients differing compare arthritis heart disease) at various degrees physical other impairments, lowest highest ends levels.