Premium
Physical function metric over measure: An illustration with the Patient‐Reported Outcomes Measurement Information System (PROMIS) and the Functional Assessment of Cancer Therapy (FACT)
Author(s) -
Kaat Aaron J.,
Schalet Benjamin D.,
Rutsohn Joshua,
Jensen Roxanne E.,
Cella David
Publication year - 2018
Publication title -
cancer
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 3.052
H-Index - 304
eISSN - 1097-0142
pISSN - 0008-543X
DOI - 10.1002/cncr.30981
Subject(s) - patient reported outcomes measurement information system , metric (unit) , comparability , item bank , medicine , function (biology) , measure (data warehouse) , scale (ratio) , quality (philosophy) , item response theory , computerized adaptive testing , computer science , medical physics , psychometrics , data mining , mathematics , clinical psychology , operations management , philosophy , physics , epistemology , combinatorics , quantum mechanics , evolutionary biology , economics , biology
BACKGROUND Measuring patient‐reported outcomes (PROs) is becoming an integral component of quality improvement initiatives, clinical care, and research studies in cancer, including comparative effectiveness research. However, the number of PROs limits comparability across studies. Herein, the authors attempted to link the Functional Assessment of Cancer Therapy‐General Physical Well‐Being (FACT‐G PWB) subscale with the Patient‐Reported Outcomes Measurement Information System (PROMIS) Physical Function (PF) calibrated item bank. The also sought to augment a subset of the conceptually most similar FACT‐G PWB items with PROMIS PF items to improve the linking. METHODS Baseline data from 5506 participants in the Measuring Your Health (MY‐Health) study were used to identify the optimal items for linking FACT‐G PWB with PROMIS PF. A mixed methods approach identified the optimal items for creating the 5‐item FACT/PROMIS‐PF5 scale. Both the linked and augmented relationships were cross‐validated using the follow‐up MY‐Health data. RESULTS A 5‐item FACT‐G PWB item subset was found to be optimal for linking with PROMIS PF. In addition, a 2‐item subset, including only items that were conceptually very similar to the PROMIS item bank content, were augmented with 3 PROMIS PF items. This new FACT/PROMIS‐PF5 provided superior score recovery. CONCLUSIONS The PROMIS PF metric allows for the evaluation of the extent to which similar questionnaires can be linked and therefore expressed on the same metric. These results allow for the aggregation of existing data and provide an optimal measure for future studies wishing to use the FACT yet also report on the PROMIS PF metric. Cancer 2018;124:153‐60 . © 2017 American Cancer Society .