File
Haan EJA1,2,3, Terwee CB2, van Wier MF4, van Deurzen DFP4, Willigenburg NW4, Pisters MF1, Kaat AJ5, Roorda LD3
1University Medical Center Utrecht, Physical Therapy Sciences, Program in Clinical Health Sciences, Utrecht, Netherlands, 2Amsterdam UMC, Location VUmc, Department of Epidemiology and Biostatistics, Amsterdam Public Health Research Institute, Amsterdam, Netherlands, 3Reade, Amsterdam Rehabilitation Research Center, Amsterdam, Netherlands, 4OLVG, Department of Orthopedic Surgery, Joint Research, Amsterdam, Netherlands, 5Northwestern University, Feinberg School of Medicine, Chicago, Illinois, United States
Background: The Patient Reported Outcomes Measurement Information System (PROMIS®) has been developed to improve the quality of health outcome measures and to reduce respondent burden. Only recently, the PROMIS®Upper-Extremity (UE) item bank v2.0, consisting of 46 items, has been developed and showed good psychometric properties in individuals with upper-extremity impairments. A Dutch-Flemish translation has not yet been developed and its validity has not been established.
Purpose: The aim of this study was to examine the cross-cultural and construct validity of the Dutch-Flemish PROMIS Upper-extremity (UE) item bank v2.0 in a Dutch population of patients with musculoskeletal upper-extremity disorders.
Methods: Four PROMIS v2.0 physical function items, measuring upper-extremity function, were translated into Dutch-Flemish. These items were combined with 42 already translated items, to develop the 46-item Dutch-Flemish PROMIS UE item bank v2.0. To examine validity, a cross-sectional design was used. Two hundred five patients referred to the Orthopedic Department of a general hospital in The Netherlands, aged 18 years or older, with musculoskeletal disorders of the upper extremity, were included between February and May 2018. Participants filled in a questionnaire with demographic and clinical characteristics, the Dutch-Flemish PROMIS UE item bank v2.0, and four legacy instruments. For cross cultural validity, Differential Item Functioning (DIF) was evaluated for language. For construct validity, a-priory hypotheses were tested for correlations with the legacy instruments. DIF was evaluated by ordinal logistic regression models. When items were flagged as potential DIF for language items, the impact of DIF was examined by plotting item characteristic curves and test characteristic curves. Correlations were quantified by Pearson's or Spearman's correlation coefficient.
Results: Eight items showed minimal DIF for language which resulted in sufficient cross-cultural validity. The Dutch-Flemish PROMIS UE item bank had a moderate correlation with the Dutch-Flemish PROMIS Pain Intensity item (r = -0.43) and strong correlations with the Disabilities of Arm, Shoulder and Hand Questionnaire(r = -0.87), the Functional Index of Hand OsteoArthritis (r = -0.86) and the Michigan Hand Outcomes Questionnaire (r = 0.81), all correlations were as hypothesized.
Conclusion(s): The Dutch-Flemish PROMIS UE item bank v2.0 has sufficient cross-cultural validity and construct validity.
Implications: After full item bank calibration is established, the Dutch-Flemish PROMIS UE item bank v2.0 can serve as a basis for short forms and Computerized Adaptive Testing, resulting in high measurement precision and low patient burden in Dutch and Flemish patients with upper-extremity conditions.
Keywords: Patient-Reported Outcomes Measurement Information, Upper-extremity, Validation
Funding acknowledgements: There was no funding for this study.
Purpose: The aim of this study was to examine the cross-cultural and construct validity of the Dutch-Flemish PROMIS Upper-extremity (UE) item bank v2.0 in a Dutch population of patients with musculoskeletal upper-extremity disorders.
Methods: Four PROMIS v2.0 physical function items, measuring upper-extremity function, were translated into Dutch-Flemish. These items were combined with 42 already translated items, to develop the 46-item Dutch-Flemish PROMIS UE item bank v2.0. To examine validity, a cross-sectional design was used. Two hundred five patients referred to the Orthopedic Department of a general hospital in The Netherlands, aged 18 years or older, with musculoskeletal disorders of the upper extremity, were included between February and May 2018. Participants filled in a questionnaire with demographic and clinical characteristics, the Dutch-Flemish PROMIS UE item bank v2.0, and four legacy instruments. For cross cultural validity, Differential Item Functioning (DIF) was evaluated for language. For construct validity, a-priory hypotheses were tested for correlations with the legacy instruments. DIF was evaluated by ordinal logistic regression models. When items were flagged as potential DIF for language items, the impact of DIF was examined by plotting item characteristic curves and test characteristic curves. Correlations were quantified by Pearson's or Spearman's correlation coefficient.
Results: Eight items showed minimal DIF for language which resulted in sufficient cross-cultural validity. The Dutch-Flemish PROMIS UE item bank had a moderate correlation with the Dutch-Flemish PROMIS Pain Intensity item (r = -0.43) and strong correlations with the Disabilities of Arm, Shoulder and Hand Questionnaire(r = -0.87), the Functional Index of Hand OsteoArthritis (r = -0.86) and the Michigan Hand Outcomes Questionnaire (r = 0.81), all correlations were as hypothesized.
Conclusion(s): The Dutch-Flemish PROMIS UE item bank v2.0 has sufficient cross-cultural validity and construct validity.
Implications: After full item bank calibration is established, the Dutch-Flemish PROMIS UE item bank v2.0 can serve as a basis for short forms and Computerized Adaptive Testing, resulting in high measurement precision and low patient burden in Dutch and Flemish patients with upper-extremity conditions.
Keywords: Patient-Reported Outcomes Measurement Information, Upper-extremity, Validation
Funding acknowledgements: There was no funding for this study.
Topic: Outcome measurement; Musculoskeletal: upper limb
Ethics approval required: Yes
Institution: Slotervaart/Reade and OLVG
Ethics committee: local institutional review boards
Ethics number: P1749
All authors, affiliations and abstracts have been published as submitted.
