Measurement Properties of the NIH-Minimal Dataset Dutch Language Version in Patients With Chronic Low Back Pain
Annemarie Boer1, Alisa L Dutmer2, Henrica R Schiphorst Preuper2
1Center for Human Movement Sciences, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands.
Study Design:
Validation study with cross-sectional and longitudinal measurements.
Objective:
To translate the US National Institutes of Health (NIH)-minimal dataset for clinical research on chronic low back pain into the Dutch language and to test its validity and reliability among people with chronic low back pain.
Summary Of Background Data:
The NIH developed a minimal dataset to encourage more complete and consistent reporting of clinical research and to be able to compare studies across countries in patients with low back pain. In the Netherlands, the NIH-minimal dataset has not been translated before and measurement properties are unknown.
Methods:
Cross-cultural validity was tested by a formal forward-backward translation. Structural validity was tested with exploratory factor analyses (comparative fit index, Tucker-Lewis index, and root mean square error of approximation). Hypothesis testing was performed to compare subscales of the NIH dataset with the Pain Disability Index and the EurQol-5D (Pearson correlation coefficients). Internal consistency was tested with Cronbach α and test-retest reliability at 2 weeks was calculated in a subsample of patients with Intraclass Correlation Coefficients and weighted Kappa (κω).
Results:
In total, 452 patients were included of which 52 were included for the test-retest study.
Validity:
factor analysis for structural validity pointed into the direction of a seven-factor model (Cronbach α = 0.78). Factors and total score of the NIH-minimal dataset showed fair to good correlations with Pain Disability Index (r = 0.43-0.70) and EuroQol-5D (r = -0.41 to -0.64). Reliability: test-retest reliability per item showed substantial agreement (κω=0.65). Test-retest reliability per factor was moderate to good (Intraclass Correlation Coefficient = 0.71).
Conclusion:
The Dutch language version measurement properties of the NIH-minimal were satisfactory.
Level Of Evidence:
N/A.
