Construct validity of instruments measuring impairments in body structures and function in rheumatic disorders: which constructs are selected for validation? A systematic review.

PURPOSE: This paper focuses on the construct validity of instruments measuring impairments in body structures and function in rheumatic disorders. The objective is: 1) to make an inventory of constructs, based on the domains of the International Classification of Functioning, Disabilities and Health problems (ICF), against which instruments measuring impairments in body structures and function were validated; 2) to analyse whether validation against a similar construct resulted in higher correlation coefficients than validation against a dissimilar construct. METHODS: In a systematic review papers were identified in which instruments measuring impairments in body structures and function for patients with rheumatic disorders were validated. The instruments identified were assessed on their methodological properties and the constructs against which they were validated. Subsequently, pooled (interclass) correlations of similar constructs and dissimilar constructs against which was validated were compared. An instrument was decided to have good construct validity, if the correlation coefficient was 0.50 or higher, and the measurement instrument in question is validated against similar constructs. RESULTS: In total 216 papers were identified analysing the validity of 42 different instruments. Only 16% of these instruments were validated against instruments that represent the most similar construct. In general, estimates of construct validity were lower when validated against dissimilar constructs, except for instruments measuring impairments in mental functions. CONCLUSION: There is a trend that validation against a similar construct yields higher correlation coefficients than validation against a dissimilar construct. If an instrument measuring impairments is validated against the most similar construct, and a criterion of r > 0.50 is applied, only 10 out of the 42 identified instruments turned out to be valid.

Reliability, validity and responsiveness of instruments to assess disabilities in personal care in patients with rheumatic disorders. A systematic review.

OBJECTIVES: The first aim was to make an inventory of available instruments and questionnaires for the assessment of disabilities in personal care in patients with rheumatic disorders. The second aim was to investigate which of these instruments have acceptable, methodological quality with regard to reliability, validity and responsiveness. The third aim was to investigate the assumption that convergent validity results in stronger correlations when validated against a more similar construct. METHODS: A computer-aided literature search (1982-2001) in several databases was performed to identify studies focusing on the clinimetric properties of instruments to assess impairments in function in patients with rheumatic disorders. Data were extracted in a standardised way and compared to a priori defined criteria. RESULTS: In total, 19 measurement instruments were included. Five out of these 19 were found to have acceptable reliability, while 12 had acceptable validity. Only three questionnaires met both criteria. Results concerning the responsiveness of these three questionnaires were conflicting. No difference was found in the strength of correlation between validation against the most similar construct versus validation against the least similar construct. CONCLUSION: It is concluded that the Arthritis Impact Measurement Scale (AIMS) is the most suitable instrument for the assessment of disabilities in personal care.

Psychometric properties of the Tampa Scale for kinesiophobia and the fear-avoidance beliefs questionnaire in acute low back pain.

The transition from acute to chronic low back pain (LBP) is influenced by many interacting factors. Pain-related fear, as measured by the Tampa Scale for Kinesiophobia (TSK) and the Fear-Avoidance Beliefs Questionnaire (FABQ), is one of these factors. The objectives of this study were to investigate, in a population with acute LBP, the reliability of TSK and FABQ through evaluation of the internal consistency, the test-retest reliability, and the concurrent validity between TSK and FABQ. One hundred and Seventy-Six patients suffering LBP for no longer than 4 weeks completed a Visual Analogue Scale for pain (VAS), the TSK, the FABQ, and a socio-demographic questionnaire. Each patient completed the VAS, TSK, and FABQ twice within 24 h. Internal consistency of TSK and FABQ scores range from alpha=0.70 to 0.83. Test-retest reliability ranges from r(s)=0.64 to 0.80 (P<0.01). Concurrent validity is moderate, ranging from r(s) =0.33 to 0.59 (P<0.01). It may be concluded that in a population with acute LBP, both the TSK and the FABQ are reliable measures of pain-related fear. In the clinical setting they may provide the practitioner a means of identifying pain-related fear in a patient with acute LBP.

Upper cervical instability: are clinical tests reliable?

SUMMARY. The aim of this preliminary study was to investigate the reliability of a selection of manual tests used for the examination of instability of the upper cervical region. Eleven children with Down's syndrome were examined by four independent examiners with different levels of experience in manual therapy. Three tests as described by Van der El (1992) were used: the lateral displacement test, the Sharp-Purser test, and the upper cervical flexion test. Scores of tests and retests were statistically analysed by calculating the percentage of agreement, and Brennan and Prediger's modified kappa, and with the binomial test. The results showed a significant agreement between test and retest for the upper cervical flexion test in three out of four investigators. Agreement between investigators was significant in four out of six combinations between two investigators, and near to significant for this test (P(bin) =.07) in the two remaining combinations. The other two investigated tests and the total score showed no tendancy towards a consistent level of significant intra- or interobserver reliability. Copyright 1997 Harcourt Publishers Ltd.