Page 36 - THE PERCEPT STUDY Illness Perceptions in Physiotherapy Edwin de Raaij
P. 36

Chapter 2
 change of three points or more, in repeated measurements, indicates a real difference within these dimensions. Two dimensions (Emotional Response, Coherence) scored 3.04, and one dimension (Personal Control) scored 3.37 as measurement error. This means that only a change of four points or more, in repeated measurements, indicates a ‘real’ difference in these dimensions. The absolute measurement error for group evaluation purposes for all eight Brief IPQ-DLV dimensions was less than one point. This means that a change of one point, in repeated group measurements, indicates a ‘real’ difference in Brief IPQ-DLV outcome out. Generalization of measurement error is limited because only patients with COPD were included in our study. However, our study design met 9 out of 11 items on the COnsensus- based Standards for the selection of health Measurement INstruments checklist which is a checklist that can be used to rate the quality of the design of studies on measurement properties25. The criterion on sample size and stable response could not be met completely. Four out of 26 patient reported improvement in health during the agreement assessment. So at least 20% of the patients reported change of health, which may have affected the results of the SDC.
For concurrent validity we hypothesized a significant association between a validated Dutch questionnaire and an illness perception dimension. However, equivalent questionnaires were only found for 5 out of 8 dimensions of the Brief IPQ-DLV. For four of these five equivalent questionnaires showed low to moderate significant correlations with the IPQ dimensions (Table 5). The Treatment Control dimensions showed a low non-significant correlation. This may be due to low variability in answers to this question23. Further analyses indicated that the scores on the illness perceptions Treatment Control dimension indeed showed low variability with a median and mode being 8, on a 0-10 rating scale.
The validity of concurrent validity assessment of illness perceptions dimensions with validated Dutch equivalent questionnaires can be debated. The underlying rationale and concepts of the questionnaires used are not exactly the same as the equivalent question of the Brief IPQ- DLV. For example, the Concern dimension (item 6) was compared with State anxiety scale27. Whether ‘state anxiety’ is the right concept to assess concurrent validity for the IP-dimension Concern is not clear. In Table 5, State anxiety showed a higher significant correlation with the Emotional Response than with Concern. Interpretation of the concurrent validity correlations must therefore be done with some caution. Unfortunately although a gold standard for comparison representing a similar (theoretical) construct is not available.
34






























































































   34   35   36   37   38