81 The Otology Questionnaire Amsterdam 4 heterogeneous group by including patients of different ages, sexes, complaints and pathologies. Furthermore, the cohort in which the responsiveness of the OQUA was assessed had an additional heterogeneous distribution of surgical procedures. Although the results confirm our hypotheses and show satisfactory results for construct validity, test-retest reliability and responsiveness, some limitations of the study should be mentioned. Unfortunately, there is no consensus in literature about the group size to evaluate validity. Fifty patients is advised by de Vet et al.6 Accordingly, 50 patients were included in the cohort for examining the OQUA’s responsiveness, but due to missing data, three patients had to be excluded. However, the remaining sample is thought to be a good representation of a general ENT practice and the number big enough to examine responsiveness. On the other hand, one could argue that this amount is critical for confirmation of responsiveness in each domain. The GRS used to examine OQUA’s complaints scores did not distinguish in the pre-determined separate domains (e.g. tinnitus, hearing loss). Rather, the GRS asked for a change in complaints overall. Using this method, complaints might have cancelled each other out in the correlation with GRS or downsized the effect. For example, at baseline (pre-surgery), hearing loss is the most prominent complaint of patient which improves in after surgery. In addition, the same patient might suffer from vertigo after surgery which was not present pre-surgery. The positive and negative result may have downsized the correlation in the responsiveness statistics. Furthermore, the patient might relate the GRS to his/hers most prominent complaint and not to the broader spectrum of complaints. Another limitation could be a partial floor effect; when patients fill-out rather low scores at the preoperative moment, there will be little to no room for improvement 3 months after surgery. The decision to calculate separate domain scores and not a total complaint score covering all domains was made based on the fact that the complaints are all separate domains in a formative model and can be seen as different (sub)constructs. Meaning that the items of the different domains not necessarily correlate with each other as can been seen in the development paper of the OQUA by Bruinewoud et al 2018.5 By using separate scores, a more accurate insight in the patients’ scala of complaints is given. This could, on the other hand, also be seen as a disadvantage, as it can be challenging to compare the results of the complaints with other cross-sectional studies in literature. Many questionnaires often use a single score for the total of complaints or a score for the complaints and impact as a whole. Up till now, studies often used a complaint- or disease-specific questionnaire. Complaint-specific questionnaires can
RkJQdWJsaXNoZXIy MjY0ODMw