Evaluation of SCL-90R and it's short forms in patients with vertigo

Abstract

Objective: Psychiatric symptom screening tests for patients who suffer from imbalance and dizziness are generally used in majority of clinics. Symptom checklist-90 Revised (SCL-90R) is one of these tests. The aims of this study are to analyze the Turkish version of SCL-90R, compare to its short version and investigate whether if it is possible to create a short version or not. Material and Method: The sample of this study consisted of 2100 patients who admitted to tertiary care neurotology department for imbalance and dizziness. Three different SCL-90R versions as the original, Brief Symptom Inventory (BSI) and Brief Symptom checklist-18 (BSI-18) were retrospectively studied. These three tests were evaluated with regards to global severity index, totalpositive symptoms and positive symptom index. The correlation between results were compared by Pearson correlation test. Structural integrity studies were also conducted with using raw data of SCL-90R. Results: There were strong correlations among the three questionnaires. While the structural distribution of questions are consisted of 10 factors in the standard form (SCL-90), the short form (containing 49 questions) which was created from our results was found to have different structural distribution. Moreover the distribution of questions were quite different than the actual structure and the questions from different factors were grouped under different factors. The short form, which was created from our results, showed similar properties with the BSI. The internal consistency reliability test showed significance for 9 of them. Eventually, we created a short form consisting of 9 factors and 49 questions. In this state, it shows similar features with BSI and contains a similar number of questions. Conclusion: Due to the difference in structural analysis, the results obtained by the sub-scales of the questionnaires should be evaluated very carefully. We think the short form we created can be used easily since the general values resulted from it are comparable to the the original questionnaire.