Effect of scrambling on the short-term reliability of the American Urological Association Symptom Index

Objectives. To determine the degree of attention patients pay to the wording of symptom scoring instruments and to determine the short-term reliability of the American Urological Association Symptom Index (AUA SI). Methods. The AUA SI was administered to 111 volunteers with a mean age ± SD of 50.8 ± 16.4 years (range 23 to 83) in the standard formatting. Without debriefing, the same volunteers were given the AUA SI a second time within 2 weeks. By sequential assignment, 65 of the subjects were given the AUA SI in the same standard format, and for the remaining 46, the sequence of the questions was scrambled and the order of the answers reversed for the second administration. In addition, the Benign Prostatic Hyperplasia Impact Index (BPH II) and the quality of life (QOL) question were answered and urinary flow rates were performed on both occasions to compare the short-term variability between the 'subjective' symptom score and the 'objective' flow rate recordings. Results. In this group of volunteers with a mean age of 50.8 years, the mean AUA SI was 11.7 points. The mean values in the unscrambled group were 12.4 and 12.9 (NS) and in the scrambled group 10.5 and 10.2 (NS) for the two administrations. Cumulative frequency distribution of differences were nearly identical. Similarly, there were no differences in the mean values for the two administrations in either group for the BPH II, the QOL question, or the peak flow rate. The reproducibility was excellent for each of the individual seven questions in both groups. There was no effect of age (younger or older than 50 years) on the reproducibility, although in general the variability of the scores was higher in older men, presumably due to a higher score in the first assessment. Conclusions. Subjects pay relatively close attention to the questions of symptom score questionnaires. The reproducibility for each individual question, as well as for the entire score, was very high in the original unscrambled and in the scrambled version. In addition, the short-term variability of the AUA SI is comparable to that of the flow rate recording. These observations should give confidence to those using such scores in clinical practice and clinical research.