INTRODUCTION AND HYPOTHESIS: The objective of this study was to determine the test-retest reliability of the cotton swab test. METHODS: Prospective, blinded, study of the cotton swab test. Subjects underwent two tests within 2-4 weeks. Agreement between resting angles and straining angles was analyzed using a Wilcoxon signed rank test. Agreement of a positive and negative test result, using 30 degrees as cutoff, was evaluated with McNemar's test and kappa statistic. RESULTS: Fifty subjects participated. The mean resting angle was 3.16 degrees (-6.0 degrees, 10 degrees) and 3.24 degrees (-5.0 degrees, 10 degrees) on repeat (p = 0.99). The mean straining angle was 34.4 degrees (20 degrees, 50 degrees) and 37.0 degrees (25 degrees, 55 degrees) on repeat (p = 0.14). McNemar's test demonstrated good agreement between observations. As the straining angle approached 30 degrees, the agreement between positive and negative exams was only 57%. CONCLUSION: There is good test-retest reliability of the cotton swab test. When the straining angle is 21-49 degrees, there is poor agreement regarding a positive or negative test result.
INTRODUCTION AND HYPOTHESIS: The objective of this study was to determine the test-retest reliability of the cotton swab test. METHODS: Prospective, blinded, study of the cotton swab test. Subjects underwent two tests within 2-4 weeks. Agreement between resting angles and straining angles was analyzed using a Wilcoxon signed rank test. Agreement of a positive and negative test result, using 30 degrees as cutoff, was evaluated with McNemar's test and kappa statistic. RESULTS: Fifty subjects participated. The mean resting angle was 3.16 degrees (-6.0 degrees, 10 degrees) and 3.24 degrees (-5.0 degrees, 10 degrees) on repeat (p = 0.99). The mean straining angle was 34.4 degrees (20 degrees, 50 degrees) and 37.0 degrees (25 degrees, 55 degrees) on repeat (p = 0.14). McNemar's test demonstrated good agreement between observations. As the straining angle approached 30 degrees, the agreement between positive and negative exams was only 57%. CONCLUSION: There is good test-retest reliability of the cotton swab test. When the straining angle is 21-49 degrees, there is poor agreement regarding a positive or negative test result.