The test-retest reliability of a questionnaire on the occurrence and severity of back pain in a German population sample

Objectives: Back pain is considered as one of the most frequent disturbances of health. The available study aimed at the evaluation of the test-retest reliability of fundamental questions on the occurrence and severity of back pain that have been widely used in international population surveys. Methods: Four hundred and eighty-seven inhabitants of Lübeck/ Germany aged 18 to 74 were mailed a first questionnaire on back pain. All respondents were resent the identical questionnaire two weeks after initial mailing. Additionally, they received a questionnaire on selected back pain items covering the time interval between the test and retest survey to identify cases with instable answering patterns due to back pain related changes, exclusively. Results: One hundred and seventy-nine respondents took part in both the test and retest survey (response rate 36.8 %) and were included in our reliability analyses. Respondents of the first questionnaire were of higher social status than the nonrespondents. Dependent on scale level, Cohen's kappa, quadratic weighted kappa, and intraclass correlation coefficients were computed, respectively. Conclusions: All evaluated questions reached good up to very good reliability scores.