The purpose of this study was to examine long-term test-retest reliability of the ISWT in clinically stable cardiac patients. Thirty patients with cardiovascular disease (15 males, 15 females; age 55 to 80 years) volunteered to participate in the study.Participants undertook two ISWTs, a minimum of 8 weeks apart. Overall, the mean distance walked in the pre-test was 502+/-161m and this did not differ from test to retest. The intraclass correlation coefficient was 0.80, indicating good test-retest reliability. Using the Bland and Altman method, there was a small mean test-retest difference (-7m). The 95% limits of agreement were large, ranging from -203m to 189m.
Over long test-retest durations, there appears to be no learning effect in the ISWT, negating the need for a practice walk. The long-term random variation in the ISWT test is larger than in previous studies, probably due to greater physiological and psychological variation in the participants over 8 weeks compared with that seen in day-to-day testing. Factors influencing long-term test-retest reliability of the ISWT require further elucidation.