RESUMO
PURPOSE: The validity of the Talk Test (TT) is well documented, but the reliability of the test is not clear. The aim of this study was to assess the absolute and relative intertester reliability of the TT in cardiac patients. METHODS: Cardiac patients (n = 64) who had completed an exercise rehabilitation program were consecutively included in the study. Patients performed a submaximal ramp cycle ergometer test with the TT as intensity indicator. Every patient was tested twice on the same day by 2 different physiotherapists with a break of maximum 45 minutes between tests. Physiotherapists were randomized to tests. Workload in watts at the first negative stage of the TT was registered as the test result. Patients and physiotherapists were blinded to test results of the first test. Absolute reliability of the TT was assessed with Bland-Altman plot, standard error of measurement, and minimal detectable change. Relative reliability was assessed using the intraclass correlation coefficient (ICC). RESULTS: Mean difference in peak workload between test and retest was 0.8 W (95% CI: -4.8 to 3.3). Limit of agreement was estimated to be +31/-32 W. Standard error of measurement was 11 W (95% CI: 10-14), and minimal detectable change was 32 W. The ICC was 0.85 (95% CI: 0.78 to 0.91). CONCLUSIONS: Although the ICC was acceptable, we found a weak absolute intertester reliability of the TT and that the test is an insufficient measure to monitor exercise intensity and safety of cardiac patients when 2 or more physiotherapists may be administering the TT.