The reliability of echocardiographic left ventricular wall motion index to identify high-risk patients for multicenter studies

Objective: To study whether the use of echocardiographic left ventricular (LV) wall motion index (WMI) is a dependable parameter for identifying patients with LV dysfunction to be enrolled in multicenter trials. Methods: Videotaped echocardiographic examinations from 200 randomly selected patients that were screened for inclusion into the DIAMOND-CHF and DIAMOND-MI trials were reevaluated by an external expert echocardiographer. WMI was calculated using the 16-segment LV model. Results: The external echocardiographer systematically found lower values of WMI than the core laboratory. The average difference in WMI was 0.18 (SD: 0.33) in the DIAMOND-CHF trial and 0.09 (SD: 0.33) in the DIAMOND-MI trial. The difference in WMI exceeded 0.33 in 34% of the patients in both trials. The cutoff value for inclusion into the DIAMOND trials was WMI <= 1.2. There was an agreement on WMI dichotomized to below or above 1.2 in 82% of the patients in both trials (kappa coefficient 0.66 for the DIAMOND-CHF and 0.55 for the DIAMOND-MI). Conclusions: Despite substantial interlaboratory variation in WMI in individual patients and a systematic lower WMI score by the external echocardiographer there was an acceptable overall agreement for identifying patients with severe impairment of LV function. This not only underscores the value of LV-WMI as a useful tool for selecting high-risk patients to be included in multicenter studies but also serves to warn against the use of rigid cutoff values for WMI in the treatment of individual patients.