Reliability of echocardiographic measurements of left ventricular systolic function in potential pediatric heart transplant donors.

Background: Echocardiogram reports, but not images, are usually available for the evaluation of potential donor hearts. To assess the reliability of local reports of potential pediatric heart donors, we compared echocardiographic measurements of left ventricular (LV) systolic function between local hospitals and a central echocardiography laboratory.

Methods: We identified all potential donors aged <18 years managed by the California Transplant Donor Network from 2009 to 2013. Echocardiograms and reports were obtained from local hospitals. All studies were reviewed in a central laboratory by an experienced pediatric cardiologist blinded to local reports. Local and central measurements of fractional shortening (FS) were compared using the Bland-Altman method (mean difference ± 2 standard deviations). LV function was categorized based on FS as normal or mild, moderately, or severely depressed.

Results: There were 70 studies from 59 donors with local and central measurements of FS. The mean difference between local and central FS was 3.9 ± 9.0. The limits of agreement ranged from -14.2 to 22. Twenty-five studies had discordant measurements of LV function, with 17 discordant by 1 category and 8 by 2 or more categories. Of 55 studies categorized as normal by local measurement, 6 were moderately to severely depressed by central review. Of 15 studies categorized as depressed by local measurement, 3 were normal by central review.

Conclusions: Local and central measurements of LV systolic function were discordant in 36% of studies. Given such discordance, efforts to obtain and view actual echocardiographic images should be part of the standard evaluation of potential pediatric heart donors.