in normal hearts than
with coronary disease and wall motion abnormalities, also explaining our
high variability. However, the aim of our study was to reflect daily
clinic, in which the abovementioned limitations are inevitable.
Interestingly, new indexes of diastolic function seemed more
reproducible compared to traditional (TR Vmax CV = 9.9-12 % vs. MV E
and A CV = 12-20 %, LA volume CV = 19-22 % and E/e’ CV = 22-36 %).
The overall moderate reproducibility found in our study is, however, not
unique (31, 32).
The 2016 algorithm was superior at classifying patients compared to 2009
(Table 4a and 4b, appendix). This emphasises the dominance of an
algorithm’s influence on grading, which can partly diminish the
consequences of a high variability of the individual parameters.
Through reclassification analysis of the total (n=356) population, did
the 2016 algorithm proved more restrictive than 2009 in classifying
patients with advanced stages of DD (table 5). Other authors have
reported similar decrease in the prevalence of DD by comparing the 2009
and 2016 algorithms (23, 33, 34). Elsewhere, the 2016 algorithm have
likewise been shown superior to the 2009 in its ability to correlate
with clinical outcome (35-37)
It is a natural chain of reasoning that a stronger reproducibility could
have yielded improved re-classification. Therefore, the
re-implementation of E/e’ average in the current guidelines (9) is
striking in light of its considerable variability. We found the
reproducibility of E/e’ average, most likely, is hampered by E/e’
lateral (intra- and inter-observer CV = 36 % and 45 % for E/e’ lateral
vs. CV = 22 % and 23 % for E/e’ septal), suggesting replacement of
E/e’ average with E/e’ septal. Further elaboration is needed for the
optimal differentiation between normal and abnormal diastolic function
and subsequent grading. Rather than more expert opinions there is a need
for studies that relate novel classification schemes containing the most
feasible, reproducible, time-efficient, and prognostic variables to
clinical outcome.