Background Although echocardiography is used extensively in clinical medici
ne, guidelines for quantitative interpretation of echocardiographic measure
ments are unavailable. The goals of this investigation were to provide an o
verview of scientific standards for formulating reference values, with clin
ical chemistry used as a model, to evaluate published echocardiographic ref
erence limits, to survey clinical echocardiography laboratories regarding t
heir interpretation of echocardiographic measurements, and to provide recom
mendations for improving the interpretation and reporting of echocardiograp
hic measurements.
Methods and Results We reviewed the original reports of the international F
ederation of Clinical Chemistry on guidelines for formulating reference val
ves. We obtained published reports on echocardiographic reference limits th
rough searches of electronic databases supplemented by a manual search of r
elevant bibliographies. We also surveyed echocardiographic laboratories in
35 adult acute-care hospitals in Eastern Massachusetts. Studies on echocard
iographic reference values were evaluated with the use of guidelines from c
linical chemistry. Responses from the 29 participating echocardiographic la
boratories were evaluated for their practice of quantitative echocardiograp
hic interpretation. There is considerable heterogeneity in the echocardiogr
aphic reference values available in the literature. There is also a lack of
agreement in the literature and among echocardiographers regarding the par
titioning of reference values (by sex, ethnicity, or age), the anthropometr
ic measure to be used for indexation, and the choice of cut-points for cate
gorizing values within the abnormal range.
Conclusions We advocate that echocardiographic reference limits be standard
ized and a consensus generated regarding the partitioning of reference limi
ts and the indexation of echocardiographic measurements, Such measures can
aid in quantitative echocardiographic interpretation and render the results
more scientific and consistent.