Introduction: QTc dispersion has traditionally been calculated from al
l 12 leads of a standard electrocardiogram (EGG), It is possible that
alternative, quicker methods using fewer than 12 leads could be used t
o provide the same information. Methods and Results: We have previousl
y shown a difference in QTc dispersion from ECGs recorded at least 1 m
onth after myocardial infarction between patients who subsequently die
d and long-term survivors, In the current study, we recalculated QTc d
ispersion in these ECGs using different methods to determine if the ob
served difference in QTc dispersion measurements between the two group
s, as calculated from 12-lead ECGs, persisted when using smaller sets
of leads, QTc dispersion was recalculated by four methods: (1) with th
e two extreme QTc intervals excluded; (2) from the six precordial lead
s; (3) from the three leads most likely to contribute to QTc dispersio
n (aVF, V-1, V-4); and (4) from the three quasi-orthogonal leads (aVF,
I, V-2), For each of the 270 12-lead ECGs examined, a mean of 9.9 lea
ds (SD 1.5 leads) had a QT interval analyzed; the QT interval could no
t be accurately measured in the remaining leads, Using the standard 12
-lead measurement of QTc dispersion, there was a difference in the fal
l in QTc dispersion from early to late ECG between the groups: 9.1 (SD
60.8) msec for deaths versus 34.4 (55.2) msec for survivors (P = 0.01
6), This difference in QTc dispersion between early and late ECGs was
maintained using either three-lead method (quasi-orthogonal leads: -2.
6 [56.2] msec for deaths vs 26.9 [54.3] msec for survivors [P = 0.003]
; ''likeliest'' leads: 8.6 [64.9] msec vs 29.5 [50.2] msec [P = 0.05])
, but not when using the other two methods (precordial leads: 19.1 [55
.5] msec vs 22 [50.8] msec [P = 0.76]; extreme leads removed: 9.2 [50.
1] msec vs 21.8 [42] msec [P = 0.13]). Conclusion: QTc dispersion calc
ulated from three leads may be as useful a measurement as QTc dispersi
on calculated from all leads of a standard EGG. Its advantages over th
e standard measurement are its simplicity and the lack of problems Wit
h lead adjustment.