GE Marquette Responder 3000, Руководство оператора

Страница: 87 / 90

The Arrhythmia Detection Program
227 490 02-C Marquette Responder® 3000 87
Appendix
The Arrhythmia Detection Program
Overview
The arrhythmia detection program used in the
Marquette Responder® Series defibrillators was
developed by Marquette Medical Systems, Inc.
After initial performance evaluation in the
laboratory,(against recognised databases .i.e. MIT*
and AHA**), independent clinical evaluation was
performed in hospital and field settings. The
hospital tests were performed in over 500
witnessed cardiac arrests and the program
performance was very successful. Pre-hospital
evaluation followed and the arrhythmia detection
program was found to perform equally well.
Moreover, the algorithm was tested against the
Marquette Hellige database with over 150 true
prehospital ECG rhythms.
Arrhythmia Detection
The “accuracy” of any arrhythmia program is
quantified in terms of;
−
Sensitivity: in this case, the ability to identify
correctly a shockable rhythm
−
Specificity: the ability to correctly identify
non-shockable rhythms
An ideal system would have a sensitivity and
specificity of both 100%, but in practice there is
always a trade-off between the two.
* Massachusetts Institute of Technology
** American Heart Association
*** Association for the Advancement of Medical Instrumen-
tation
The consensus opinion published by
AAMI***(TIR #2-1987), is that specificity should
be as high as possible, even at the sacrifice of
some sensitivity. This approach minimizes the
likelihood of shocking a non-shockable rhythm.
However, this does mean that a system may not be
able to identify every shockable rhythm.
The arrhythmia detection program of the
Marquette Responder® 3000 achieves a sensitivity
of 86 % and a specificity of 99.8 %. These values
refer to ventricular fibrillation and ventricular
tachycardia rhythms.
VT is shockable in the presence of a non-
perfusing, high rate VT. There is no consensus as
to what constitutes a “high rate”. Therefore, in
order to achieve a high specificity a rate of 120
bpm is used to identify shockable VT.
The algorithm uses a ‘Majority Decision’
approach to advising a shock, whereby two-of-
three ECG analysis segments must be in agree-
ment(each segment is 4 s).
The overall performance of the algorithm was
tested in ambulance trials. It is important to bear
in mind the many factors that can affect the
performance of such a system. These include skin
preparation, electrode quality and placement, as
well as patient or vehicular motion.