User Manual Part 4

H-6
CLINICAL STUDY - SUMMARY OF GDT1000 SEN
SING ACUTE STUDY
Tab le H-4. S ummary of Sensing Performance - Second Phase (continued)
Sensitivity
(Rate o f
Undersensing)
Specificity
Positive
Predictive
Value
(Rate of
Oversensing)
Appropriate
Sensed
Beats
Inappropriate
Sensed
Beats:
Undersense
Inappropriate
Sensed
Beats:
Oversense
Left
Ventricular
Channel
99.98%
(0.016%)
99.99% 99.99%
(0.008%)
12,227 2 1
Totals 35,8 31 2 169
Comparing the performance between the two phases, there were 1,056
inappropriate sensing events out of 55,027 signals (1.919%) in the first phase
of the study, and a total of 171 inappropriate sensing events out of 35,998
signals (0.475%) in the second phase of the study, reflecting a 7 5. 2% reduction
in inappropriate sensing events from phase one to two.
Oversense Events
During the analysis of the first phase data, some unexpected oversense events
were identified. There were a total of 831 oversense events in the RA (615) and
RV (216) channels in phase one. The majority of the RA and RV oversense
events were attributed to an artificial event introduced while pacing. This type
of oversense was observed in 6 patients in the RA channel and 7 patients in
the RV channel. The results for the second phase of the study demonstrated
that oversensing artificial events observed in the first phase o f the study were
successfully eliminated by using the upgraded GASD system. There were no
artificial events introduced in the second phase of the study.
In the second phase, a total of 168 oversense events in the RA channel were
observed in one patient. This patient had an intrinsic P-R interval greater
than 300 ms. In order to complete the AP/VS test scenario, the device was
programmedwithaLRL=80bpmandAVDelay=300ms,whichisthe
maximum allowable AV Delay in a CRT-D device. At the end of the AV Delay,
no intrinsic activity occurred and the device paced both ventricles. These paced
beats were oversensed by t he atrial channel. If the atrial blanking period were
programmed to a larger value, atrial oversensing would have been eliminated.
Therefore, excluding this patient’s AP/VS test scenario from the analysis, there
were no undersensing or oversensing events in the RA channel.
In one patient, the single oversense event in the LV channel was caused by
noise.
- DRAFT -