User Manual Part 2

CLINICAL STUD Y - MADIT II
C-21
at implant u sing a catheter method and 24 (12 percent) using the ICD for
induction; there was no data on the method of induction for 6 patients.
The Occurrence of ICD Therapy for VT, VF, or VT/VF Combined
Therapy for VT was defined as antitachycardia pacing (ATP) or ICD shock
delivered by the device in an attempt to stop an arrhythmia as reported by the
enrolling center. The rap y for VF was defined as the delivery of ICD shock
therapy. The endpoint for VT/VF was defined by the occurrence of either VT or
VF therapy. The occurrence of therapy for each of these groups is provided in
Table C-12 on page C-21. All analyses were Cox regression analyses, stratified
by enrollment center, with time to VT, time to VF or time to VT/VF therapy
as the res pective endp oint.
Table C -12. ICD patients receiving one o r more therapies
Type of ICD Therapy
a
Number of Patients Percent of Patients with
Therapy E pisodes
VT (ATP or shock)
89
15.4%
VF (Shock only)
36
6.2%
VT/VF VF (ATP + shock)
114
19.7%
a. Some patients received both types of therapy.
Predictions of VT and VF Therapy in ICD Patients
A statistical analy sis was pe rformed to e valuate whether inducibility at EP
testing provides predictability of the potential effectiveness o f a n ICD. To this
end, the occurrence of each of the three endpoints defined above (VT, VF
and VT/VF), in ICD patients with EP testing were evaluated. Analyses were
done by Cox proportional hazards regression, stratified by enrollment center.
(Table C-13 on page C-22)
A list of potential risk factors was considered for these endpoints, such as age,
gender, and standard cardiological variables like NYHA class, EF, etc., and
developed a parsimonious regression m odel in the 583 ICD patients identified
above. GENDER and BUN (dichotomized at up to 25 versus 26 and over) were
observed as potential risk factors for these endpoints, with males and elevated
BUN associated with increased occurrence of these endpoints. Further
analysis in v e s tiga ted whether inducibility added an y additional, indepe nd ent
predictive power for each of these endpoints.
The conclusion was that inducibility increases the risk of VT events by perhaps
60 percent (p=0.07) and decreases the risk of VF events by perhaps 50
- DRAFT -