Page 9 Heart Failure Remote Monitoring Evidence From the Retrospective Evaluation of a Real World Remote Monitoring Program

method. The Cox proportional hazard model using time to death and hospitalization as endpoints was
used to estimate hazard ratios (HR). Multiple hospitalizations were accounted for in the model. All

analyses were stratified on the matched pairs [8], and we also adjusted for age, gender, race, ejection
fraction, and New York Heart Association classification (NYHA). A two-sided P<.05 was considered as
significant. All analyses were performed using data analysis and statistical software, STATA 12 version.

Results

Sample

Figure 2 depicts the sample selection process. A total of 510 patients were enrolled in the CCCP from
Jan. 1, 2008, to Aug. 31, 2012. Of these, 116 patients were excluded from matching. A total of 348

patients, 174 enrolled in CCCP and 174 match controls, were included in the final analysis.
The majority of the sample population were Caucasian males with an average age of 77 years in both
groups, but they differed by marital status. The median duration of follow-up, 365 days, was similar in
both groups. The baseline rates of hospitalization, 120 days prior to CCCP enrollment, did not differ

significantly in the 2 groups (HR=1.02, 95% CI 0.83-1.24, P=.87). All patients were followed up for a
maximum duration of 1 year from the period of program enrollment for CCCP patients and their
corresponding controls. All baseline characteristics are summarized in Table 1.

pg. 9