Page 6 Heart Failure Remote Monitoring Evidence From the Retrospective Evaluation of a Real World Remote Monitoring Program

Methods

Overview

This study is a retrospective analysis to evaluate the effect of CCCP on clinical outcomes in patients
enrolled in the program in a 1:1 match cohort study. Matching on potential confounders is a
methodology that is commonly adopted to increase efficiency [8, 9]. It is generally suitable for

situations where the investigators have access to large population data sources [9].
CCCP participants were compared with control patients, within the same health care system, not
enrolled in the CCCP. The control patients received the usual standard of care at MGH. A 1:1 individual
matching was done to identify controls for each CCCP participant by selecting a control patient that had

a hospitalization within 30 days of the corresponding CCCP’s patient index hospital admission. Every
patient enrolled in the CCCP program must have an index hospitalization. The index admission is the last
heart failure–related hospitalization a patient must have prior to enrollment in the CCCP. Other matching
parameters are age ±2 years, race, and gender. We used the matching without replacement method. In

this method, once a matched patient from the control population is selected, they are no longer eligible
for subsequent selection. The best matched control was selected to maximize the precision of the
analysis. Gender and race were the first considerations in selecting the best match, followed by the
nearest age and index admission date in that order.

Study Population

All subjects included in this study are patients with a diagnosis of heart failure receiving care at MGH.
Eligible participants were English-speaking heart failure patients, who had a Partners HealthCare primary
care provider or cardiologist that utilized the electronic medical record. They were also required to have

a hospital admission to a Partners HealthCare hospital to be eligible to participate in the CCCP. Eligibility
requirement for enrollment in the program included that patients must have a diagnosis of heart failure,
must have an MGH care provider, and must have been flagged as having a high risk for readmission and

pg. 6