Adherence to the Use of Home Telehealth Devices by Veterans with Heart Failure

Va Greater Los Angeles Healthcare System, Los Angeles CA

Investigators

Linked publications, trials & patents

Abstract

? DESCRIPTION (provided by applicant): Problem: Poor patient adherence to use of home telehealth (HT) devices has been proposed as a possible reason for negative results in randomized controlled trials of patients enrolled in HT programs for heart failure (HF) management. Objectives: The objectives of this retrospective, descriptive study in the Veteran population with heart failure newly enrolled in the VA HT program are to: (a) describe patient adherence to the use of HT devices in the first year after enrollment; (b) examine the relationship between patient adherence to the use of HT devices (at one month, three months, six months, and 12 months after enrollment) and health services use at six and 12 months after enrollment; and (c) identify patient characteristics that influence patient adherence to the use of HT devices at one, three, six, and 12 months after enrollment. Methods: This is a retrospective cohort design study using existing VA data from the national VA HT Database and the VA Corporate Data Warehouse and will include all Veterans with HF who enrolled in the VA HT Program between 1/1/14 and 6/30/14. The Andersen Behavioral Model will be used to guide the study. Patient adherence will be calculated as ratios using dates that the patient logged in to use their HT device during a specific time period as the numerator, and the number of days in that specific time period as the denominator (e.g., 30 days for one month, 60 days for three months, etc.). Use of health services include emergency room (ER) visits, primary/ambulatory care clinic visits, cardiology clinic visits, hospital admissions, hospial readmissions, time to hospital readmission, and bed days of care. Finally, potential patient variables that may influence patient adherence to the use of HT devices include (1) predisposing (i.e., age, marital status, race, religion, gender, caregiver availability, living alog status, medication refill history, health service use during the year prior to HT enrollment, enrollment in my HealtheVet, and presence of an implantable cardiac device, pacemaker, or left ventricular assist device), (2) enabling (i.e., eligibility for VA services, income, and availabiliy of non-VA insurance), and (3) need factors (i.e., Charlson Comorbidity Index, Care Assessment Need Score, mental health diagnoses, reason for enrollment in a VA HT program, and severity of heart failure). Available health care system factors (i.e., urban/rural setting, size of facilit, distance to assigned facility, and academic affiliation) will be obtained as control variables. Descriptive statistics will be used to summarize sample characteristics and determine the distribution of adherence at various time frames. Patterns of adherence will also be described and coded as yes/no depending on how daily adherence is observed over time in a categorical pattern (e.g., early adherent versus late adherent, or adherent weekly versus daily, etc.). Analysis of patterns using this type of adherence will use the general estimating equation (GEE) technique and have the facility as a covariate in the correlation structure to adjust for potential hierarchical clustering. To examine the relationship between adherence and health service use, the GEE technique will again be used because data is expected to be highly correlated within patients and within facilities. To determine factors that influence adherence, factors found to be significant in bivariate analyses will be placed in a multivariate multilevel logistic regression model to determine the relationship with low/high adherence using a 70% cut off for each time period with all independent variables being entered simultaneously. A block entry design will also be used for each set of variables based on the Andersen Model.