Helping cancer-family caregivers with end-of-life home symptom management: Initial evaluation of an automated symptom monitoring and coaching system.

Abstract

85 Background: Family caregivers face the task of providing end of life care (EOL) in the home setting and need effective caregiving support. Although not previously tested in EOL care, automated symptom monitoring has shown promise for cancer patients on active treatment. This study tests an automated monitoring and coaching system for family caregivers during home hospice. Methods: Family Caregivers (FC) of hospice cancer patients (PT) (N=243 dyads) were monitored daily using an automated telephone symptom monitoring system, and randomly assigned to receive the symptom care intervention (SC) (N=119) or usual care (UC) (N=124). All FCs called the monitoring system daily, reporting by proxy, past 24 hour presence, severity and FC distress on 11 common PT symptoms (0–10 scale). In addition, FCs reported their own fatigue, sleep, mood and anxiety severity (0-10 scale). The SC intervention FCs received automated tailored symptom care coaching based on reported patterns; in addition symptoms moderate or greater generated alerts to their hospice nurse. Results: Most FCs and PTs were White (95; 94%), 66% of FCs and 48% of PTs were female with the Mean age for FCs 60 and 73 for PTs. Most frequent PT symptoms were fatigue (70%), pain (64%), poor appetite (54%), anxiety (39%), and change in thinking (38%). PT symptom severity levels were highly related to FC distress. Preliminary analyses indicates that FC vitality (FC combined symptoms) improved more in the SC group (p=.003, mixed effects model). Furthermore PT symptom severity, categorized as Mild (<=3), Moderate (4-6), or Severe (>=7), provided strong preliminary evidence of SC intervention benefit in each of the first four weeks in hospice care and overall (p=.003). Rapid onset of the SC benefit was suggested by Kaplan-Meier analysis of time to first all-zero symptom profile (p=.06 log-rank). The benefit of the SC intervention continued throughout the hospice stay until the last several days of life when there was no difference with UC. Conclusions: This is the first reported evaluation of automated symptom monitoring in home EOL care and provides strong preliminary evidence of benefit for cancer family caregivers and patients. Clinical trial information: NCT02112461.