A randomized control trial to determine necessary intervention elements to achieve optimal symptom outcomes for a remote symptom management system.

12008 Background: Interventions to improve patient-reported cancer symptom outcomes (PROs) are often multicomponent. In order to scale efficacious interventions, it is important to know if all components are needed to achieve the greatest benefit. We deconstructed the Symptom Care at Home (SCH) monitoring and management system, a multicomponent intervention found to be highly efficacious in reducing symptom burden for patients at home. SCH remotely monitors daily PROs, provides tailored automated self-management coaching based on the severity pattern reported for 11 monitored symptoms and alerts a nurse practitioner (NP) of poorly controlled moderate to severe symptoms. The NP follows up to intensify symptom care utilizing the SCH decision support system based on national guidelines. Methods: A 5 group randomized controlled design was utilized with 755 participants assigned to PRO assessment plus: Gp 1, (N = 143) automated self-management coaching, Gp 2, (N = 144) automated self-management coaching plus an activity tracker, Gp 3, (N = 147) NP follow up without decision support, Gp 4, (N = 154) NP follow up utilizing decision support or Gp 5, (N = 167) automated coaching plus NP follow up utilizing decision support- the complete SCH intervention. The primary outcome was maximum likelihood estimation of overall symptom severity over the duration of study participation (62 day median). Results: Patients were recruited from sites in Utah and Georgia. The majority of patients were female (60.3%), married (59.9%), 60.7% Caucasian and 34.9% Black, with a mean age of 59.4 years and Stage 3 or 4 cancer (65.7%). Most common diagnoses included breast (17.1%), lung (13.9%), colorectal (12.9%), pancreatic (8.6%) and ovarian (8.4%). Adherence rate for daily symptom reports was 73.8% with no differences between groups. 84.2% of patients reported one or more moderate or severe symptoms- most commonly fatigue (70.1%), pain (64.8%) and trouble sleeping (56.0%). Group 5- the complete SCH intervention had significantly greater symptom relief than any other group (Gp 5 vs Gp 1 or Gp 2 automated coaching both p <.001; Gp 5 vs Gp3 NP without decision support p =.039; Gp 5 vs Gp 4 NP with decision support p =.017). The two NP groups did not differ significantly but were significantly better than either automated coaching groups. The two automated coaching groups did not differ significantly from one another. Conclusions: PRO assessment with automated coaching and NP decision-supported follow-up for moderate to severe symptoms achieved better symptom reduction than the individual components. Better symptom relief can be obtained for patients at home through a synergistic, multicomponent intervention that combines a tailored symptom coaching component with notification and follow up by clinicians when symptoms rise to moderate or severe levels- both components contribute to better outcomes. Clinical trial information: NCT02779725.