T96. RESULTS OF THE HUMMINGBIRD STUDY A MULTICENTRE, PRAGMATIC TRIAL OF A DIGITAL MEDICINE SYSTEM

BACKGROUND: This was a multicentre, 8-week, single-arm, open-label, pragmatic trial to explore the acceptance and performance of using the Digital Medicine System (DMS) with health care professionals (HCPs) and adult subjects with schizophrenia, schizoaffective disorder (SAD), or first episode psychosis (FEP) on an oral atypical antipsychotic (aripiprazole, olanzapine, quetiapine, or risperidone). METHODS: Subjects received an initial introduction to the DMS, and had HCP visits at screening/baseline, Week 4, Week 8/early termination (ET), and as directed by the HCP for the duration of the subject’s participation in the trial. Safety and tolerability data was collected and evaluated on an ongoing basis, as assessed by the frequency and severity of serious adverse events (SAEs), and device-related non-serious adverse events (AEs). Subjects were monitored on the DMS technology by the HCPs through review of the HCP dashboard data at a minimum of every 2 weeks and to make changes to the current treatment plan and therapy at their discretion. The study initiated in May 2018 and concluded in September 2019. NCT03568500 RESULTS: Fifty-five (55) subjects were screened, and forty-three (43) subjects were treated and included in the sample analysis. Subjects enrolled were on average 34.4 years old. There were twenty-eight (28) male and fifteen (15) female subjects in the study. The most common reasons for discontinuation was subject withdrawal of consent (5 subjects, 11.6%), loss to follow-up (4 subjects, 9.3%), and adverse events (4 subjects, 9.3%). The primary endpoint was the proportion of days with good patch coverage during the trial, calculated by the number of days with good patch coverage divided by the total number of trial days for each subject. Over the duration of the study, subjects had 63.4 % (SD = 26.6%) of days with good patch coverage during the trial time. No notable differences were observed across the disease types, with schizophrenic subject 64.3% (SD=20.2%), FEP subject 62.5% (SD = 27.5%), and SAD subject 63.0% (SD = 37.7%). The secondary endpoint was subject adherence, defined as the proportion of detected ingestible events markers (IEM) over the expected during the trial days with good patch coverage. Overall, subjects had a mean of 86.6% (SD = 14.5%) IEMs detected during the good patch coverage days. The proportion of IEM detected by disease type were: FEP 91.0% (SD = 7.4%), schizophrenic 88.9% (SD = 8.1%), and SAD 72.3% (SD=25.7%). There were nine (9) subjects (20.9%) who had eleven (11) AEs during the study, all were non-serious. Out of all the AEs, nine (9) subjects had nine (9) treatment emergent adverse events (TEAE) that were all reported as medical device site irritation, and four (4) subjects discontinued the study due to AEs. DISCUSSION: In conclusion, during the 8-week treatment with the DMS, subjects reported good patch coverage 63.4% of the time on average for this pragmatic study. Using the DMS, subjects’ medication adherence reached 86.6% on average when subjects were having good patch coverage. The acceptance and performance of DMS is considered safe and there were no SAEs associated with its use. FUNDING: This study was supported by Otsuka Pharmaceutical Development & Commercialization, Inc.