Using Mental Health Telemetry to Predict Relapse and Re-hospitalization in Mood Disorders

The mood disorders (major depressive disorder, bipolar disorder, and dysthymia) are a significant public health issue: mood disorders affect approximately one in four people during their lives; in total, over 8 million Canadians are affected by mood disorders, costing the economy over $6 billion annually. Mood disorders are generally recurrent: approximately 50% of cases of major depressive disorder will relapse at least once; 15% will run a chronic course. With bipolar disorder, a chronic course of episodes of depression and/or mania intermixed with episodes of normal mood is typical. At Sunnybrook (SHSC), 75% of inpatient mental health admissions are due to mood disorders. Clinically, with mood disorders, earlier recognition of symptom changes in patients provides greater potential for early intervention and suppression of relapses, which in turn leads to reduced outpatient resource utilization, fewer Emergency Department visits, and fewer readmissions.

Ecological momentary assessment (EMA), a body of work developed in the late 90s and early 00's, focuses on using data collected from subjects living their daily lives in their natural environments using minimally invasive techniques to improve the quality of the data collected. Mental health telemetry (MHT) is an evolution of EMA. MHT was developed here at Sunnybrook in partnership with the Faculty of Medicine at the University of Toronto; it uses wirelessly networked handheld computers - typically, cell phones - to collect self-report data on symptoms of illness and then transmit it in real-time to a central database. Principal advantages of MHT over EMA include (i) the ability to time- and date-stamp data, thus eliminating retrospective record completion, and (ii) the ability to observe and monitor data flow in real-time, without having to wait for participants to upload or deliver the data (e.g., at their next doctors' appointment).

MHT has been used to collect rich sets of longitudinal self-report ratings from patients while they live their daily lives in the community; populations studied include adults with depression, premenstrual syndrome, and bipolar disorder (for as long as 18 months), as well as teens with mood swings (for as long as nine months). Adherence to daily (or more frequent) symptom reporting using MHT in these studies has been very good, with adherence rates of 75±5 % over the duration of the studies.

MHT has considerable potential to reduce rates of inpatient re-hospitalization for mood disorders. EMA methods have previously shown to be superior to standard patient follow-ups for detecting medication response in depression14 or days ill in bipolar disorder18. While numerous previous studies have looked at general patterns of recurrence and prognostic factors in the long-term course of major depressive disorder4;23;24;27 and bipolar disorder, none have looked at using EMA or MHT data to monitor for signs of imminent relapse in individual cases. The rich data stream provided by MHT will allow us to do this for the first time, similar to how ECG telemetry is routinely used for diagnostic and prognostic purposes. Doing so will enable clinicians to trigger early community- or outpatient-based interventions, thus reducing the likelihood of relapse and / or hospitalization.

One limitation of the MHT studies to-date has been the platform-specific nature of the MHT systems used: rather than using platform-independent software, the systems deployed to date have been custom-created for specific cell-phone platforms (e.g., the Kyocera pdQ 1900, the Palm Treo family of SmartPhones, or the Motorola QA30). As a result, until now, MHT has been limited use by participants in research studies who were provided with a research-funded cell phone.

Therefore, to explore the feasibility of using MHT to reduce re-hospitalization in patients with diagnosed mood disorders, we will:

1. Upgrade our existing patient MHT software (for recording MHT) to a web-based platform, thus removing all hardware dependency from MHT and allowing anyone who has a web-capable cell phone to use MHT essentially without cost;
2. Upgrade our existing clinician PATH software (for viewing patients' MHT), optimizing its user interface to make it easy for clinicians to rapidly and effectively visualize their patients' telemetry;
3. Distribute clinician PATH software to interested clinicians within the SHSC Department of Psychiatry, to enable clinicians to receive daily, patient-generated, quantitative symptom severity ratings as well as medication adherence and side-effect reports from their patients. (Clinicians will use this information in collaboration with their patients to aid in monitoring patients' mental status over the course of this study.)
4. Over the duration of the study, distribute patient MHT software to a random sample of patients with (i) a mood disorder being discharged from Sunnybrook's inpatient mental health ward (F2) or (ii) bipolar disorder being followed in the outpatient psychiatry clinics at Sunnybrook; then,
5. Compare re-hospitalization rates and quality-of-life measures of MHT users over a 6 month period to a control group of similar patients not using MHT.