Patients suffering from depression have a high risk of relapse and
readmission in the weeks following discharge from inpatient wards. Electronic
self-monitoring systems that offer patient-communication features are now
available to offer daily support to patients, but the usability, acceptability,
and adherence to these systems has only been sparsely investigated.
We aim to test the usability, acceptability, adherence, and clinical
outcome of a newly developed computer-based electronic self-assessment system
(the Daybuilder system) in patients suffering from depression, in the period
from discharge until commencing outpatient treatment in the Intensive
Outpatient Unit for Affective Disorders.
Patients suffering from unipolar major depression that were referred from
inpatient wards to an intensive outpatient unit were included in this study
before their discharge, and were followed for four weeks. User satisfaction was
assessed using semiqualitative questionnaires and the System Usability Scale
(SUS). Patients were interviewed at baseline and at endpoint with the Hamilton
depression rating scale (HAM-D17), the Major Depression Inventory (MDI), and
the 5-item World Health Organization Well-Being Index (WHO-5). In this
four-week period patients used the Daybuilder system to self-monitor mood,
sleep, activity, and medication adherence on a daily basis. The system
displayed a graphical representation of the data that was simultaneously
displayed to patients and clinicians. Patients were phoned weekly to discuss
their data entries. The primary outcomes were usability, acceptability, and
adherence to the system. The secondary outcomes were changes in: the
electronically self-assessed mood, sleep, and activity scores; and scores from
the HAM-D17, MDI, and WHO-5 scales.
In total, 76% of enrolled patients (34/45) completed the four-week study.
Five patients were readmitted due to relapse. The 34 patients that completed
the study entered data for mood on 93.8% of the days (872/930), sleep on 89.8%
of the days (835/930), activity on 85.6% of the days (796/930), and medication
on 88.0 % of the days (818/930). SUS scores were 86.2 (standard deviation [SD]
9.7) and 79% of the patients (27/34) found that the system lived up to their
expectations. A significant improvement in depression severity was found on the
HAM-D17 from 18.0 (SD 6.5) to 13.3 (SD 7.3; P<.01), on the MDI from 27.1 (SD
13.1) to 22.1 (SD 12.7; P=.006), and in quality of life on the WHO-5 from 31.3
(SD 22.9) to 43.4 (SD 22.1; P<.001) scales, but not on self-assessed mood
(P=.08). Mood and sleep parameters were highly variable from day-to-day.
Sleep-offset was significantly delayed from baseline, averaging 48 minutes (standard
error 12 minutes; P<.001). Furthermore, when estimating delay of sleep-onset
(with sleep quality included in the model) during the study period, this showed
a significant negative effect on mood (P=.03).
The Daybuilder systems performed well technically, and patients were
satisfied with the system and had high adherence to self-assessments. The
dropout rate and the gradual delay in sleep emphasize the need for continued
clinical support for these patients, especially when considering sleep