Earlier work involving the University of Oxford Symptom Monitoring System established that technology-assisted symptom monitoring was acceptable to patients over a period of 36 weeks with 75% compliance [4], meaning that, on average, patients reported symptoms in response to prompts 75% of the time over follow-up. The OXTEXT program was dedicated to developing and validating the True Colours remote symptom monitoring system for patients, at a larger scale. Across several projects, the program revised the software after in-depth patient consultation, established a cohort of well-characterized patients with BD by using the improved system, determined the potential cost-effectiveness of this remote capture tool in clinical service, and tested remote mood monitoring as a potential intervention via RCTs (OXTEXT research studies 1-6). Several publications resulted from these studies, largely from the OXTEXT-1 cohort comprising up to 367 patients (≥16 years of age) from Oxfordshire, with a Diagnostic and Statistical Manual of Mental Disorders-IV diagnosis of BD (BDI, BDII, or BD-not otherwise specified), with some patients completing up to 81 months of continuous weekly mood measures. Compliance and acceptability of the True Colours system in the OXTEXT-1 cohort were excellent, with low attrition (<2%) and a median of less than 8% of weeks of missing data that did not differ by key sociodemographic factors or by mood score [10]. This pilot work has demonstrated support for the feasibility of True Colours as a remote mood monitoring system in patients with mood disorders [4,10], and it has lent important insights into detectable mood instability that differentiates clinical course [5] and other BD patient characteristics, including cognitive functioning [10,11].

The OXTEXT-2 study [12] assessed participant’s compliance with monitoring, their mental health resource use (including hospitalizations and face-to-face and phone contact with mental health staff), and service and medication costs, before and during their first 12 months of engagement with True Colours. Compliance with monitoring was high, with a median response completion rate of 92% for both Web-based and SMS symptom reporting and all patients continuing to report during the duration of the study. The introduction of True Colours was thought likely to reduce service costs, but this was not supported in OXTEXT-2. In fact, when associated with enhanced specialist care, medication costs increased over the first year of monitoring. This illustrated that studies of any digital addition to care need to account for all possible confounders relating to mood monitoring and mental health service costs. OXTEXT-2 did not examine nonmental health service costs, and larger economic evaluations of the True Colours system are required and are being conducted.

The True Colours system was also utilized as part of a psycho-education intervention for 121 patients with BD in an RCT (OXTEXT-6) [13]. The Facilitated Integrated Mood Management (FIMM) [14] study condition involved True Colours mood monitoring, a psycho-education manual, and individual sessions with a facilitator. This was compared with Manualized Integrated Mood Management, which only involved the psycho-education manual. Patients in the FIMM arm showed better knowledge of BD, and greater BD knowledge was associated with a high number of months in remission over 1-year follow-up [13]. Of note, True Colours in isolation is not intended as an intervention, but it may improve symptoms via insight into patients about their symptoms and closer, more accurate monitoring by clinicians, which will require further study.

The OXTEXT-7 study commenced in 2013 involving a trial rolled out to all 11 community mental health treatment services across Oxfordshire and Buckinghamshire titled as Feeling Well with True Colours (FWwTC). The goal of FWwTC was to offer patients a self-monitoring system that could allow care interventions to be tailored to the individual. Patients and clinicians create tailored symptom monitoring schedules on the basis of the type of symptom measure, frequency of prompts (weekly, daily, and several times a day), and reminder frequency. This study was a stepped-wedge, cluster randomized design. In this design, all services eventually implemented FWwTC, but the time at which they were trained to implement FWwTC was randomized to compare outcomes in treatment services before and after the introduction of FWwTC. The aim of this phase of OXTEXT was to apply True Colours to other patient populations (including those experiencing depression, anxiety, psychosis, alcohol and drug use, and BD) and test the feasibility and cost-effectiveness of such a tool in a larger scale secondary care setting. Experience from this trial is currently being synthesized, and it has proved heuristically useful [15], although uptake across clinical services was a challenge, illustrating the considerable barriers to innovation that persist in the National Health Service and other medical services.

Taking advantage of the developments in digital technology and ubiquity of mobile phones, the AMoSS study introduced a mobile phone app, Mood Zoom, to facilitate a higher frequency of symptom monitoring and included wearable devices as measures of objective symptoms. The Mood Zoom questionnaire comprises mood state descriptor items that are rated on a scale from 1 to 7 [18], which could be completed several times a day. Mood Zoom was used alongside weekly True Colours mood monitoring to help understand, in greater detail, mood episodes and mood instability in patients with BD and borderline personality disorder, as well as healthy volunteers in a sample of 139 patients with 3 months of continuous data (as per protocol) but with over 12 months of continuous data (for those willing to continue). The introduction of a mobile phone app also enabled the collection of passive background data, such as number of texts or calls and geolocation [19], which could reflect proxies of behavior associated with BD and how they are associated with mood, an emerging area with promise for the identification of behavioral markers of impending BD-related episodes [20]. Quantitative [18,21-23] studies have supported the feasibility and acceptability of the use of the Mood Zoom app and True Colours for daily and weekly symptom monitoring in patients with BD, borderline personality disorder, and controls. Specifically, attrition was low in the AMoSS cohort, with only 1 subject withdrawing and 8 subjects being excluded because of providing data for less than 2 months. Median adherence for the Mood Zoom and weekly measures was greater than 80% and 85%, respectively, and it remained stable over the study follow-up [18]. A qualitative study of 20 subjects from the AMoSS cohort provided support for the fact that reporting on symptoms once daily was of no inconvenience, and it was felt that the system contributed to insights into personal symptoms and patterns [24]. Additional themes from this study highlight the importance of tailoring patient preferences into symptom reporting tools.

In recent studies, additional objective physiological measures, derived from Fitbit and wrist-worn accelerometers, were included, along with daily and weekly mood monitoring [22] as well as the proteus patch [21,25] that provides an estimate of heart rate. These studies have contributed insights into detectable variability of sleep patterns in patients with BD and borderline personality disorder, which map onto observable symptoms of low and irritable mood [21] and variability in mood [25]. The additional add-on of wearables offers an exciting line of inquiry into objective symptoms of illness-alleviating biases relating to subjective reporting of symptoms. This potentially supports downstream applications of True Colours, with the inclusion of additional devices for the measurement of objective symptoms, which will be important for deeper insights into early signs of disease.

The BDRN [17] adopted the True Colours system, engaging 815 research participants (815/4080, 19.97% of invited existing BDRN participants) with mood disorders [26]. BDRN participants with a diagnosis of BDII were more likely to register with True Colours. Approximately 78.2% (637/815) of registered participants completed 3 months of symptom reporting, approximately 51.1% (413/808) of the participants completed more than 1 year, and some participants continued mood monitoring for up to 3 years, demonstrating the feasibility of such a remote mood monitoring system at a larger scale.

An international application of True Colours is from the Canadian Flourish High-risk Offspring Study [27], recruiting young offspring of a parent with BD. The Flourish group has piloted the Web-based True Colours monitoring system in 50 high-risk offspring of a bipolar parent and 108 control offspring of psychiatrically well parents. Compliance was good over 30 days, with approximately 80% and greater than 90% of high-risk and control offspring completing daily ratings, respectively, and no difference in compliance between study groups. Daily mood scores significantly differentiated the high-risk from control offspring, and irregularity in weekly mood and anxiety scores was higher in high-risk offspring with remitted major mood disorders compared with those with no lifetime history of major mood disorders [27].

Additional studies from the University of Oxford have made use of the True Colours system to elucidate mood variability in BD, involving determining the different nonlinear time series processes of mood instability and analytic techniques for appropriately detecting it from high-frequency time series data [5,28], as well as its associations with mental imagery [29].

RCTs of treatment efficacy in psychiatric disorders are expensive and lengthy, given the needed follow-up time for full Diagnostic and Statistical Manual of Mental Disorders threshold mood episodes to develop. Traditional endpoint assessments using paper and pencil questionnaires or clinician-rated diagnostic episodes also ignore clinically significant symptoms not meeting full diagnostic threshold between episodes [30] and cognitive dysfunction [31], which could be used to determine earlier and more proximal treatment effects. Several RCTs have used True Colours as both primary electronic outcome assessments and secondary higher frequency outcome measurements. For example, a 12-week double blind RCT (CEQUEL) [32] assessed combination therapy with quetiapine plus lamotrigine versus quetiapine monotherapy plus lamotrigine placebo on depressive symptoms in 266 patients (≥16 years) with BD, recruited across 27 different United Kingdom clinics. Another completed single blind RCT (OASIS) [33] of 3755 university students across the United Kingdom used True Colours to measure outcomes to determine the effectiveness of a Web-based cognitive behavioral therapy for insomnia and other psychiatric symptoms, including psychosis, mood, and anxiety.

Other mood-related applications of True Colours for outcome assessment in ongoing RCTs include the OxLith [34], aimed to compare lithium with placebo on mood instability in adult patients with BD; a trial assessing the clinical effectiveness and cost-effectiveness of lithium versus quetiapine augmentation for treatment-resistant depression [35]; and OxCaMS [36], which aims to assess the impact of a calcium channel blocker on cognition and brain activity in adults with mood instability. Finally, the Oxford Brain Body Research into Eating Disorders study [37] involves a pilot trial to assess the safety, acceptability, and feasibility of deep brain stimulation in patients diagnosed with severe eating disorders. Other funded large trials involving the True Colours system under development include the Pramipexole Therapy in Treatment Resistant Depression and Bipolar Depression (PAX-D and PAX-BD) [38,39].

As the research and clinical utility of True Colours became evident, it naturally branched out to other populations and research contexts. The Cognition and Mood Evolution across Time study is aimed at measuring cognition and brain activity in healthy participants with various levels of mood instability—a useful application of True Colours, with the inclusion of daily mood monitoring and cognitive tasks [40].

The True Colours system has also been modified for community outpatients, with a diagnosis of psychosis using forensic psychiatric services (FOXWEB risk violence tool). This research application involved the development of a Web-based violence risk monitoring tool for psychosis, which provides visual feedback of patient scores to clinicians to guide risk assessment [41], and this is being further piloted in inpatients.

The Brief Interventions for Self-Harm (BIRSH) clinic [42] has piloted True Colours for self-harm prevention in patients (13-65 years) presenting to accident and emergency departments. The aim of this ongoing research and service evaluation application was to determine the effectiveness of a new clinical service incorporating remote symptom monitoring to reduce self-harm repetition and health service costs.

The True Colours inflammatory bowel disease (IBD) group has expanded the True Colours schedule to include daily measures of ulcerative colitis and Crohn’s disease symptoms, as well as fortnightly quality-of-life and other validated measures of disease activity. The initial aim of the True Colours IBD project was to develop and test the feasibility of a predictive index of IBD. A 6-month pilot in 66 patients supported the initial feasibility of this system, with 76% adherence rate for daily measures and 86% patient retention [43]. Further work has supported associations between daily IBD symptom measures and biological measures of disease activity [44], and this has facilitated the prediction of whether escalation of therapy or clinical investigation would be needed [45]. Qualitative findings from this work suggest that patients felt more in control and empowered by the True Colours IBD system [43].