Background
In bipolar disorder research, there has during the last decade been an emerging shift in illness paradigm from a focus on affective episodes to an increasing focus on inter-episodic mood instability [
1,
2]. Many patients with bipolar disorder remain symptomatic during inter-episode periods and experience significant subsyndromal day-to-day or week-to-week mood swings that are of greater severity than those experienced by healthy individuals and appear to reflect illness activity [
2]. Further, these subsyndromal mood swings seem associated with high risk of relapse, hospitalization and impaired functioning [
1,
3‐
5]. Continuous monitoring and assessment of mood instability and other variables possibly reflecting illness activity in detail, including measures of duration, severity and frequency of symptoms, may therefore be clinically relevant since it would allow for early intervention on subsyndromal symptoms and ultimately prevention of full-blown affective episodes. Self-reports are ubiquitous in psychiatric research, and various mood charting instruments for self-monitoring are frequently used in the management and monitoring of depressive and manic symptoms in patients with bipolar disorder. Traditionally these mood charting instruments have been paper-based, such as the National Institute of Mental Health LifeChart Method (NIMH-LCM) [
6], the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BP) the Mood Chart (mood chart no longer available online) and the ChronoSheet [
7] and have been shown valid compared to clinical rating scales for depression and mania [
8,
9]. Paper-based mood charting instruments can be viewed as facilitating tools helping patients with bipolar disorder gain illness insight, facilitate patient empowerment, teach patients to recognize early warning signs of recurrence of affective episodes and enable individualized characterization of mood instability in detail. However, several issues limiting the usefulness of paper-based mood charting instruments have been addressed, such as low compliance and potential recall bias when reporting data retrospectively, i.e. where patients complete batches of daily ratings at a single time (sometimes referred to as
hoarding or
backfilling) [
10‐
13]. During recent years there has been an increasing growth of e-mental health technologies [
14], including electronic platforms offering tools for self-monitoring of mood. The electronic approach for self-monitoring of mood offers ecological momentary assessments [
15], a monitoring technique for assessment in real-time and in naturalistic settings, offers the ability to verify the timing and compliance of data collection, eliminates the need for costly and error-prone data entry, may help remind patients to perform the self-monitoring and may have higher usability than paper-based versions.
However, it remains unclear whether the severity of self-monitored mood registered using electronic self-monitoring tools is a valid measure compared to validated clinical rating scales for depression and mania, which are currently used as the golden standard for assessing the severity of depressive and manic symptoms in patients with bipolar disorder. Furthermore, it remains unclear to what extent the use of electronic mood self-monitoring tools affects clinically relevant outcomes, and importantly whether there may in fact, be harmful effects, e.g. self-monitoring of mood symptoms may induce depressive ruminations that may result in increasing severity of depressive symptoms [
16]. An understanding and overview of these aspects is crucial in order to guide the use and development of IT platforms for electronic self-monitoring of mood in bipolar disorder.
The objectives of the present systematic review were thus 1) to evaluate the validity of electronic mood self-monitoring tools compared to validated clinical rating scales for depression and mania and 2) to evaluate the evidence of the effect of electronic mood self-monitoring tools on clinically relevant outcomes in randomized controlled trials (RCT).
This is the first systematic review of electronic self-monitoring of mood in patients with bipolar disorder.
Discussion
This is the first systematic review of the evidence of the validity of electronic mood self-monitoring tools using IT platforms as methods for assessing mood in adult patients with bipolar disorder compared with clinical rating scales for depression and mania. Further, the evidence of the effect of electronic mood self-monitoring tools on clinically relevant outcomes in RCTs was assessed. A total of 13 published articles were included. The included articles were heterogeneous, employing various monitoring IT platforms and included different clinically relevant outcomes.
Electronic self-monitored mood was found valid compared to clinical rating scales for depression in six out of six studies comprising a total of 179 outpatients [
12,
54,
56,
58,
59,
64], but only two studies found a correlation between electronic self-monitored mood and a validated rating scale for mania [
55,
64]. Thus, the present review suggests that despite different bias issues it seems possible for patients to validly evaluate the severity of their depressive symptoms, but specifically difficult to report emerging manic symptoms in a valid way and may be due to decreased illness insight during hypomania/mania [
70]. Furthermore, as mentioned most of the studies included only outpatients who did not present with severe levels/ high scores on clinical rating scales for depression and mania. The validity of the electronic self-monitoring tools may be both overestimated as well as underestimated disregarding the possible difficulty of self-monitoring the severity of symptoms in more severe cases.
The studies included describe convergent validity of electronic self-monitoring tools. Other variables such as activity level and sleep length could represent parameters for self-monitoring that may correlate with clinical rating scales for depression and mania, but content validity was not investigated in the present review. Further, the reliability and predictive validity of the electronic self-monitoring tools were not investigated in the present review.
Paper-based self-monitoring tools for depressive mood registered using different types of paper-based tools has been shown to significantly correlate with the scores on clinical rating scales for depression in a number of studies [
7,
9,
71‐
73]. The severity of self-monitored manic mood has been shown to correlate with the scores of clinical rating scales for mania in a number of studies [
8,
9,
74,
75], however, not in all studies [
7]. Further, other parameters such as activity level may correlate with the scores of clinical rating scales for mania [
7]. Two of the RCTs included researcher blinded validated clinical rating scales for depression and mania as outcomes [
16,
62]. These studies could potentially investigate the validity of electronic mood self-monitored compared to clinical rating scales for depression and mania and thus further contribute to the knowledge in this area.
It should be mentioned that a paper by the authors analyzing the validity between mood self-monitoring and clinical rating scales for depression and mania has been accepted for publication [
76]. This study used a smartphone-based mood self-monitoring tool and found a significant correlation between self-monitored mood and validated ratings scales for both depression and mania, respectively.
Using electronic self-monitoring tools may offer solutions regarding issues of low compliance and potential recall bias that are present when using paper-based self-monitoring tools. However, the results presented in this review suggest that it seems difficult for patients to evaluate manic symptoms calling for other new and more objective real-time electronic methods to monitoring the severity of manic symptoms in patients with bipolar disorder.
To provide a more complete and inclusive picture of the scientific research on electronic self-monitoring the RCTs using electronic self-monitoring tools as a part of an intervention were evaluated. No study investigated the sole effect of electronic self-monitoring of mood as an intervention in itself, but investigated the effect of different electronically delivered intervention programmes with electronic self-monitoring of mood represented as part of the intervention. The evidence of the effect of electronic self-monitoring was limited by methodological issues and by a lack of RCTs. Notably, three of the RCTs investigating the effect of electronically delivered interventions did not report on any researcher blinded outcomes [
60,
61,
63] and thus introduces bias issues on the validity of the results from these studies and introducing the risk of overestimating or underestimating the beneficial and harmful effects of the interventions. One RCT [
16], by the authors, reported on potentially harmful effects of electronic self-monitoring with more depressive symptoms but fewer manic symptoms in the intervention group. A paper by
Scott & Colom discuss the issues of a differential effectiveness of psychological interventions for manic and depressive phases [
77], and points out that the reasons for these differential effects are not clear. Manic prodromes are more distinct and may be easier to detect and treat more quickly and effectively with pharmacotherapy than depressive episodes [
78]. On the contrary, depressive symptoms are more difficult to differentiate from normal day-to-day problems and may have a more gradual onset and prolonged duration [
79]. Considering electronic self-monitoring a psychological intervention, the potential harmful effects on depressive symptoms as suggested by the findings from the RCT by the authors [
16] highlight that electronic self-monitoring should not be uncritically used or implemented in clinical practice and that important aspects need further clarification before it is implemented as a standard tool.
If there would be an effect of electronic self-monitoring on the severity of depressive or manic symptoms, then self-monitoring would influence the variables it measures (mood). Whether that would be a threat to the reliability and validity is unknown and should be investigated further.
In westerns countries nearly everyone has at least one device that would be able to handle electronic self-monitoring. The use of computers and/or tablets as tools for electronic self-monitoring of mood require some technical skills by the user and can be quite expensive to acquire. However, most people have access to a computer or tablet and know how to interact with simple software systems. Computers and tablets allow for storage and visual presentation of self-monitored data making recognition of symptom patterns possible, thus potentially providing tools for increasing the patients’ illness insight and empowerment. PDAs represent a tool that is possible for the patients to carry with them during the day making continuous real-time electronic self-monitoring in naturalistic settings is possible. However, when using a separate and non-standard electronic device for self-monitoring of illness activity the risk of stigmatization is present [
56]. Other electronic devices could replace PDAs as an electronic self-monitoring tool in the years to come. Unlike computers and PDAs, smartphones offer opportunities for continuous electronic self-monitoring in naturalistic environments that cannot be achieved using other types of IT platforms. Since most people carry their cell phone with them during most of the day and use it for communicative purposes through various platforms the risk of stigmatization is not present. Furthermore, the number of smartphone users worldwide has been estimated to reach 2.5 billion people by 2017 [
80] which makes smartphones an obvious tool for electronic self-monitoring.
Patients willing to participate in studies using these kinds of novel and technical electronic interventions could represent a more motivated and technically oriented group of patients with higher degree of illness insight and willingness to use the electronic self-monitoring tools in question. As can be seen from some of the included articles, compliance to self-monitoring was highly variable, and patients participating were quite young. Elucidation of possible technical barriers for using an electronic device for self-monitoring among non-technically oriented patient groups, perhaps in patients with higher age than in the included studies, could be of interest. In addition, future studies should provide more information on excluded patients and reasons for declining to participate in studies in order to allow the readers to better assess the level of generalizability of the study results.
None of the included studies provided information regarding the economical part of the development and maintenance of the electronic self-monitoring IT platforms, and is likely a relevant factor in the development of future efforts in this area as well as for the clinical implementation.
Limitations
Some limitations to the present review should be mentioned. Telemedicine in general and e-mental health are areas under great expansion and the investigations in this area are published in very diverse forms and places reflecting that this is an research area in the intersection between two areas of research- medical research and IT research. Therefore, conducting a search strategy that is able to capture all relevant scientific literature is a challenge. A search on Google Scholar alone on electronic self-monitoring in bipolar disorder resulted in 595.000 hits. Many commercial websites and smartphone applications (for both Android, iOS and Windows) for electronic self-monitoring exist in the App store and the Google Play store. The search strategy for the present review reflects that we aimed at systematically collecting and reviewing published scientific studies in order to get an overview of the validity of electronic mood self-monitoring compared to validated clinical rating scales and the evidence status related to using these kinds of electronic self-monitoring tools.
Most of the included studies had a daily and momentary frequency of self-monitoring, but one study used a frequency of self-monitoring on a weekly basis. A recent paper discussed differences in momentary and retrospective trait self-report techniques pointing out that retrospective self-monitoring is influenced by peak moments with greater salience of moments that occur closets in time to the assessment [
13].
The interventions of the included RCTs were heterogeneous and often mood self-monitoring was a part of an intervention, which incorporated mood self-monitoring alongside other psychological interventions. In addition, the RCTS employed various monitoring IT platforms and included different clinically relevant outcomes. Thus, comparing not only different mood self-monitoring tools but also different interventions as well as outcomes is a big challenge.
It should be noted that the authors did not have access to the various electronic self-monitoring tools reviewed apart from one of the tools for smartphones [
16,
59]. Further, as pointed out by others [
44] all of the self-monitoring tools are different from one another allowing for calculation of many different measures of illness activity and also making it difficult to compare findings across studies.
One author (MFJ) selected all papers and extracted all data, and one of the co-authors (KM) independently checked the extracted data.
It should also be noted that in most of the articles the patients had to provide the hardware for the electronic self-monitoring themselves and no information regarding economical compensation was given. None of the included articles provided information regarding the cost of developing and maintaining the electronic software and the amount of possible technical problems with the electronic self-monitoring systems. These aspects are likely relevant factors that should be evaluated in future studies.
It would be interesting to investigate the validity between electronic mood self-monitored and validated clinical rating scales for depression and mania and the effect of electronic self-monitoring as an intervention in a non-technically oriented group of patients with bipolar disorder, and further to elucidate these aspects during full-blown affective episodes.
Competing interests
MFJ has been a consultant for Eli Lilly and Lundbeck. KM has no conflicts of interest. LVK has within the last 3 years been a consultant for Lundbeck and AstraZeneca. MFJ, LVK, MF and JB are the main researchers and authors on the MONARCA system.
Authors’ contributions
MFJ, KM and LVK designed the study and analyzed the extracted data. MFJ, KM, MF, JB and LVK wrote the paper. All authors contributed to and have approved the final version of the paper.