Use of mobile apps and technologies in child and adolescent mental health: a systematic review

Child and adolescent mental health

Mental health (MH) and well-being is rapidly growing as the new frontier in child and adolescent health. The increased interest is justified by the publication of troubling statistics showing decreasing levels of happiness in children1 and increasing recognition of their MH problems.

One in 10 children and young people (CYP) are estimated to present with a diagnosable MH condition by conservative estimates2; however, some studies have demonstrated a prevalence as high as 19%.3 As an example, data from the UK show that despite expansion of the teams and increased staffing levels,4 Child and Adolescent Mental Health (CAMH) teams are experiencing continuous surge in referral numbers (figure 1).

It is difficult to definitively say if this corresponds with a true increase in the prevalence of MH issues as this might be a reflection of greater societal awareness and reducing stigma around the topic. Equipping young people with knowledge and providing coping strategies is now a core component of education curriculums,5 and this is supplemented by easily accessible information and discussion forums on internet and media platforms which allow them become more active and seeking help when necessary.

Unfortunately, the increased numbers of CYP requiring CAMHS intervention is placing the service under untenable pressure. Taking Scotland as an example, one in five children referred to CAMHS breach the 18-week maximum waiting time goal set by the Scottish Government,6 and despite assurances that spending on MH would be increased as a priority, MH expenditure has gone down when compared with overall National Health System budgets.7 This means that the service has to find ways to adapt in order to continue providing high-quality care. The economic cost of poor MH in Scotland is currently estimated at around £10.7 billion,8 and this will only rise if nothing is done.

Mobile applications (‘apps’) and their utilisation in child and adolescent mental health

Arguably, the most significant and impactful change to society in the last 50 years has been the digital revolution, the introduction of digital technology and the internet to all aspects of life; this includes interest in using them in the service of health (widely accepted applications already in use are provided in online supplementary file 1: Digital Health Tool. Source: IQVIA Institute; The Growing value of Digital Health; 2017).

With the advent of smartphones, mobile telephones capable of computer functionality, digital health-related content is in everyone’s pocket: 64.3% of the UK population owns a smartphone, with an even larger proportion regularly going online,9 so it is not surprising that the market of downloadable applications or ‘apps’ has also embraced digital health content. More than 318 000 health apps (h.apps) are currently available through multiple platforms (App Store, Google Play) with more than 200 new health-related apps being added everyday.10 However, their market penetration is on the whole poor, with only 15% reaching more than 5000 downloads. Despite this, it was estimated that by the end of 2017, 50% of mobile phone users have downloaded at least one health app11 (online supplementary file 2: The number of Digital Health Apps 2013, 2015 and 2017. Source: IQVIA Institute; The Growing Value of Digital Health; 2017). A recent Australian study12 about adolescent’s perceptions of online therapies for m.health problems showed that 72.0% (n=217) said they would access an online therapy if they experienced an m.health problem and 31.9% would choose an online therapy over traditional face-to-face support. The most valued benefits reported were alleviation of stigma and increased accessibility, and the knowledge of online therapies was found to positively predict perceived helpfulness and intended uptake. This suggests that h.apps are an acceptable resource and can be a useful vehicle for enhancing access to evidence-based monitoring and self-help.

There are significant obstacles to overcome before this goal can be achieved, the greatest of which concerns the oversight and regulation of h.apps. Apps developed in academia and supported in clinical trials are slow to reach the consumer marketplace; meanwhile, proliferation of industry-developed apps on consumer marketplaces has been high.13 Anyone can create and upload an app for public use but there is, at present, little to no regulation or certification.14 This is one of the greatest barriers to physicians recommending apps to their patients; an American study found that 42% of physicians would not feel comfortable or safe prescribing use of apps without regulatory oversight,15 the risks of doing so, is succinctly put by Boudreaux et al.14 ‘the decision to recommend an app to a patient can have serious consequences if its content is inaccurate or if the app is ineffective or even harmful’.14 16 It is also of note that high-quality guidance on how to judge the validity of commercially available apps is severely lacking,14 17 with guidance limited to personal familiarity with apps or user ratings. Some tentative efforts have been made to address these problems. A UK Digital Health Apps Library18 was launched in April 2017, and although still in its beta phase of testing, the website aims to provide information on which apps are accessible and safe to use, with apps listed as NHS approved or under testing.

In order to remain successful and relevant, apps creators must also embrace the constantly evolving technology and operating systems; if they do not, apps can quickly become obsolete. It is therefore crucial that h.apps are constantly assessed and adapted to reflect advances in the health sector and the business/consumer sector.14 Another important consideration is the language: 56% of the apps are available only in English,14 making its use in an increasing multicultural society more difficult.

Approximately 60% of marketed h.apps come under the category of ‘Wellness Management’ and are focused on health maintenance, facilitate tracking and modification of personal lifestyle behaviours. The remaining are categorised as ‘Health Condition Management Apps’, some of which are disease specific, the most popular being for MH and behavioural disorders (eg, autism and alternative communication, depression, anxiety and attention deficit and hyperactivity disorder).14 19

The current CYP are ubiquitous consumers of technology; indeed, this generation has been labelled as ‘iKids’20 and ‘digital natives’,21 and children of all ages are spending ever increasing amounts of time online.22 Around 69% of children use a mobile phone and by the age of 12, >80% of young people accessing the internet via smartphones download and use mobile apps.23

There is also well-founded concern as to the effect of mobile technology usage on CYP. High-frequency use of mobile phones has been related to sleep disturbance and depression24 and less accurate response to higher level cognitive tasks,25 and more Facebook friends has been found to be predictive of personality disorders.23 Despite these worries, it seems unlikely that CYP will be prised away from their mobile phones. In light of this, apps seem to be an ideal avenue to explore in the realm of CAMH maintenance and management (online supplementary file 3: Disease-Specific Apps by Therapy Area. Source: IQVIA Institute; The Growing Value of Digital Health; 2017).

Study selection and analysis

A literature search of six electronic databases was conducted for relevant publications between January 2007 and May 2019. Databases searched included MEDLINE, PsychINFO, EMBASE, Allied and Complementary Medicine, Health Technology Assessment Guide and the Cochrane Register of Controlled Trials. Searches were conducted with keywords pertaining to MH, well-being and problems, age of the study population and words pertaining to mobile or internet apps or interventions (online supplementary file 4: Table with Search Terms Used).

Database-specific filters of human population and English language were applied; following the database search, reference lists were screened to ensure an exhaustive search. The resulting titles and abstracts were manually screened according to the inclusion and exclusion criteria detailed in online supplementary file 5: Table with Inclusion and Exclusion Criteria.

Online supplementary file 6: PRISMA Flow Diagram details the manual screening process of the search results against the inclusion and exclusion criteria. It should be noted that if at any stage there was equivocation around the suitability of papers for inclusion, they were advanced to the next stage of the process for more rigorous scrutiny.

Study inclusion

Initial database searching with the filters of human population, English Language and 2007–2019 publication date identified 1013 results with three additional results identified through screening of appropriate reference lists. Duplicate results were removed yielding a total of 692 records for screening. The titles and abstracts of these unique citations were assessed for eligibility. After this process, the remaining 92 full-text articles were assessed against the inclusion and exclusion criteria with a total of 6 articles included in the final review. In these six remaining articles, three of them reported about the same randomised clinical trial (ClinicalTrials.gov NCT00794222), and to avoid confusion, they were tabulated together since the research design and intervention are identical.

Study characteristics

Study characteristics are detailed in table 1.

Study risk of bias

All studies were assessed using a tool adapted from the Cochrane Risk Of Bias Tool (http://handbook-5-1.cochrane.org/). In general, performance bias due to difficulties in blinding participants and personnel was an issue, as well as difficulty in assessing the use of selective reporting. Full results of the risk of bias assessment can be seen in table 2.

Study participants

The studies included participants whose age ranged from 10 to 29 years. Only two studies sampled exclusively participants between the ages of 13 and 17 years.26 27 In the Jang et al.28 study, 81.4 %  of participants were aged 10 to 29 years but no further breakdown of the groups’ composition in terms of ages was provided.28 The sum of mean age of participants in the mobile-type trial was 18.1 years.29–31 Additional demographic data of participants were recorded in all studies, with none reporting any statistically significant differences between the comparison groups. However, in all of them, the number of females was disproportionately greater than the number of males.26 28–31

Two of the studies exclusively comprised data from participants with mild to severe emotional or MH difficulties, either formally diagnosed26 or indicated by scores on a validated scale.29–31 As the aim of the MEMO trial27 was to evaluate the effectiveness of the intervention in preventing the onset of depression, prospective participants scoring in the clinical range as having MH difficulties on several validated scales were excluded.

Incentives to participation were employed in the design of some studies. The mobile-type studies offered $30 phone credit to reimburse the young participants, and participating general practitioners received quality assurance points towards their continuing personal development,29–31 while the MEMO trial offered ringtones and games.27

Mobile technology characteristics and mobile phone ownership

Three studies used asynchronous features of mobile phones (apps and multimedia messaging);26 28–31 the remaining study concerned the use of a mobile phone-based interactive voice response (IVR) system.26 Studies evaluated the effect of using the mobile-type app,29–32 the MEMO-CBT multimedia mobile phone programme,27 an unnamed depression and suicide screening app28 and the IVR follow-up system for reporting mood post discharge from inpatient care.26 With the exception of the unnamed app described by Jang et al,28 all technologies used were designed specifically for CYP.26 27 29–31 At the time of writing, none of them is available for public use.

Mobile phones were required for participation in all included studies, some of which relied on participant’s personal mobile phone27 28 and others worked in conjunction with mobile technology companies to provide participants with mobile phones and pre-paid SIM cards.29–31 The MEMO trial research team provided free internet access.27

Mental health and healthcare service outcomes

All studies reported MH outcome data.26–28 30 31 As evident in table 1, the main areas of intervention were depression, anxiety and suicide. One study also reported health service outcomes data31 (table 1).

Mobile technology usage

Few of the included studies reported data on the technology usage. Whittaker found that 19% of participants viewed more than half of the multimedia messages and reported that participants greatly overestimated the proportion of messages they viewed.27 The mobile-type trial reported that only 52.6% of participants in the intervention group received what the researchers considered to be the minimum effective dose with the remainder of participants not active enough on the app to reach this dose level.30 Johansson reported 71% of participants achieved an answering IVR call frequency of 100%.26

Child and adolescent feedback

Although not the aim or primary outcome measure of any of the included studies, results of the MEMO-CBT included reports from the majority participants that they found the programme to be helpful, with a slightly higher proportion (84%) of the intervention group giving positive reports than the control group (82%) but this difference was nonsignificant (p=0.46). There were also nonsignificant differences between the intervention and control groups in sharing of messages with others (37.9% and 40.0% shared messages, respectively).

Limitations

There are several limitations to this review. Owing to time constraints, only the results of ‘full’ studies such as RCT and cross-sectional studies were included; ideally, results of all available literature would have been included.

In addition, confining the search to articles available in the English language may have limited the power of the review, particularly in the area of developing countries, applicability to low-income countries being low. The small number of studies and small sample sizes also limit the power of the review, as does the varying quality and bias of the papers reviewed.

Also four out of six of the studies included participants over the age of 18 years and no study included any participants under the age of 10; therefore, they can only be applied to adolescents and young adults and not to the ‘child’ population. The focus of the majority of the included work on depression and anxiety disorders restricts broad application to all areas of MH.

It must also be noted that the evolving market and pace of technology development may mean that the literature included is already outdated and obsolete,13 and without consulting experts on the relevance of the technology described it is impossible to know the value of the studies.

These limitations make the conclusions drawn tentative at best.

Recommendations for future research

As there is a distinct lack of research in the area in less developed countries, attention should be given to the use of these interventions in those countrieswhere high penetration of m.tech has already been described[55].

The ethical issues involved do require further examination, not only confidentiality and clinical risk concerns, but also the potential to medicalise normal MH and put clinicians under increased pressure to keep up-to-date with all patient self-monitoring.43

Currently, the interest for most researchers lies in the development and testing of apps they themselves have had a hand in developing. In the future, it may be of more benefit that existing technologies and apps are evaluated by diverse teams of researchers in order to properly evaluate them and come close to an ‘ideal app’.

As an Australian study has already demonstrated, scope to understand adolescents" use and diversity of preferences should also be assessed to better inform planned co-design processes in creating digital MH tools for CYP.44

In addition, continued efforts must be made to make clinicians aware of the available apps, and an agreed and validated system for assessing their quality must be developed.