Problems and Barriers related to the Use of Digital Health Applications: A scoping review

Background: The digitization of healthcare has also led to a steady increase in the adoption and use of mHealth apps. With the aim of fully realizing the benefits of mHealth apps, Germany is the first country in the world to cover the costs of mHealth apps through statutory health insurance. While numerous opportunities are known about how mHealth apps can improve patient health and make healthcare more efficient, not as much is known about the problems and barriers that stand in the way of optimal usage. Objective: This scoping review aims to map and categorize the evidence on problems and barriers related to the use of mHealth apps. Methods: Systematic searches were conducted in MEDLINE, EMBASE, and PsycINFO. Additional searches were conducted on JMIR and on websites of relevant international organizations. Inclusion criteria were: Publications that (1) dealing with apps corresponding to those approved in the German healthcare system, (2) addressing problems and barriers related to the usage of mHealth apps and (3) were published between January 1, 2015 and June 8, 2021. Study selection was performed by two reviewers. The manuscript was drafted according to the PRISMA extension for scoping reviews (PRISMA-ScR). The analysis of the included publications and the categorization of problems and hurdles were performed using MAXQDA. Results: The search identified 1479 publications. Thereof, 21 met inclusion criteria. Further eight publications were included from the focused search. Identified publications were analyzed for problems and barriers studied, which were classified into ten categories ('validity', 'usability', 'technology', 'use


Table of Contents
No, I do not wish to publish my submitted manuscript as a preprint.
2) If accepted for publication in a JMIR journal, would you like the PDF to be visible to the public?
Yes, please make my accepted manuscript PDF available to anyone at any time (Recommended).
Yes, but please make my accepted manuscript PDF available only to logged-in users; I understand that the title and abstract will remain v Yes, but only make the title and abstract visible (see Important note, above).I understand that if I later pay to participate in <a href="http

Introduction
Since the development of the iPhone in 2007, the proliferation of apps has steadily increased.Especially mobile health (mHealth) solutions such as web apps or native apps are more and more diffused and provide manifold approaches to support users' health.They can be applied in health monitoring and surveillance, for health promotion and raising awareness, communication and reporting, data collection, telemedicine, emergency medical care, point of care support, and decision support [1].
With the aim to benefit from the potential of new technologies such as mHealth apps for health care, the Digital Healthcare Act was introduced in Germany in December 2019.Hereby particular mHealth apps with a low risk class (I or IIa according to the Medical Device Regulation (MDR) or, within the scope of the transitional provisions, the Medical Device Directive (MDD)), so called "Digital Health Applications (DiGA)", became part of the German health care system [2,3].During the corresponding approval process, the "Fast-Track Process for DiGA", mHealth apps have to fulfill a predefined set of criteria.Among other things, these aim to prevent safety issues, problems with data privacy and security and to guarantee benefits either in form of medical benefit or patientrelevant structure and process improvements for the patient [4].mHealth apps that meet these requirements can be included in the so-called DiGA directory.Apps listed in this directory are reimbursable by the statutory health insurers.Currently, about 30 DiGA are listed and are subsequently reimbursable.While the German Fast-Track Process for DiGA currently is unique in the world, it has been announced that it will also be applied in France [5].
Many publications address possible benefits of mHealth apps.For example, mHealth apps for behavior change (either as a standalone intervention or as part of a larger intervention) have been shown to positively impact health outcomes compared to standard care and can be a useful adjunct in behavior change health interventions [6].In addition, Liu et al. examined the effectiveness of mHealth apps for assisted self-care interventions in patients with type 2 diabetes and/or hypertension and found that they were effective in improving blood glucose levels and blood pressure control [7].Wang et al. systematically reviewed the effectiveness of mHealth apps for monitoring and managing mental health symptoms or disorders and found that they have the potential to monitor or improve symptoms of certain mental health disorders, such as anxiety, stress, alcohol disorder, sleep disorder, depression, suicidal behaviors, and post-traumatic stress disorders [8].Finally, rising DiGA prescription numbers and strong interest from physicians and psychotherapists indicate that DiGA are expected to have a potential to improve care and, in some cases, fill existing gaps in care [9].Nevertheless, as in other sectors and areas of health care, problems and barriers might arise in the context of mHealth apps as well.Therefore, an integrated application of mHealth by health care systems requires a comprehensive analysis of problems and barriers to address potential challenges and risks in advance adequately.To our knowledge, problems and barriers related to mHealth, have not been gathered systematically, yet.Such a compilation would be the precondition to analysis weather certain problems are in need for further governance and regulation during the processes of development, approval, dissemination or utilization.Therefore, this research paper aims to search literature systematically, that identified problems and barriers related to the use of those mHealth apps similar to DiGA.Problems and barriers identified will be compiled and categorized.

Methods
A scoping review was conducted to identify problems and barriers related to the use of mHealth apps.The research was guided by the five mandatory stages for scoping reviews proposed by Arksey and O'Malley [10] which were further developed by Levac et al. [11].The manuscript was drafted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for scoping reviews (PRISMA-ScR) [12].The corresponding research protocol was published in JMIR Research Protocols [13].

Search strategy
A systematic search for articles published between January 1 st , 2015 to June 8, 2021 was conducted in bibliographic databases (MEDLINE, EMBASE and PsycINFO).The search strategies were developed through discussion (GG & CS) and with aid of an experienced researcher (SN).
In order to develop a suitable search string for the systematic search in MEDLINE, EMBASE and PsycINFO, the MIP scheme including methodology (all methodologies), issues (problems and barriers related to mHealth apps), and participants (main focus on patients and health care providers) was adapted [14].Subsequently, the search terms and links among them were defined.Searches for defined terms were restricted to the occurrence in abstract, title and keywords.If there were indexing terms (MeSH, Emtree) the search string was extended accordingly.The final search strategy for each database can be found in the research protocol [13].
Results were loaded into EndNote reference management program (version X9).To supplement additional evidence JMIR was searched on January 18 th and 19 th of 2022 and reference lists of included studies were investigated on eligible articles.The search in JMIR was performed by using the search function on the journal's website.For this purpose, the problem terms were combined with either the term 'mHealth app' or 'mobile app'.This adjustment was made due to a consensus paper recommended by the editor [15].Apart from the bibliographic databases and reference lists, gray literature sources such as reports, guidelines and working papers were searched via institutional websites.A full list of considered institutions can be found in the corresponding research protocol [13].
Search for gray literature was conducted depending on the institutional website in question.If available, search fields were used to identify publications by using search words related to mHealth apps.Otherwise, relevant subpages with reference to the topic of mHealth apps were searched.

Eligibility criteria
Inclusion criteria were the focus on problems and barriers related to the use of mHealth apps which were similar to the German DiGA concept.Journal papers were included if they were peer-reviewed, published in 2015 or afterwards and were written in English, German or French.Papers were included irrespective of their research method.See Textbox 2 or the research protocol [13] for detailed inclusion criteria.The criteria that had to be met for mHealth apps to be classified as DiGAlike can be derived from the exclusion criteria in Textbox 3. Reviews and app assessments on mHealth app categories, which in principle could also be implemented as DiGA or have already been implemented were also assessed as DiGA-like and consequently included as well.Textbox 2. Inclusion criteria  Articles mentioning problems and barriers related to the use of mHealth apps  A problem term mentioned in the abstract or title relates to the use of mHealth apps  Publication with focus on mHealth app  Included mHealth apps were similar to DiGA  Article published in 2015 or afterwards  Language: English, German or French Exclusion criteria were not providing answer to the research question or not having at least one of the predefined problem terms related to the investigated app ('difficulty', 'obstacle', 'problem', 'issue', 'challenge', 'barrier') in the title or abstract.Further articles were excluded if the investigated mHealth apps were not similar to DiGA (not for patient use, no relation to illness, injury or handicap, for primary prevention or not achieving its medical purpose through the main digital functions), the publication date was before 2015 or the language was other than English, German or French.Furthermore, research protocols and conference abstracts were excluded.The exclusion criteria are presented in Textbox 3 or more detailed in the research protocol [13].

Article Screening and Data Extraction
After downloading citations and transferring them into EndNote, duplicates were removed.Screening and selection were performed in two steps.In a first step, two reviewers (GG & CS) assessed titles and abstracts independently.In a second step, articles included for full text screening were assessed using the exclusion criteria by the same reviewers independently (cf.table 2).In case of disagreement, conflicts were resolved by a third person (SN).The two reviewers (GG & CS) used MAXQDA to independently mark and extract relevant text characteristics of included articles.A previously developed data-charting form was used for extraction.Extracted data consists of metadata, such as article characteristics as well as information related to the underlying research question -problems and barriers related to the use of mHealth apps.Thus, relevant items were author, year, study country, study participants, type of study, underlying diseases as well as problems and barriers related to the use of mHealth apps.

Synthesis of results
After the evaluation of included studies, the results were summarized in a descriptive way.Identified problems and barriers were grouped into clusters.Whenever a problem or barrier arose that could not be sorted into an already existing cluster, a new cluster was created.Finally, the respective clusters were appropriately named according to the problems and barriers they contained.Additionally, the results were summarized, systemized and presented in tables (cf.appendix 1, appendix 2).

Selection of sources of evidence
The systematic search yielded 1,479 articles after removing duplicates (cf.Fig. 1).Of these, 72 studies were screened in full-text (cf.Appendix 3) and subsequently, 21 studies met the inclusion criteria .Furthermore, three studies were identified by screening the references of included studies [37-39] and five studies were identified by the search in JMIR [40][41][42][43][44].The search on institutional websites did not yield any further results.In total, 29 studies were included into this scoping review.

Identification of studies via databases Identification of studies via other methods
Records removed before screening:

Synthesis of results
Problems and barriers identified in included studies were categorized into ten major groups.Included studies usually addressed several different problems and barriers.Appendix 2 gives an overview of which categories of problems and barriers were found in which articles.The 10 groups included 'validity', 'usability', 'technology', 'use and adherence', 'data privacy and data security', 'patientphysician relationship', 'knowledge and skills', 'individuality', 'implementation', 'costs' (cf. Figure 2).For more detail see below.

Validity
Problems with validity were addressed in 16 articles [16, 18, 21, 23-30, 33, 35, 39, 40, 44].Thereof, three described quantitative studies [24,27,35], four qualitative studies [16,18,21,39], six were mixed-design [25,26,30,33,40,44] and three were reviews [23,28,29].Problems addressed in these studies were mainly in the area of contents, outcomes and user input.Problems in validity concerning contents and outcomes of some mHealth apps were due to a lack of accordance with clinical standards.A fundamental problem was especially found in missing empirical evidence [24,27,29].Some content was declared as inappropriate [18], wrong [24] or ambiguous [24,33,40].Patients as well as health care staff stated that depth and quality of information was often not suitable [21,39].One quantitative study found that few apps provided details about the underlying formulae used for calculations [24].In two studies, some users criticized that functions did not meet their needs [35,40].Lack of added value was perceived or assumed in qualitative and quantitative studies [21,25,26,27,44].Some studies described mHealth solutions as inferior to usual care [21,23,29,30] or mentioned adverse effects or even harm [16,21,27,28,39].One mixed-design study found that physiotherapists were skeptical about whether a hybrid setting including an mHealth app could be conducive to build and maintain a robust working alliance between patients and physiotherapists [25].
Besides problems with app content, an app assessment study found problems with user inputs and their validation [24].In another study, patients described problems with changing entered values [39].Only few apps provided guidance based on user-entered data.This was especially important in mental health crises or risk of suicide [27].Some apps caused inappropriate alerts after incorrect data entry in the settings component [44].Patients manipulated the generated results by deliberately entering wrong values in order to receive better feedback from the app [39].Thereby, the medical benefit of the app could be reduced.Apps including physical exercises faced the problem of validation, too.Lack of feedback on correctness of exercise executions led to a feeling of insecurity and wrong execution of exercises [25].
Health care professionals described app measurements, calculations and resulting data provided to the user as imprecise and inaccurate [16].App assessment and interviews with patients revealed that wrong results despite correct input values were a problem [24,39].In one study, health care providers criticized that mHealth applications were not able to consider each aspect related to subjects separately [33].

Technology
Technical problems were mentioned in 18 articles [16, 18-20, 22, 24-27, 29, 33, 37-40, 42-44] and were found in both devices and software.Three quantitative studies [22,24,27], five qualitative studies [16,18,37,39, 42], nine mixed-design studies [19,20,25,26,33,38,40,43, 44] and one review [29] described problems related to technology.A case study with an older user found dependence on technological support [37].Patients faced issues related to hardware in lack of free storage space on the smart phone, short battery life [20] and use of small devices with small screens such as smartphones [44].Some patients still use feature phones with limited functions compared to smartphones and accordingly cannot use apps [22].Software-related issues were seen in functionality [27,43], challenges with software updates and technical issues with operating systems [44].Patients reported bugs, glitches or intermittent screen freezes in two apps [16,44].Technological failures may lead to subsequent errors induced by physicians [33].Besides the technological problems, it was stated that detecting these potential issues prior to app distribution is a challenge [24].
A further issue was seen in compatibility.Problems for patients could result from incompatibility or difficulties between apps and running devices [18,20,26,38], between running and external devices [19,20] as well as connection to a server [40] or electronic health records [42].Some articles mentioned technical problems, which were not described any further [19,25,29,37,39,44].One mixed-design study referred to technical difficulties with access but did not provide further details [38].
Patients and health care professionals perceived lack of motivation [18,21] and lack of engagement in users [16,19,25,40,43] as reasons for low adherence and high number of dropouts.In one study users expressed that they had forgotten to use the app [20].Some patients perceived the app as an additional burden or found it overwhelming [19,42,44].Some participants did not update goals in order to conform to expert recommendations and to keep goals achievable [19].Otherwise patients expressed concerns of being judged if they did not complete or miss lessons [37].Two qualitative studies revealed that some patients used the app not according to given advices [39,42].Some social situations [19,20,24,25,37] or disease specific contexts [37] were reported as being problematic.Environmental influences [19,26] and special use cases [19] were further issues.One review described that distraction by other online activities could be a problem [29].
Patients pronounced that lack of pause option [26] and difficulties with integration in everyday life can be a significant barrier to adoption [44].A further risk was a possible interference of technology use on relationships [40].
Lack of time was a major factor decreasing usage [20,31,42,44].Repetitive, long, complicated and boring contents reinforced the problem and might lead to even less time spent for mHealth app use [19,21,25,26,29,38].Qualitative studies and reviews described that the lack of human factor also affected usage and adherence.In other words, mHealth apps lacked personal touch, empathy and further complex aspects of human interaction [17,21].Communication was sometimes seen as ineffective [29].Some people will not use mHealth apps and reject them because they see their recovery as a process that depends only on the health care professionals caring for them [16].
First, concerns about weak security arise in the context of data security and privacy [34,35].Thus, a study of chronically ill patients found that 37.2% of them reported being concerned about the disclosure of personal information [35].Further problems regarding data security and privacy were access without permission [16] and possible breaches of and concerns about confidentiality [16,27,39,33,42].One app assessment study emphasized in its discussion that when data is stored on provider servers, there is an increased risk that data will be used or sold for undesirable purposes [27].Finally, health care practitioners mentioned problems regarding the identification of individuals by unauthorized data access [21] and patients worried about unpredictable consequences of data leaks [34].
The other problem was transparency of data handling [40].While some apps did not provide privacy policy [27,28], others were not clear or difficult to understand [21,27].

Doctor-patient relationship
In this context, 'Doctor-patient relationship' stands for all types of relationships between patients and health care providers.Thus, it includes therapeutic relationships as well.Problems in this category were described in 13 articles [16, 21, 23, 25, 27, 29-32, 34, 38, 40, 44] and include 'the attempt to replace the clinician', 'lack of a therapeutic alliance', 'negative impact on the relationship', 'information inequalities' and 'the question of responsibility'.This category was addressed in two quantitative studies [27,31], four qualitative studies [16,21,32,34], five studies with mixed-design [25,30,38,40,44] and in two reviews [23,29].Both patients and health care professionals mentioned problems with the lack of face-to-face contact.
Both assumed preferences for face-to-face communication for some patients and providers [21,34].
Health care professionals particularly emphasized the lack of nonverbal and para-communication associated with face-to-face conversations [21].Nevertheless, the spectrum of physician replacement ranges from taking over individual decisions [40] to complete replacement [21,23,44].One review described significantly lower treatment effects due to substitution of face-to-face intervention [29].
Absence of a treating person resulted in lack of therapeutic space [38] and therapeutic alliance [21,25] considered vital for successful therapeutic care [21].Without human support, one study found difficulties with user engagement in active components [31].Problems with 'Doctor-patient relationship' also occurred when mHealth apps were integrated into the treatment process.Even if technology can assist in health care, concerns in regards of interference with relationships were pronounced by patients [34] and health care providers [21].
Physiotherapists saw the problem that especially mutual trust could suffer from continuously monitoring a patient [25].
Limited capacity to export or download data reports reduces the ability to communicate directly from the app with others [27].Information asymmetries can arise and specialists could end up in situations where patients receive treatment results before they do [32].Lastly, responsibilities are altered and might lead to new problems.Physicians expressed the concern that they have to handle additional data or alerts and that use of mHealth apps could lead to detraction of patient's own self-management [16].Physicians who did not engage as leaders in digital interventions were also seen as problematic in one study [30].
Little, bad or no experience with apps is seen as a major problem [18,19,42] and fosters the issue of low abilities and confidence with technology use [16,19].While young individuals showed few difficulties in app usage, older patients, especially those with conditions such as dementia [27] or declining cognitive functions [37], face difficulties in app use [19].A special problem is that senior users have more problems because they use their mobile phones just for known functions and discourage themselves from learning new technology through trial and error [22].
A major barrier for app use is found in its perception.Irrespective of individual apps, some patients believe that mobile phones [34] and apps [22,35] are complicated and difficult to use.Patients also might feel dismissed because they see inferior care in digital products compared to face-to-face contact [21].
Besides the above-mentioned problems in digital literacy [41], literacy in general and numeracy were found to be a barrier for app use [21,24].In two studies, participants did not understand specific app functions [19,33].
Clinicians experienced similar problems as patients.Low experience and skills were found frequently [16,19,21,23,29,30,39].Some clinicians even had a more negative attitude regarding this type of interventions than patients [17].Others expressed lack of confidence in the integration of technology in health care [21].

Individuality
A further problem mentioned in 15 articles [16, 17, 19, 21, 25, 26, 29, 33, 35-37, 39, 40, 43, 44] is the intention or capability to customize mHealth apps to the individual needs of patients.One quantitative [35], four qualitative [16,21,37, 39], eight mixed-design studies [19,25,26,33,36,40,43,44] as well as two reviews [17,29] included problems and barriers corresponding to individuality.This is for example expressed in the fact that mHealth apps are usually not adapted to each individual [17, 19, 33, 35-37, 40, 44].Thus, authors discussed the difficulty in designing attractive and useful programs for all patients which are at least as effective as standard therapy [17,36] and pronounced the difficulties due to the diversity of the target users, especially in terms of age [33, 35, 37] and diseases [35].Furthermore, authors described individualization of functions due to perceptual impairments [37] and motor or physical issues [44] to be problematic.In one study, patients indicated that the goals set by the app were too simple and that the app could not be customized to their needs as much as necessary [40].Different functions are affected by a lack of individualization.Patients and health care professionals expressed that exercise programs often consist of a fixed number of different standard exercises [16,25,26], that data input is limited to imprecise standardized possibilities [39] and communication provided by the mHealth app is unadjusted [21,43,44].A special problem is the so called 'cold start problem'.It describes the need of time at the beginning of the intervention to personalize app content to the user profile through artificial intelligence [26].

Implementation
Implementation of mHealth apps into health care system faces different problems.Problems related to implementation were found in four quantitative studies [22,24,31,35], five qualitative studies [16,21,39, 42], five mixed-design studies [30,38,40,41,44] and four reviews [17,23,28,29].Barriers to access were seen as a problem for implementation.These occurred due to a lack of infrastructure, socioeconomic conditions or social reasons.Lack of access (e.g.lack of smartphones or broadband and computers) is a fundamental barrier for the use of mHealth apps [22,35,38,42].Disparities in access subsequently foster concerns that only a fraction of users benefit from apps in health care [21,22,29].Issues in the context of equity may stem from income or disability and result in non-equally distributed devices and connectivity [28].Further barriers to access concerned stigma and culture [31] as well as language [28,31].Though, no further information was given on these barriers.
Further problems concerned transferability of study effects to real-world care, organizational barriers, such as lack of capacity or preparedness of health care systems and reimbursement structures.Successful transfer into clinical practice was seen as a problem [17,23,41].Many questions, for example regarding modes of action or for which target groups app-based therapy is most suitable, are still unanswered.Thus, mHealth programs showing effectiveness in experimental settings do not necessarily show good results in real health care settings [17].Staff members reported low expectations and low confidence in the ability of national health care systems to implement digital tools [21].Barriers for implementation were lack of health system readiness, organizational resistance to change and policy uncertainties [44].Approval of apps, e.g., by the U.S. Food and Drug Administration (FDA), focuses on safety and minimal effectiveness thresholds and does not provide enough information for decision makers [28].Reimbursement options are not uniform [28] and a lack of collaboration among stakeholders, such as developers, health care professionals and patients, in the design and development process affects acceptance and adoption [16].
Low acceptance is a twofold problem.On the one hand, some professionals have less interest in information in apps, than in paper based information.This was highlighted by patients commenting that care providers always asked for paper forms despite information being provided in app format [39].On the other hand, professionals need to be open to the use of mHealth apps because their strong leadership engagement and promotion are fundamental for mHealth use [30,35].Health care professionals see usage of digital solutions as an additional burden placed on them [21,40,42] and expressed fear regarding the complexity of, and the responsibility for, identifying and managing risk [21].Interacting with mHealth apps was frequently seen as obstructive for workflows [21,28,42].Three further problems concerning the implementation were: Firstly, some app manufacturers were not available and did not respond to requests [24,42].Secondly, some users expressed the amount of choice being overwhelming [16] and finally, frequent app updates, requiring evaluation of new and confirmation of old functions were potential problems [24].
One study mentioned potential costs for apps as a concern [18].Furthermore, a problem for patients is the lack of opportunity to test and evaluate apps before they are purchased [27].The costs usually have to be borne by patients and might lead to socio-economic inequalities [21].Another problem for patients is the lack of opportunity to test and evaluate apps before purchasing them [27].
Just as with traditional health services, health care practitioners need time to integrate mHealth apps into treatment.Yet the effort is often not reimbursed [28,29,42,44].Therefore, providers demanded that time used for mHealth interventions be compensated in the same way as face-to-face treatments [28].Some clinicians questioned the value of investing in mHealth apps and preferred investing in staff training and staff employment rather than digital tools [21].One article did not specify the problem of costs [31].

Principal findings
This scoping review maps the evidence on potential problems and barriers related to the use of mHealth apps fulfilling the basic criteria of DiGA.The inclusion criteria were in particular (1) low risk class (I or IIa), (2) usage by the patient, (3) relation to illness, injury or handicap, (4) not for primary prevention and (5) the medical purpose is achieved through the main digital function.To our knowledge it is the first scoping review on this topic.In total 29 papers on mHealth apps met the inclusion criteria.The included studies showed a large heterogeneity and identified problems and barriers were often only a byproduct in included articles.
The majority of the studies originate from English-speaking countries.Thereof, eight stemmed from the USA [18, 27-31, 42, 43], five from the UK [19,21,24,40,41] and two from Ireland [24,28].Four studies originated from Asian countries (cf.Appendix 1).Two each from China [35, 37] and Korea [22,34].Despite the presupposed DiGA similarity of the apps described in the studies, none of the included studies came from Germany.Included studies differed strongly in terms of study design (cf.Appendix 1).Eight studies had a qualitative and five studies had a quantitative design.Furthermore, 12 mixed-method studies and four reviews were included.Most included studies used interviews [16, 19, 25, 30, [19,20,38].Study populations investigated also varied widely (cf.Appendix 1).While some studies focused on relatively balanced study populations [31], other studies included very specific populations such as people with military background suffering from post-traumatic stress disorder [30] or an older woman suffering from insomnia [37].Identification of relevant aspects and categorization of problems and barriers was done by two reviewers, independently.The categorization was done by clustering aspects to consistent groups.New groups of problems and barriers were compiled, if an identified problem could not be matched into the groups, already existing.This proceeding revealed ten major categories of problems and barriers on a super ordinated level: 'validity', 'usability', 'technology', 'use and adherence', 'data privacy and data security', 'patient-physician relationship', 'knowledge and skills', 'individuality', 'implementation' and 'costs'.
The categorization into the 10 problem groups is an approach to systematize problems and barriers in the context of mHealth usage identified in the literature.In addition to the categories defined by the scoping review, it would be conceivable to include further categories or subcategories.For example, "demographics" could be such a category.This could include identified problems such as the problems of older patients to use the app or socioeconomic inequalities that pose problems to the access.Although it would in principle be conceivable to define other problem categories, our research approach has proven to be well suited to identify relevant categories.All identified problems and barriers could be clearly assigned to a category.In addition, "actuality" would be another problem category that could be considered, as outdated content or technology could, in the worst case, lead to a compromise of patient safety.Unfortunately, however, in this review no relevant texts including this type of problem were identified.Our final 10 categories were formed qualitatively based on the available evidence and should be used as a fundamental basis for further discussion and research.While conducting the scoping review and interpreting the results, it was suggested that there might be some correlations between the problem categories.Thus, one category might have a direct influence on another category.For example, such correlations were suspected between 'use and adherence' to mHealth apps and their 'usability' or between 'implementation' and 'knowledge and skills'.However, since these are not confirmed results of the scoping review, these assumptions should be pursued in further studies.
Our results show that research in the area of problems and barriers is still rare compared to research of opportunities and possibilities.The problem categories identified can be attributed to the mHealth apps themselves on the one hand, and their integration into the healthcare system on the other.Regarding app-level problems (e.g.'Validity', 'Usability', 'Technology', 'Data Privacy and Data Security'), there are already quality assessment tools especially developed for mHealth apps [45][46][47][48] that aim to ensure the quality of apps.Other issues, such as 'Use and Adherence', 'Patient-Physician Relationship', 'Knowledge and Skills', 'Implementation', and 'Costs', affect the entire healthcare system.In contrast to quality assurance approaches at the app level, such approaches do not exist yet at the system level.Especially in this area more research is needed.Only if the integration of mHealth apps in the health care system succeeds as a whole, patients will sustainably benefit from the new technology.In order to achieve this goal, it is mandatory to explore those problems and barriers affecting various stakeholders.Not only scientists but also policy makers should have a special focus on these types of issues and address them within research and regulations.
In Germany, mHealth apps applying for the DiGA directory are examined initially for safety and suitability for use, data protection and information security, interoperability, robustness, consumer protection, ease of use, support of healthcare providers, quality of medical content and patient safety as well as evidence of positive healthcare effects [4].Other categories of problems such as 'use and adherence', 'Doctor-patient relationship' and especially 'implementation' and 'costs' are not sufficiently addressed, especially in high-quality studies, and need further investigation.
The identified problem categories can serve as a starting point for further research.For the less well studied ones, systematic studies of higher quality and scoping reviews should delineate the field; for the better studied problem categories such as 'validity', 'usability', 'technology' and 'data privacy and data security', systematic reviews might be more useful to gain insights.However, in addition to further reviews in this area of research, it is important to consider the results of primary studies.

Limitations
This review has some limitations.The search was not restricted to certain study types to capture a broad evidence base and to include aspects currently under discussion.Therefore, besides quantitative and qualitative studies, narrative reviews were included as well, as long as they met the inclusion criteria.Often, included articles addressed the problems and barriers only incidentally.
Only by including these different types of items, enabled us to create a broad evidence base.This might be a starting point for further research on certain categories such as implementation.
Since the systematization of the individual problems has taken a lot of time, more recent evidence should also be examined.Here, however, the approach of examining problem categories should be explicitly pursued.
During the screening process, we did not determine the agreement between the two reviewers or the kappa coefficient.Nevertheless, in cases of disagreement we have involved a third person.Thus, the inclusion or exclusion of texts has been done qualitatively.Furthermore, we have listed the studies that were excluded in the full-text screening with the reason for each in Appendix 3 to make our investigation more transparent and comprehensible by third parties.

Conclusions
Findings of this scoping review are not only relevant for DiGA but for all kinds of mHealth apps.Ten categories of problems and barriers were found.Issues at the app level such as 'Validity', 'Usability', 'Technology', 'Data Privacy and Security', 'Individuality' are addressed in several studies and are partly taken into account in quality assurance systems, problems and barriers related to the level ('Use and Adherence', 'Doctor-Patient Relationship', 'Knowledge and Skills', 'Implementation' and 'Costs') of health care system are rarely extensively studied.To optimize the integration of mHealth apps into healthcare, further research is essential, especially in the area of system-related problems.
In addition to serving as a starting point for further research, it is imperative that identified problems and barriers are considered in the development of new mHealth apps.

Funding
This paper is part of a larger research project (Continuous quality assurance of DHA ("QuaSiApps")).The project is funded by the Federal Joint Committee (G-BA).The funders had no influence on the study design, the conduct of the study, or the decision to publish or prepare the manuscript.We acknowledge support by the Open Access Publication Fund of the University of Duisburg-Essen.
medical Internet research, 20 (8) No specific app or device 6   (physical activity apps) Does not apply!Quantitative: Questionnaires The survey was conducted once.

Patients:
218 patients (f: 133) This study investigates on the current usage, willingness to use and barriers to using physical activity apps of Chinese patients with chronic diseases.The app was applied in a six-session treatment.

Qualitative:
Case study

Patients:
One 64-year-old-Chinese woman The utilization, advantages, and limitations were investigated.Patients used the app for 14 days in a home setting.Subsequently, a mixedmethods evaluation was performed.

Quantitative and Qualitative:
Questionnaire, semi-structured interview After the 14-day usage period, patients were asked to fill out the questionnaires.

Textbox 3 .
Exclusion criteria  Did not provide answer to the research question  The problem term mentioned in the abstract or title was not related to the investigated mHealth app  Publication does not focus on mHealth apps  Examined mHealth apps fulfill one or more of the following criteria: o Not used by the patient o No relation to illness, injury or handicap o Primary prevention o The medical purpose is not achieved through the main digital functions  Research Protocol or Conference Abstract  Article published before 2015  Language other than English, German or French

Figure 2 .
Problems and Barriers related to the use of DiGA (n = number of included articles mentioning the respective category)

Overview on included articles
https://doi.org/10.1136/bmjopen-2017-020906[37] Chen, Y. X., Hung, Y. P., & Chen, H. C. (2016): Mobile Application-Assisted Cognitive Behavioral Therapy for Insomnia in an Older Adult.Telemedicine journal and e-health : the official journal of the American Telemedicine Appendix 1: Quantitative and Qualitative: Feedback, questionnaire and interview After 8 weeks of intervention veterans completed a satisfaction questionnaire and a patient feedback interview.

Studies assessed in the full-text screening and exclusion criteria
Rules of engagement in mobile health: what does mobile health bring to research and theory?