Introduction Violence against women is a pressing global health problem that is being met with a new intervention strategy—mobile applications. With this systematic review, we provide an initial analysis and functional categorisation of apps addressing violence against women.
Methods We conducted a systematic online search conforming with Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines to identify apps addressing violence against women in five World Bank regions (Europe and Central Asia; North America, Latin America and the Caribbean; Middle East and North Africa; South Asia; and sub-Saharan Africa). Applications with location of initiation in mentioned regions and ≥100 downloads were included. Data on sector, target group(s), year of release, location of initiation and implementation were extracted. By means of a structured qualitative content analysis, applications were then categorised according to their main functions.
Results Of 327 relevant applications, 171 were included into the systematic review and assigned to one of five identified categories of main functions, respectively: emergency, avoidance, education, reporting and evidence building, and supporting apps. The largest proportion (46.78%) consisted of emergency apps, followed by education, reporting and evidence building, supporting and avoidance apps in descending order. With regards to the geographical distribution of app categories, significant (χ2(20)=58.172; p=0.000) differences among the included regions were found.
Conclusion A vast proportion of apps addressing violence against women primarily draw on one-time emergency or avoidance solutions, as opposed to more preventative approaches. Further research is necessary, critically considering questions of data security, personal safety and efficacy of such mobile health interventions.
- systematic review
- prevention strategies
- public health
- health policy
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
What is already known?
Violence against women is a preventable global health issue of epidemic proportions, affecting approximately one-third of the global female population.
Recently, violence against women has been met with a new intervention of increasing global significance and scope, mobile applications (apps); despite their popularity and potential to heavily influence public health strategies such apps have not yet been systematically reviewed in the scientific literature.
What are the new findings?
Most apps addressing violence against women (46.78% out of 171 apps included into the systematic review) primarily use short-term emergency functions.
However, an increasing amount of apps offer education, reporting and evidence building and supporting functions as the prevailing feature.
What do the new findings imply?
Although all mentioned app functions can be beneficial to address violence against women, the dominance of emergency apps should be critically assessed, as it indicates that the currently prevailing mobile health strategy is highly skewed towards one-time solutions for isolated events of violence against women.
Further research on the benefits, safety, efficacy and sustainability of apps in the context of already existing traditional intervention strategies addressing violence against women is necessary.
Violence against women (VAW) is a major global health issue and a gross violation of women’s human rights that affects more than one-third of the global female population.1 It does not spare any geographical region, ethnicity, age or social class.2 Eliminating VAW is pivotal to achieve gender equality and women’s empowerment globally (Sustainable Development Goal 5).3 4 In the past years, mobile applications have begun to gain importance in the fight against such violence.5 6 The presumed common aim of these apps is to help victims, bystanders, and/or health workers implement different prevention or response strategies. Notably, the United Nations Commission on the Status of Women supports the development and use of mobile technologies as tools in the response to VAW in its general recommendation no. 35.7
However, to date, there has been no systematic review of such apps that is not restricted to apps concerning personal safety5 or rape6 and that discusses them in the realm of global health as it has been done for other mobile applications.8–10 This paper aims to fill parts of this research gap in the form of a systematic review on apps addressing violence against women in five World Bank regions (Europe and Central Asia; North America, Latin America and the Caribbean; Middle East and North Africa (MENA); South Asia; and sub-Saharan Africa). It aims to contribute to the discussion on the opportunities and challenges of mobile health (mHealth)11 solutions for VAW globally.
As defined by the WHO, mHealth is an area of electronic health that provides health services and information to both patient populations and healthcare providers via mobile technologies such as mobile phones, especially smartphones.11 Although the definition additionally entails further instantiations of mHealth, such as personal digital assistants,11 12 it is to be noted that, in this review, mHealth against VAW is referred to in terms of smartphone apps against VAW.
We used a systematic online search strategy to identify a apps that addressed VAW. We first developed a review protocol based on Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement recommendations.13 Our methodology with regards to identification and categorisation was in accordance with previous systematic app reviews.14 Building on our findings, we performed a structured qualitative content analysis15 of the included app descriptions. At the point of the data collection for the systematic review at hand, no systematic data compilation of apps addressing violence against women had been published.
Between 1 August and 23 August 2018, we carried out a systematic Google search as well as a systematic online search of two app reselling platforms (App Store and Google Play; see figure 1). For app identification on the reselling platforms, we searched apps in all group categories of the reselling platforms mentioned above, including all suggested search results for the following predetermined search items in English, German, French and Spanish: ‘Violence against women’, ‘gender-based violence’ (GBV), ‘anti gender-based violence’ (anti-GBV), ‘sexual harassment’, ‘women security’ and respective translations. For the Google search, mentioned search items were linked with the term ‘app[lications]‘ by the Boolean operator ‘and’. Scientific databases (PubMed, Medline, The Cochrane Library, Google Scholar and JSTOR) were also searched but did not reveal any relevant articles. Two authors (ES and KE) conducted the structured app search independently in the same time frame (see above). Discrepancies on eligibility were resolved by iterative discussion and consensus.
In this review, apps relevant to violence against women (see figure 1) were identified per mention of our search terms (see Identification) in the title and/or description of the app. If apps found in grey literature matched with those found on reselling platforms, information was obtained from both sources. However, they were only counted once and assigned to the respective reselling platform. Identical duplicates were removed.
Eligibility was determined by year of release (2010 up to and including 2018), number of downloads (≥100), interface language (English, German, French and Spanish) and explicit location of initiation in accordance with our regional subgrouping (ie, explicit location of initiation in exactly one of the included regions, see Data extraction).
Data of eligible apps were extracted from app descriptions on reselling platforms, respective app websites and non-academic journals (ie, newspaper articles, press releases and blog entries). Information obtained was coded, if applicable, for the following variables:
Sector (public, private or both).
Location of initiation.
Location of implementation.
Target group(s) (only victims, only non-victims, victims and non-victims).
Year of release.
Number of downloads.
Availability on reselling platform(s) (App Store, Google Play or both).
Coding per location of initiation (ie, location of initial development; as opposed to location of implementation, ie, all regions where an application is available) was further used for grouping with regards to location. The regions of initial development were organised in accordance with the World Bank regions into: Europe and Central Asia; North America, Latin America and the Caribbean; MENA; South Asia; and sub-Saharan Africa. East Asia and Pacific was excluded due to the small number of apps identified with location of initiation in this region; furthermore, applications whose location of initiation remained elusive were excluded. In total, n=12 apps were excluded due to these two criteria (see figure 1).
Qualitative content analysis: categorisation with regards to app function
Inductive categorisation with regards to the prevailing app function was based on structured qualitative content analysis and categorisation according to Mayring.15 Two independent raters (ES and KE) conducted the grouping by main function independently in accordance with the composed coding guidelines. In total, five distinct categories of apps were identified: emergency, avoidance, education, reporting and evidence building, and supporting apps (see table 1).
Overall, 32 out of 171 apps were found to combine more than one of the functions listed in table 1. For these apps, coding rules regarding relevance of app functions were obtained from the order of listed functions and key words signifying the importance of the functions to the respective app. In coding for group affinity, a function was defined as main function as soon as it was listed as first in the order of functions and accompanied by one of the following keywords: ‘Basic’, ‘main’, ‘fundamental’, ‘important’, ‘prevailing’, as opposed to ‘additionally’, ‘in addition to’, ‘furthermore’, ‘moreover’, ‘apart from’. In applying these coding rules, all 32 apps could be categorised to one of the mentioned groupings. The inter-rater reliability of the reviewed apps indicated strong agreement of 0.94. 8 apps out of 171 were categorised differently. This mismatch was resolved by iterative discussion and consensus.
Quantitative data were analysed with the Statistical Package for the Social Sciences (SPSS Statistics for iOS, IBM SPSS 26). For comparison of app quantities in different functional categories to sectors, target groups or geographical distribution regarding location of initiation, a χ2 test of homogeneity was deployed, respectively.
Patient and public involvement
Patients were not involved in this systematic review.
A total of 171 applications were used for a mapping of region-specific distribution patterns of apps (see table 1). This showed that 45 apps out of 171 regionally identifiable apps (26.32%) had their location of initiation in South Asia. Other regions with large numbers of apps were Europe and Central Asia with 42 apps (25.56%) and North America with 33 apps (19.30%). The remaining three regional subcategories had relatively few apps with 23 apps in Latin America and the Caribbean (13.45%), 17 in sub-Saharan Africa (9.94%) and 11 in the MENA region (6.43%, see figure 2A).
Target group and sector distribution
82.46% of all apps reviewed were directed at (potential) victims of VAW, that is, women, only (see figure 2B). In this paper, the term ‘victims’ is used to refer to both survivors and ‘potential victims’, that is all individuals potentially at risk. Additionally, the review found that a majority of apps (77.19%) were developed in the private sector (see figure 2C). In this review, the term ‘private sector’ is to be understood as non-governmental institutions/organisations and includes for-profit and non-profit organisations. ‘Public sector’, in turn, refers to governmental institutions. In a few cases (5.26%), apps were developed in collaboration between private and public sector institutions and were therefore classified as ‘both’.
The download of most (76.61%) apps identified in this review was available free of charge (f.o.c.) and did not include any additional cost in the sense of in-app purchases (see figure 2D). In some cases, the individual app’s cost remained unclear meaning that its actual cost could not be retrieved with the search strategies used in this review.
Global app function distribution
A proportion of 46.78% of all 171 apps addressing violence against women in our review was made up of apps whose primary function was to offer immediate help in emergency situations, for example, by automatically alerting a nearby police station or preselected contacts if activated by the respective user. These emergency apps (app categories, see table 1) constituted the largest proportion of apps globally. The second largest group was composed of education apps (21.05%) followed by reporting and evidence building (14.04%), supporting (12.28%) and avoidance apps (5.85%, see figure 3A).
Education apps (see table 1) are predominantly designed to provide information on the unacceptability and, if the case, unlawfulness of VAW, as well as on the benefits of egalitarian gender roles in society and in relationships. Several apps in this category target health professionals and social workers offering training resources and information in order to deepen the expertise on how to engage with women who have been victims of violence or are at risk.
Reporting and evidence building apps allow the user to report incidents of sexual violence and/or harassment via their mobile phone. The reported information about the nature and location of respective incidents are then further aggregated into street maps of incidents. This app category focuses specifically on visualising high prevalence of acts of VAW in public spaces.
Supporting apps (see table 1) aim to mitigate the social, psychological and physical repercussion of for survivors of violence by mediating between survivors and local support resources. This app category specifically targets domestic violence as an essential dimension of VAW. Apps from the avoidance category help users avoid potential at-risk situations by mainly offering all-female taxi services at night.
With regards to distribution of app categories, significant (χ2(20)=58.172, p=0.000) differences between all regions mentioned were found, as depicted in figure 3B. In many regions, emergency apps represented the largest functional category (see figure 3B). Also, as shown by download numbers (obtained from reselling platforms, if explicitly stated there), emergency apps represented the most frequently purchased category globally (see figure 3D).
Significant differences were also found regarding distribution of functional app categories and sector distribution (χ2(8)=22.501, p=0.004), with most emergency apps being developed by the private sector (see figure 3E).
Global trends with regard to app feature
An overall eightfold increase in app development rates was observed when comparing the years 2010 and 2018. Furthermore, changes in main function distribution of developed apps were identified, namely, a noticeable relative increment of the development of supporting, reporting and evidence building, and education apps (see figure 3C). To test whether these trends in primary app function are statistically significant, apps were aggregated to those released in the time frame 2010–2014 and 2015–2018, respectively, and compared with app function categories (see table 1), which showed a non-significant correlation (χ2(4)=4.202; p=0.379).
In this systematic review of apps addressing violence against women, we have assessed the global distribution of apps with regards to different parameters, particularly function, location of initiation and target group. With download numbers exceeding 160 million (these numbers were obtained from explicitly stated download numbers on mentioned sources only), it is easy to see that such apps have the potential to heavily influence the field of interventions to address violence against women.
Potential benefits of mHealth solutions may, for example, include a lowered threshold to report and/or seek help in case of violence as it has been documented that even in places where non-mHealth interventions are in place, 55%–95% of women remain hesitant and fearful of reporting such incidences in person.16–18 The barriers for reporting might be alleviated, for example, by anonymous reporting functions that certain apps offer (ie, reporting and evidence building apps). Furthermore, interaction between survivors and existing support institutions can potentially be facilitated by means of such apps, for both survivors of violence and respective support institutions (ie, supporting apps). Also, information and communication technologies (ICT) in general,10 19 20 and apps in particular, offer the opportunity for comparably easy and quick access to information (ie, education and supporting apps), potentially in multiple languages.21 Mentioned benefits of mHealth strategies especially concerns those who do not have direct access to health facilities where they could seek help in cases of violence but, in turn, to a smartphone—a scenario that often applies to underserved rural populations, particularly in low-income and middle-income countries (LMICs).22–24 However, in LMICs, it must be noted that while many individuals nowadays have access to a phone, access to a smartphone is relatively less common and often limited to phone-sharing between spouses, other family members or even a broader community. Although the exact prevalence of phone-sharing in LMICs remains unknown,25 studies suggest that it is a common practice that has stark implications,26 particularly for the safety of app-users as phone use usually requires permission of its main (commonly male) owner (see Conclusion).
Based on our findings, several important shortcomings of mHealth approaches addressing violence against women could be identified. As shown in this review, the biggest proportion by type of app function was made up of emergency apps for women as (potential) victims of violence. This dominance of emergency apps should be critically assessed, as it indicates that the prevailing mHealth strategy emphasises individual, one-time solutions for isolated events of violence. This focus was especially pronounced in regions with higher rates of VAW (eg, South Asia, see figure 3B)1 and among those apps developed by the private sector (see figure 3E). Yet, violence against women has been established as a widespread, structurally embedded and complex global problem, which is rarely experienced as a one-time individual incident.1 2 However, the overall shift in focus of apps in the recent years towards more preventative and inclusive mHealth strategies against VAW (ie, education, reporting and evidence building, and supporting apps) that, among other things, broach underlying (ie, societal and structural) aetiologies of VAW as indicated by our data (see figure 3C) is a favourable development.
Factors associated with violence against women are reflected in the so-called ‘socio-ecological’ or ‘ecological’ framework, a widely used theoretical model that considers how factors across four levels (ie, individual, relationship, community and societal) interact to put individuals at increased (or decreased) risk for experiencing and/or perpetrating violence.18 27 28 This model is extensively used, notably by the WHO18 28 and the Centers for Disease Control and Prevention,27 to delineate the multiple levels at which different factors operate to influence risk; in turn, evidence suggests that factors at all these levels should be addressed—separately or jointly—for successful interventions against violence against women.18 In contrast, apps oftentimes disregard the relationship, community and societal level, respectively, as our data suggest a global under-representation of education, reporting and evidence building as well as supporting apps (see figure 3A). Reporting and evidence building apps help shed light on the systematic societal level of sexual violence and harassment in public spaces as a global phenomenon. In turn, supporting apps focus on the relationship level of (mainly) domestic violence and help connect (potential) survivors of the latter to already existing ‘traditional’ support structures. This way, these app categories shift the focus from the individual to the relationship, community and societal level of VAW.
Too little attention is being paid still to mHealth-delivered primary prevention against violence against women, for example, in the form of education apps that explicitly highlight the relevance of gender equality, as ‘[t]he promotion of gender equality is an essential part of violence prevention’ in public health.16 Gender inequality, more specifically female inferiority, in turn, is a socio-structural concept underlying VAW and therefore should be addressed by education apps. In general, it is paramount for mHealth intervention strategies not to be seen as an isolated solution but rather as strategy that complements the long-term public health goal of prevention.17 Naturally, there is no need for every single app to address all factors or levels of the socioecological framework on its own; however, an overall balance of different approaches addressing all levels jointly in each region could provide better results.
In addition, the use of apps to promote change of men’s attitudes and behaviours as a potentially important element in the prevention of VAW seems to be poorly developed.16 18
Another important aspect is access in terms of app cost, which can be a factor limiting their effect, especially in LMICs. As described earlier (see Results, Target group, sector and cost distribution), the majority of apps can be downloaded free of charge leading to a relatively equitable access to different app-based interventions. However, it must be noted that many apps that include call or SMS sending functions, for example, emergency apps, come along with respective charges for such calls not by the application itself (through download or in-app purchases) but by the user’s network provider. Consequently, access to and use of many apps with a notification function of that kind still remains inequitable.
To effectively and responsibly seize the opportunities of mHealth technology in preventing and/or addressing violence against women, the issues above should be made visible and discussed.
In our review, we moreover refrained from further qualitative assessment of the respective applications, as has been done before for personal safety apps,5 since app download was partly restricted with regard to app store as well as geographical location and activity. Selective assessment of only those applications that were accessible to us would have led to a biased app sample and thus a biased quality analysis. However, our qualitative content analysis was based on extensive app descriptions (see Methods and Qualitative content analysis: categorisation with regards to app function), and we included all apps that adhered to our inclusion criteria (see figure 1), going far beyond apps that can only be downloaded in Germany which was the location of our study. This way, we were able to include more applications than any other review on this topic—and also more than the vast majority of reviews on mHealth applications in general32—thus leading to the first extensive review of apps addressing violence against women by which global trends may be identified.
Importantly, it should be noted that the review methodology used in this study clearly draws on the concept of categorisation. Each app was inductively categorised with regards to its prevailing function into one of five categories; as such, it is not a nuanced review of all functions an app entails. This approach was chosen due to its practicability for frontline health and social workers for whom an overview about which applications address VAW in what way in their respective region is a first step to know what kind of app(s) their patient population is using and which are available for them. This information might also be relevant for policy makers who are involved in planning and regulating both mHealth-based and non-mHealth-based interventions against violence against women.
In light of the high global prevalence of violence against women1 and the expanding field of mHealth technologies,11 as well as the high download numbers of apps (see figure 3D), this review presents a first step for further in-depth research on mobile approaches addressing VAW globally. The overall scarcity of discussions of such apps in the scholarly literature is striking. Also, previous reviews have not captured the whole spectrum of these apps, as they solely concern applications for personal safety5 or against rape,6 while in this systematic review, a considerable amount of apps for education, reporting and evidence building, and supporting purposes addressing all aspects of violence against women, not exclusively personal safety or rape, were identified.
Based on our systematic review, the following issues need to be taken into account in further explorations of apps against VAW.
As discussed above, such apps should address individual and relationship, community and societal factors, all of which contribute to violence against women and can be modified as intervention studies on non-mHealth approaches have shown.18 App development and implementation should be based on the most up-to-date scientific evidence on interventions for the prevention of and response to VAW. For as long as there is no evidence on the efficacy of mHealth approaches against VAW, recommendations on non-mHealth strategies will provide guidance on prevention, for example, the ‘RESPECT Women’ Framework endorsed by twelve UN, bilateral and multilateral agencies.18
Furthermore, it is not sufficient for apps to aim to address the broader social and institutional context of violence against women without integrating their approach into this broader context, namely linking them with existing institutions (eg, self-help groups, NGOs, women organisations and so on), and taking into account the social norms and institutions that continue to uphold VAW (eg, lack of targeted training on VAW in the police force, legal and healthcare system, lack of promotion of gender equality in education and so on).22 35–38 Thus, collaborations between mHealth and ‘traditional’ approaches should be actively sought,23 subordinating the technology to the overall aims of preventing violence against women and mitigating its impacts; and also integrating app functions into other domains that are not primarily focussed on addressing violence against women.
As an important step in the further evaluation of apps and app functions addressing violence against women, their potential purposes, risks and effects need to be assessed more thoroughly, for example, in the form of qualitative studies with users and other stakeholders, or systematic intervention studies. It will be paramount to understand potential harms such as data privacy breaches or loss of personal contact with supporters or other meaningful interventions resulting from a shift to digital technology. Furthermore, issues of safety need to be addressed as it must be ensured that women are not inadvertently put at more risk by using mHealth technology.18 This is especially the case for LMICs where mobile phones are often shared35. Hence, important confidentiality as well as safety issues of the user need to be taken into account in developing apps.35 36 In our review, we could identify security provision for the user mainly by password-secured access to the respective app. Many emergency app functions also included the option of placing an emergency call with a locked screen (eg, the apps Guardly mobile and Shake2alert; see online supplementary table 1). This way, the app may be accessed without anyone other than the user taking note of it. However, more research is needed on the efficacy, extent and sufficiency of such functions. For this, interviews with potential users and stakeholders of such apps as well as systematic intervention studies are needed for further evaluation. Such studies should include an assessment of risks and cultural acceptability of the mHealth intervention, as well as user satisfaction.39 40 Furthermore, the issues of potentially inaccurate data input, misinterpretation of the information, as well as issues regarding language and literacy problems must be taken into account when assessing mHealth intervention strategies in general41 42 and against VAW in particular.43 44 Misinterpretations can additionally occur due to lack of verbal and non-verbal cues in most apps.45
Taken together, mHealth solutions certainly are not the panacea for violence against women and should not be viewed as replacement for existing strategies. Nevertheless, if used effectively and integrated appropriately into existing interventions, apps or app functions addressing violence against women may present a promising tool in global health. As for many mHealth interventions, available evidence on effectiveness and overall benefit is sparse37 and, thus, needs to be further investigated in the future.
Handling editor Seye Abimbola
KE and ES contributed equally.
Contributors All authors designed the paper, read drafts and contributed written parts as well as commentaries on parts written by the coauthors. KE and ES conducted the systematic app search and categorisation, under supervision of VW and CG-M, and wrote the first draft. Equal authorship applies to KE and ES.
Funding The study was supported by the German Federal Ministry of Education and Research (BMBF); grant number: 01GP1791.
Disclaimer CG-M: The views expressed are the views of the author and do not represent policy of the WHO. VW: The funder had no role in the preparation of the manuscript or the decision to submit. The corresponding author (VW) has access to all the data in the study and final responsibility for the decision to submit for publication.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Patient consent for publication Not required.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement All data relevant to the study are included in the article or uploaded as supplementary information.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.