International Journal of Medical Informatics 110 (2018) 1–9

Contents lists available at ScienceDirect

International Journal of Medical Informatics

journal homepage: www.elsevier.com/locate/ijmedinf

Research Paper

A biopsy of Breast Cancer mobile applications: state of the practice review T

a,b,⁎ c d,e f,g,h i
G. Giunti , D.H. Giunta , E. Guisado-Fernandez , J.L. Bender , L. Fernandez-Luque
a
Salumedia Tecnologias, Seville, Spain
b
University of Oulu, Oulu, Finland
c
Internal Medicine Research Unit, Department of Internal Medicine, Hospital Italiano de Buenos Aires, Buenos Aires, Argentina
d
University College Dublin, Dublin, Ireland
e
Insight Centre for Data Analytics, University College Dublin, Dublin, Ireland
f
ELLICSR Health, Wellness and Cancer Survivorship Centre, Princess Margaret Cancer Centre, University Health Network, Toronto, ON, Canada
g
Centre for Global eHealth Innovation, Toronto General Hospital, University Health Network, Toronto, ON, Canada
h
Dalla Lana School of Public Health, University of Toronto, Toronto, ON, Canada
i
Qatar Computing Research Institute, Hamad Bin Khalifa University, Qatar Foundation, Doha, Qatar

A R T I C L E I N F O A B S T R A C T

Keywords: Background: Breast cancer is the most common cancer in women. The use of mobile software applications for
mHealth health and wellbeing promotion has grown exponentially in recent years. We systematically reviewed the breast
Breast cancer cancer apps available in today’s leading smartphone application stores and characterized them based on their
Gamification features, evidence base and target audiences.
Cancer
Methods: A cross-sectional study was performed to characterize breast cancer apps from the two major smart-
eHealth
phone app stores (iOS and Android). Apps that matched the keywords “breast cancer” were identified and data
Connected health
was extracted using a structured form. Reviewers independently evaluated the eligibility and independently
classified the apps.
Results: A total of 1473 apps were a match. After removing duplicates and applying the selection criteria only
599 apps remained. Inter-rater reliability was determined using Fleiss-Cohen’s Kappa. The majority of apps were
free 471 (78.63%). The most common type of application was Disease and Treatment information apps
(29.22%), Disease Management (19.03%) and Awareness Raising apps (15.03%). Close to 1 out of 10 apps dealt
with alternative or homeopathic medicine. The majority of the apps were intended for patients (75.79%). Only
one quarter of all apps (24.54%) had a disclaimer about usage and less than one fifth (19.70%) mentioned
references or source material. Gamification specialists determined that 19.36% contained gamification elements.
Conclusions: This study analyzed a large number of breast cancer-focused apps available to consumers. There has
been a steady increase of breast cancer apps over the years. The breast cancer app ecosystem largely consists of
start-ups and entrepreneurs. Evidence base seems to be lacking in these apps and it would seem essential that
expert medical personnel be involved in the creation of medical apps

1. Introduction a steady rise [3,4] and this cancer is no longer thought of as an acute
illness but rather a chronic condition. This means that we need to treat
Breast cancer is the most common cancer in women both in the breast cancer as such, with a focus on long term goals and wellbeing
developed and less developed world [1]. It is estimated that worldwide promotion [5]. Breast cancer survivors must be aware of the long-term
over 508 000 women died in 2011 due to breast cancer [1]. Incidence consequences of their treatment and be given information to encourage
rates vary greatly worldwide, ranging from 27 per 100,000 in Middle a proactive approach to their overall health [5,6]. Many authors now
Africa and Eastern Asia to 92 in Northern America and 89 in Western claim that the ever increasing number of breast cancer survivors require
Europe [2]. Low incidence rates from less developed regions however new models of care. These models should include a personalized needs
are probably due to a lack of early detection programmes. Fortunately, assessment; a self-management based care approach as well as in-
thanks to advancements in treatments, breast cancer survivorship is on dividualized follow-up and support [7]. Also, the rising cost of

Abbreviations: App, mobile software application; Avg, average; CI, confidence Interval; DG, Diego Hernán Giunta; EG, Estefanía Guisado-Fernández; GG, Guido Giunti; IQR, interquartile
range; JLB, Jackie Bender; LF, Luis Fernández-Luque; mHealth, mobile health; OS, operating system; SHF, Santiago Hors-Fraile; SE, standard error
⁎
Corresponding author at: Salumedia Tecnologias, C/Historiador Juan Manzano, S/N, Edificio Palmera Center – Oficina 28, Spain – 41089 – Dos Hermanas (Sevilla).
E-mail addresses: drguidogiunti@gmail.com, guidogiunti@salumedia.com (G. Giunti).

https://doi.org/10.1016/j.ijmedinf.2017.10.022
Received 25 October 2016; Received in revised form 11 September 2017; Accepted 31 October 2017
1386-5056/ © 2017 The Authors. Published by Elsevier Ireland Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/BY-NC-ND/4.0/).
G. Giunti et al. International Journal of Medical Informatics 110 (2018) 1–9

healthcare presents a compelling argument for developing more effec- 2.3. Selection criteria
tive health education and communication strategies aimed at im-
proving health outcomes [8,9]. Apps were included if the title and/or store description of the app
Connected Health is a healthcare delivery model that aims to was about breast cancer or breast cancer conditions, or cancer in gen-
maximize healthcare resources and provide increased, flexible oppor- eral, but contained specific mentions about breast cancer. A small
tunities for patients to engage with clinicians and better self-manage random sample (10%) was independently reviewed by two reviewers
their care using technology [10]. The use of mobile software applica- with ample mHealth experience (GG and JLB) who evaluated the
tions (apps) for health and wellbeing promotion has grown ex- eligibility of the apps against the selection criteria. In order to assess
ponentially in recent years [11]. Between 2013 and 2014 the global use clarity of the selection criteria, inter-rater reliability was assessed using
of smartphones increased by 406 million, reaching 1.82 billion devices Fleiss-Cohen’s Coefficient. Basic and “premium” versions of the same
(up 5% in a year), and Internet usage via mobile devices has increased app were considered as separate apps as were versions of the same app
by 81% in one year [12]. Mobile health (mHealth) is defined as the for different operating systems. This distinction was considered because
delivery of healthcare or health related services through the use of of the phenomenon of mobile device fragmentation in which different
portable devices [13]. There are currently thousands of healthcare re- versions of the same app must co-exist due to version capabilities or
lated mobile software applications (mHealth apps) available through store submission processes. This distinction is also common practice in
app stores [14]. This rapid proliferation of mHealth apps makes it in- this type of systematic app reviews [20,30]. Disagreements were re-
creasingly difficult for users, health professionals, and researchers to solved by consensus involving a third reviewer when necessary.
identify and assess which apps may be helpful and which may be in-
effective or even harmful. Concerns regarding the absence of healthcare 2.3.1. Inclusion criteria
professionals involvement in app development has been raised time and
time again [15–19]. • title and/or description is about breast cancer or breast cancer re-
Bender et al. [20] published a review in 2013 exploring the dis- lated conditions
tribution of cancer mHealth apps across the four major smartphone • title and/or description is about cancer in general but contains in-
platforms at that time, which found that most apps (45%) focused on formation about breast cancer
breast cancer. The focus of this review, however was only to assess apps
for the general public. Another review published in 2014 [21] targeted 2.3.2. Exclusion criteria
apps related to breast disease but it did not provide in depth study of
apps focused on breast cancer. Moreover, the past few years have seen a • description is not written in English
dramatic change in manufacturers and operating system (OS) market • duplicates from the same store
share, with some big players having almost disappeared today (ie: • title and/or description is not about breast cancer (ie. astrology,
Symbian [22]). Finally, the use of game elements in non-game contexts, breastfeeding, breast augmentation, chicken breast recipes, etc.)
commonly called gamification [23] has been gaining traction in health • title and/or description is about other specific type of cancer (ie.
apps and is now a popular strategy in both commercial and academic Pancreatic Cancer App)
fields [24–27], however current gamification prevalence in breast
cancer apps is unknown. 2.4. Data extraction
We systematically reviewed the breast cancer apps available in to-
day’s leading smartphone app stores and characterized them based on Data was automatically extracted from the store description of the
their features, evidence base and target audiences. app using the software application 42matters. Data extracted included
app information on: year of release, costs, downloads, ratings, title of
app, app description, categories, tags, languages, app websites,
2. Methods screenshots, etc.
GG and EG independently manually reviewed that information ex-
2.1. Study design tracted using structured forms by reading the store descriptions and
websites of the app that had unclear store descriptions or did not pro-
A cross-sectional study of breast cancer apps was performed to vide screenshots to extract information on: origin (eg, healthcare re-
characterize apps from the two major smartphone app stores: iTunes lated agencies, non-governmental organizations, universities, etc),
App and Google Play Store, which together represent more than 98.9% evidence base, features and intended audiences.
of the smartphone app market share [22]. Building upon the approach
used by Bender [20] we systematically searched both stores to identify 2.5. Data coding and classification
all relevant apps and provide a systematic presentation and synthesis of
the characteristics of the apps. Apps were classified based on their main purpose as described in the
store description into only one category following Bender et al. [20]
classification and our own scheme. If the purpose of the app was not
2.2. Setting clear from the description, a proper classification was discussed among
reviewers until consensus was reached. To ensure classification quality,
On February 24th 2016, we searched both stores from the United another randomly produced sample (10%) was selected from the list of
States using the keywords “breast cancer”. The iTunes App Store is a included apps and two reviewers with ample mHealth experience (GG
digital distribution platform developed and maintained by Apple Inc., and EG) classified them. Inter-rater reliability was once again assessed
for mobile apps on iOS. The iTunes App Store had 2 million apps using Fleiss-Cohen’s Coefficient.
available as of June 2016 since its launch in 2008 [28]. Google Play
store (originally the Android Market) is a digital distribution service 2.5.1. Application purpose
operated and developed by Google. It serves as the official app store for The application purpose classification scheme follows the work
the Android operating system. The Google Play store reached over 2.2 done by Bender et al. [20]:
million apps as of June 2016 since its launch in 2008 [28]. We down-
loaded all apps that partially or fully matched the keywords using the • Awareness-raising: tools to raise public recognition of cancer as a
software application for audience targeting called 42matters [29]. societal problem.

2
G. Giunti et al. International Journal of Medical Informatics 110 (2018) 1–9

• Fundraising: tools to attract financial resources for cancer control. might contain gamification elements. Due to the novelty of gamification
• Promote an organization: encourage awareness about a charitable as a discipline, specialists were defined as individuals with more than 3
organization raising awareness and funds for cancer or providing years of first-hand work experience designing gamification interven-
support to people affected by cancer. tions. Gamification elements include: self-representation with avatars;
• Disease and treatment information: provide general information three dimensional environments; narrative context (or story); feedback;
about cancer (eg, disease or treatment options). reputations, ranks, and levels; marketplaces and economies; competi-
• Prevention: provide information and practical tools to avoid cancer, tion under rules that are explicit and enforced; teams; and time pres-
including the recurrence of cancer. sure. These elements are in line with what was described by Reeves and
• Early detection: provide information and tools to assist in the Read [31] and the current body of literature used or discussed in the
identification of cancer before the emergence of symptoms or signs. literature for impacting health behavior [24–26,32,33]. Fleiss’ Kappa
• Disease management: provide information and practical tools to was calculated between the two specialists and the outcome was used as
deal with the medical, behavioral, or emotional aspects of cancer. our “gold standard” for gamification in our data set. More information
• Support: provide access to peer or professional assistance. on this process is detailed in another article [34].

Additionally, apps with an alternative or complementary medicine 2.6. Statistical methods

component (such as healing prayers, hypnosis, etc) were flagged for
later analysis. Categorical variables are presented as absolute and relative fre-
quencies. Quantitative variables are presented as mean and standard
2.5.2. Application origin deviation or median with interquartile range depending on distribution.
The app title, description and developer and/or uploading entity Landis & Koch’s standards for Fleiss-Cohen’s Coefficient are used [35].
were analyzed and coded based on the following criteria: Statistical analysis was performed using STATA version 13.

• Healthcare related Agency: hospitals, clinics, pharmaceutical cor- 3. Results

porations or governmental organizations directly related to health-
care (ie Public Health branches). The complete flow of this study is shown in Fig. 1.
• Governmental Agency: any governmental agency or organization
not directly involved in healthcare (ie IT departments). 3.1. Selection
• Non-governmental Agency: any organization that is neither a part of
a government nor a conventional for-profit business (ie Cancer A total of 1473 apps matched the search terms of “breast cancer”, of
Foundations). which 692 matches were from the iTunes App Store and 781 from the
• Educational Organizations: any educational organization such as Google Play Store. A random sample (n = 146) was independently
Universities, Colleges, Libraries or Schools not directly related to reviewed by two reviewers (GG and JLB) following the selection cri-
healthcare (ie Science School Projects). teria. Inter-rater reliability was determined using Cohen’s Kappa and
• Conferences and Journals: scientific journals, patient and/or med- found to be more than acceptable at 0.91 (SE 0.04 CI 95% 0.83 −
ical conferences. 0.98). One reviewer applied the inclusion/exclusion process with the
• Small and Medium-sized Enterprises: startups, software developing remainder of app sample. After removing duplicates only 599 apps met
companies or any other private organizations that identified them- the eligibility criteria (317 for iOS and 282 for Android).
selves as an enterprise and not individuals (ie Digital Health
Startups). 3.2. Classification
• Patient Organizations: societies or organizations that specialize both
in general health improvement as well as illness-specific objectives Another randomly produced sample (n = 68) was selected from the
and offer support groups (ie Patient Empowerment Organizations). list of included apps and two reviewers (GG and EG) independently
• Individuals: developers or uploader entities who are listed as in- classified them following the classification scheme. Cohen’s Kappa was
dividuals or have not identified themselves as enterprises (ie John also acceptable at 0.80 (SE 0.02 CI 95% 0.76–0.84) therefore one re-
Smith). viewer completed the classification of the remaining sample.
Table 1 shows a description of the app population that we explored.
Whenever discrepancies were found between descriptions and de- The vast majority of apps were free 471 (78.63%) and this remained
veloper or uploading entities, description was considered instead. constant independently of the OS (Android: 235 (83.33%); iOS: 236
(74.45%)). Android app developers classified their apps mostly in
2.5.3. Application target audience “Health and Fitness” and “Medical” categories (Fig. 2). iOS developers
The app descriptions were analyzed to assess intended target audi- used different tags with a greater dispersion as can be seen in Fig. 3.
ence and coded based on the following criteria: English language was predominant for both OS. Chinese, Japanese,
Korean and Russian followed English for Android; while German,
• Patient-oriented: intended to be used by the general public, patients Spanish and French followed for iOS. Android had a high number of
and/or their family members. missing language information for these apps.
• Physician-oriented: intended to be used by healthcare professionals Using the information available for each app about its release date,
or students from health related fields. we can see yearly growth in breast cancer apps and changes in app
market distribution over the years (Fig. 4).
2.5.4. Application credibility and disclaimers
App descriptions were explored for any type of disclaimer of use or 3.2.1. Application purpose
attribution of content, such as the presence of References. Based on our classification, the most represented type of application
are Disease and treatment information apps (29.22%), followed by
2.5.5. Gamification in data set Disease Management (19.03%), then Awareness Raising (15.03%) and
In order to assess the presence of gamification elements, two ga- Prevention apps (10.18%). These proportions held true for iOS and
mification specialists (SHF and GG) examined all of the included apps’ Android. Almost 1 out of 10 apps dealt with alternative or homeopathic
description and title independently and marked those they suspected medicine (see Table 2); in some cases these apps included misleading

3
G. Giunti et al. International Journal of Medical Informatics 110 (2018) 1–9

Fig. 1. Study Flow.

information such as using hypnosis or meditation to “cure” breast Only a quarter of all apps (24.54%) had any kind of disclaimer about
cancer. usage and less than a fifth (19.70%) had any mention of references or
source material. References and disclaimers for iOS apps were close to a
3.2.2. Application origin quarter for each (25.55% reference and 25.87% disclaimer) while in
SME and Individuals develop more than half of the breast cancer Android disclaimer presence almost doubled that of references (23.05%
mHealth apps. Only 1 app was developed by a Patient Organization. See disclaimer and 13.12% references). See Table 4. shows the presence of
Table 3 for more information. Fig. 5 and Fig. 6 show the proportion of references in app description per year.
app types for each developing agency group.
3.2.7. Gamification in data set
3.2.3. Application target audience After reviewing our app data set, the two gamification specialists
The majority of the apps were intended for patients (75.79%) and determined that 19.36% (n = 116) might contain gamification ele-
this was true for both Android (78.01%) and iOS (73.82%). See Table 2. ments. Apps with gamification elements in iOS doubled the amount of
Android apps (25% vs 13%). The inter-rater reliability between spe-
3.2.4. Application ratings cialists on this was 0.84 (SE 0.06, CI 0.71–0.96).
Apps can be rated by users using a five star rating system to indicate
their opinion. Less than half of the apps were rated (41.57%); Android 4. Discussion
apps were more frequently rated than iOS apps (69.15% vs 17.03%)
(see Table 1). 4.1. Principal results

3.2.5. Application downloads This is the first study to provide an in-depth analysis of the breast
Number of app downloads was only available for Android applica- cancer applications available to consumers. Our study paints a picture
tions. The store provides this information in terms of download ranges of the current ecosystem and the active stakeholders involved in it. We
(see Table 1). Close to a quarter of Android apps (24.47%) were classified apps based on their features and characteristics: the most
downloaded more than 1000 times but only a handful has been common type of breast cancer apps were about disease and treatment
downloaded more than 10,000 times (6.73%). information, followed by disease management and awareness raising
apps. We assessed the presence of references to source material to un-
3.2.6. Application credibility and disclaimers derstand the reliability of the information offered to consumers. Our
Table 3 also shows references and disclaimers per operating system. work is also the first to evaluate the extent of gamification elements

4
G. Giunti et al. International Journal of Medical Informatics 110 (2018) 1–9

Table 1
Basic app characteristics.

Android iOS All

n 282 317 599

Commercialization
Free 235 (83.33%) 236 (74.45%) 471 (78.63%)
Paid 47 (16.67%) 81 (25.55%) 128 (21.37%)
Rated 195 (69.15%) 54 (17.03%) 249 (41.57%)
Users rating† (avg) 58.62 5.21 30.36
Users rating† (median 2 (11) 0 (0) 0 (5)
IQR)

Rating (number of stars)‡

★ 2 (0.71%) 0 (0.00%) 2 (0.33%)
★★ 4 (1.42%) 6 (1.89%) 10 (1.67%)
★★★ 15 (5.32%) 9 (2.84%) 24 (4.01%)
★★★★ 81 (28.72%) 29 (9.15%) 110 (18.36%)
★★★★★ 93 (32.98%) 10 (3.15%) 103 (17.20%)

Number of downloads※
1–5 19 (6.74%)
5–10 9 (3.19%)
10–50 54 (19.15%)
50–100 23 (8.16%)
100–500 85 (30.14%)
500–1000 23 (8.16%)
1000–5000 32 (11.35%)
5000–10000 10 (3.55%)
10000–50000 16 (5.67%)
100000–500000 3 (1.06%)
Not Available 8 (2.84%) 317 (100.00%)

†Number of users who have rated the app; ‡Apps are rated based on a 5-star rating Fig. 3. App categories and tags for iOS.
system; ※Number of downloads are provided as a range by Google, this information is not
provided for iOS apps.
absence of healthcare professionals involvement in app development
continues to be raised time and time again. These concerns revolve
around app design and app content alike [15–19]. Production of
medical apps from non-medical stakeholders has benefits in terms of
creativity in design of apps and bridging disciplines, but there remains a
larger concern regarding the credibility of medical information within
such apps. It would seem essential that expert medical personnel be
involved in the creation of medical apps yet this seldom happens.
Areas such as breast cancer prevention and early detection should
have more active participation from healthcare organizations and
governmental agencies, however this is not the case. There is little in-
volvement from these stakeholders. Emerging trends in software design
like user-centered design or participatory design are characterized by
involving end-users in the phases of design [36–38]. The value of in-
cluding relevant stakeholders and users as part of the design team is
well recognized, and an essential aspect of good design practice for
adults and children alike [39–43].
The lack of academic reference found in our study is consistent with
Mobasheri’s previous work [21] and the available literature for other
disciplines. Authorship disclosure within app descriptions was highly
inconsistent making it difficult to determine involvement so unless
specified, no involvement was assumed. The US Food and Drug Ad-
ministration (FDA) has regulations for medical smartphone apps which
directly influence patient treatment [44], however most medical apps
are not formally evaluated under the current guidance [45,46]. Sources
and references were absent in almost 80% of app descriptions which is
alarming considering that over 70% of breast cancer apps regarded
Fig. 2. App categories for Android.
disease management, disease information and awareness.
The work done by Bender et al. [20] in 2013 showed how breast
present in breast cancer applications. cancer applications occupied a predominant role in cancer mHealth
Based on our study, the breast cancer app ecosystem largely consists apps. This is consistent with the steady increase in breast cancer apps
of start-ups and individual entrepreneurs. This is evidenced by more that our study shows for both operating systems. This pace could also be
than 60% of all breast cancer apps being developed either by SME or explained by the increasing penetration of smartphones [12] as well as
individuals. As such, it’s strange that both disease management and the continuous growth of mHealth applications [11]. However, not all
disease and treatment information applications were mostly developed of these apps are continuously maintained by their developers so the
by SME and individuals instead of healthcare related agencies. The danger of outdated information is present. App developers seem to be

5
G. Giunti et al. International Journal of Medical Informatics 110 (2018) 1–9

Fig. 4. New apps per year by OS and apps with references. Total
number of apps displayed on top of each year bar. Release date was
missing in 38 apps. * only first two months.

Table 2
Apps per type by OS.

All

n 599
Awareness-raising 90 (15.03%)
Fundraising 39 (6.51%)
Promote an organization 54 (9.03%)
Disease and treatment information 174 (29.04%)
Prevention 61 (10.18%)
Early detection 34 (5.67%)
Disease management 114 (19.03%)
Support 33 (5.51%)
Gamification elements 116 (19.36%)
Alternative Medicine 58 (9.68%)
Patient-oriented 454 (75.79%)
Physician-oriented 167 (27.88%)

Table 3
References, Disclaimers and developing agency by OS.

Android iOS All

n 282 317 599

References 37 (13.12%) 81 (25.55%) 118 (19.70%)
Disclaimer 65 (23.05%) 82 (25.87%) 147 (24.54%)
Healthcare related Agency 26 (9.22%) 52 (16.40%) 78 (13.02%)
Governmental Agency 2 (0.71%) 0 (0.00%) 2 (0.33%)
Non-governmental Agency 42 (14.89%) 53 (16.72%) 95 (15.86%)
Educational Organizations 2 (0.71%) 5 (1.58%) 7 (1.17%)
Conferences and Journals 27 (9.57%) 41 (12.93%) 68 (11.35%)
Small and Medium-sized 133 (47.16%) 103 (32.49%) 236 (39.40%)
Enterprises
Patient Organizations 1 (0.35%) 0 (0.00%) 1 (0.17%)
Individuals 48 (17.02%) 64 (20.19%) 112 (18.70%)

mentioning and citing sources for their apps more each year as can be
seen in Fig. 4. It would be interesting to study if this is in respond to Fig. 5. App type per developing agency for Android.
consumer demands or the market maturing.
The submission process and data organization is different for the
Gamification trends seem to be present also in breast cancer apps as
two stores so one to one comparison was not possible. There are sub-
1 in 5 apps contained some game element as reported by our experts.
stantial differences from store to store. Google Play uses a category
Another interesting find was the relatively few of alternative or com-
structure for internal classification while iTunes store uses tags and
plementary medicine apps in this topic. One in ten breast cancer apps
categories. However, it is interesting to note that Android apps classi-
deal with healing through hypnosis, herbal replacements or faith
fication system allows faster identification of medical or health related
healing. To our knowledge there are no other studies that cover the
apps (Fig. 2).

6
G. Giunti et al. International Journal of Medical Informatics 110 (2018) 1–9

version very often. This is now likely to change as apps are approved in
a matter of hours and Apple will let developers keep historical ratings
and reviews [52].

4.2. Limitations

This study has two main limitations. One is the fact that search al-
gorithms within stores return partial matches as well as full matches so
it’s possible for some apps to have been missed in our search. Keywords
used could have left out potentially valid breast cancer applications.
The other is its reliance on app store descriptions for data extraction
and classification. It was beyond the scope of this project to download
all the applications, many of which require payment for installation.
Although it is possible that in some instances developers may only
disclose sources, features or affiliations once in-app, this seems unlikely
given that such features are positive selling points and would therefore
likely be mentioned in app store descriptions if present. Furthermore,
app store descriptions are the only information available to most con-
sumers when deciding whether to download a given application.
We used apps available in the United States stores only which might
have caused some apps to be excluded (published in the UK or Canadian
stores for example). Structural differences between stores made it im-
possible to compare certain aspects (ie. iTunes doesn’t disclose number
of downloads per app; categories differ among stores, etc.). Restricting
app stores to Apple and Android based smartphones could also in-
troduce a selection bias as proportions might differ in less popular
platforms such as Windows or Blackberry phones.
It’s important to note that iTunes App Store and Google Play Store
have different processes and steps for app submission. These differences
can account for variability in the presence or absence of certain de-
scriptive elements. Fields that are required in one store might be op-
tional in the other.
Our breast cancer app review focuses on all breast cancer apps
Fig. 6. App type per developing agency for iOS. commercially available to patients, health professionals and public in
general. To our knowledge, this is the first study to assess the pre-
Table 4 valence of gamification in breast cancer apps and to describe the pre-
References and disclaimers per app type. valence of alternative and complementary medicine in the breast cancer
app population. Although previous works may have considered evi-
References Disclaimer
dence-base or medical professional involvement in development, our
iOS Android iOS Android
study also depicts what proportion each type of developing entity takes
of the available population.
Awareness-raising 1 (2.27%) 1 (2.17%) 1 (2.27%) 0 (0.00%)
Fundraising 0 (0.00%) 0 (0.00%) 0 (0.00%) 1 (4.76%) 5. Conclusions
Promote an 1 (3.03%) 0 (0.00%) 0 (0.00%) 2 (9.52%)
organization
Disease and 46 (45.54%) 17 (22.97%) 43 (42.57%) 30 (40.54%) This study analyzed a large number of breast cancer-focused apps
treatment available to consumers. Most breast cancer apps were designed for
information patients and focused on Disease and Treatment Information, Disease
Prevention 13 (41.94%) 3 (10.00%) 13 (41.94%) 8 (26.67%)
Management and Awareness Raising. Use of these applications to em-
Early detection 2 (11.76%) 0 (0.00%) 2 (11.76%) 4 (22.22%)
Disease management 18 (30.51%) 16 (29.09%) 21 (35.59%) 19 (34.55%) power patients and encourage preventive strategies, monitor symptoms
Support 0 (0.00%) 1 (5.88%) 2 (12.50%) 1 (5.88%) and behaviors, and provide effective interventions is appealing; how-
ever there continues to be a lack of involvement from healthcare pro-
fessionals, and a lack of properly cited source material or references in
prevalence of alternative or complementary medicine in mHealth apps. these applications.
Usually, App Store’s star rating systems and download ranges are The continuous growth of breast cancer applications shows a very
used to indicate popularity and indirectly measure the “success” of active ecosystem driven mainly by SMEs and Individuals with the pa-
these apps. These methods are not ideal as beyond the stores’ star rat- tient as the ultimate target of the app. Apps for healthcare professionals
ings published; little information on the quality of apps is available. were mainly conferences or journal apps with little in the way of
Using this popularity criteria yields little or no meaningful information helping patient education. This presents an interesting opportunity to
on app quality as has been discussed on occasions [47]. The lack of improve these patient facing apps and address the lack of healthcare
standardized quality measures continues to be concerning, as app use providers end of the equation.
carries risk and can lead to adverse outcomes for both patients and
clinicians [48–50]. Additionally, the way the stores handle ratings as Future research
app versions evolve affects the star system. In iOS for example, ratings
and reviews are limited to the most recent version as these reset when Understanding why a breast cancer application is successful is
apps receive upgrades. iTunes App Store approval process would complex as there is no standard for what a successful breast cancer app
usually take up some time [51] and developers wouldn’t release a new may be. Thus, what makes it successful remains uncertain. We intend to

7
G. Giunti et al. International Journal of Medical Informatics 110 (2018) 1–9

explore our current data set and look for associations between variables worldwide: sources, methods and major patterns in GLOBOCAN 2012, Int. J. Cancer
to create a model that would allow to weigh in the different factors that 136 (5) (2015) E359–E386, http://dx.doi.org/10.1002/ijc.29210.
[3] P.L. Bedard, L.L. Siu, Tilting the balance of dose modification for oral anticancer
could account for a successful breast cancer app. drugs? J. Clin. Oncol. 32 (15) (2014) 1537–1539, http://dx.doi.org/10.1200/JCO.
For both stores the main focus of development was towards patients 2014.55.2372.
and not healthcare professionals which poses questions as to why. This [4] S.N. Weingart, E. Brown, P.B. Bach, et al., NCCN task force report: oral che-
motherapy, JNCCN J. Natl. Compr. Cancer Netw. 6 (Suppl. 3) (2008).
presents a gap in the current repertoire of mHealth apps that should be [5] B.I. Bodai, P. Tuso, Breast cancer survivorship: a comprehensive review of long-
explored. term medical issues and lifestyle recommendations, Perm J. 19 (2) (2015) 48–79,
More study should also be done to understand how to improve http://dx.doi.org/10.7812/TPP/14-241.
[6] J. Cho, S.Y. Jung, J.E. Lee, et al., A review of breast cancer survivorship issues from
healthcare professional involvement during app design and testing. survivors’ perspectives, J Breast Cancer. 17 (3) (2014) 189–199, http://dx.doi.org/
10.4048/jbc.2014.17.3.189.
Author contributions [7] O. Pagani, M. Eicher, Survivorship −What women need after Breast cancer, Europa
Donna 50 (2014), pp. S43–S44, http://dx.doi.org/10.1016/S0959-8049(14)
70074-7.
GG conceived, designed, led overall study conduct, carried out the [8] I.R. Bardhan, M.F. Thouin, Health information technology and its impact on the
data collection, led analysis and interpretation of the data, and drafted quality and cost of healthcare delivery, Decis Support Syst. 55 (2) (2013) 438–449,
the manuscript. DG participated in overall study conduct, analysis and http://dx.doi.org/10.1016/j.dss.2012.10.003.
[9] K.E. Thorpe, C.S. Florence, P. Joski, Which medical conditions account for the rise
interpretation of study data. GG, JLB and EG carried out the data coding in health care spending? Health Aff. (Millwood) (Suppl) (2004), http://dx.doi.org/
and data interpretation. LF contributed to the analysis and interpreta- 10.1377/hlthaff.W4.437.
tion of study data, and conceptualized, reviewed and suggested mod- [10] University College Dublin Definition of Connected Health, e-health, m-health, tele-.
http://www.connectedhealthireland.com/what-is-connected-health/. Accessed
ifications to contents related to rigor and presentation of results. All March 14, 2016.
authors read and approved the final manuscript. [11] W.T. Riley, D.E. Rivera, A.A. Atienza, W. Nilsen, S.M. Allison, R. Mermelstein,
Health behavior models in the age of mobile interventions: are our theories up to
the task? Transl. Behav. Med. 1 (1) (2011) 53–71, http://dx.doi.org/10.1007/
Conflicts of interest s13142-011-0021-7.
[12] Cisco & Co, Cisco Visual Networking Index: Global Mobile Data Traffic Forecast
None. Update, 2015–2020, (2016) http://www.cisco.com/en/US/solutions/collateral/
ns341/ns525/ns537/ns705/ns827/white_paper_c11-520862. html.
[13] R. Whittaker, Issues in mHealth: findings from key informant interviews, J. Med.
Summary Points Internet Res. 14 (5) (2012) e129, http://dx.doi.org/10.2196/jmir.1989.
[14] R.-G. Jahns, The market for mHealth app services will reach $26 billion by 2017,
Research2Guidance, (2013) http://research2guidance.com/the-market-for-
• Breast Cancer patients and survivors represent a unique group [15]
mhealth-app-services-will-reach-26-billion-by-2017/.
A.D. Hamilton, R.R.W. Brady, Medical professional involvement in smartphone
who must be aware of the long-term consequences of their
apps in dermatology, Br. J. Dermatol. 167 (1) (2012) 220–221, http://dx.doi.org/
treatment and be given information to encourage a proactive 10.1111/j.1365-2133.2012.10844.x.
approach to their overall health [16] C.L. Ventola, Mobile devices and apps for health care professionals: uses and ben-

• The use of mobile software applications for health and well-

[17]
efits, P T 39 (5) (2014) 356–364 http://www.ncbi.nlm.nih.gov/pubmed/4029126.
S. O’Neill, R.R.W. Brady, Colorectal smartphone apps: opportunities and risks, Color
being promotion has grown exponentially in recent years. Dis. 14 (9) (2012) 530–534, http://dx.doi.org/10.1111/j.1463-1318.2012.03088.x.
• A 2013 cancer mobile applications review exploring the dis- [18] S.J. Wong, G.A. Robertson, K.L. Connor, R.R. Brady, A.M. Wood, Smartphone apps
for orthopaedic sports medicine −a smart move? BMC Sports Sci. Med. Rehabil. 7
tribution of all types of cancer in mHealth apps, which found
(1) (2015) 23, http://dx.doi.org/10.1186/s13102-015-0017-6.
that most apps (45%) focused on breast cancer. [19] K. Connor, R.R.W. Brady, A. De Beaux, B. Tulloh, Contemporary hernia smartphone
• A breast cancer specific mobile applications review took place applications (apps), Hernia 18 (4) (2014) 557–561, http://dx.doi.org/10.1007/
s10029-013-1130-7.
in 2014 but the smartphone app industry has dramatically
[20] J.L. Bender, R.Y.K. Yue, M.J. To, L. Deacken, A.R. Jadad, A lot of action, but not in
changed since then. the right direction: systematic review and content analysis of smartphone appli-
• The most represented type of application are Disease and cations for the prevention, detection, and management of cancer, J. Med. Internet
Res. 15 (12) (2013) 1–15, http://dx.doi.org/10.2196/jmir.2661.
treatment information apps (29.22%), followed by Disease
[21] M.H. Mobasheri, M. Johnston, D. King, D. Leff, P. Thiruchelvam, A. Darzi,
Management (19.03%), then Prevention (10.18%).
Smartphone breast applications − What’s the evidence? Breast 23 (5) (2014)
• Our study found that the breast cancer app ecosystem largely 683–689, http://dx.doi.org/10.1016/j.breast.2014.07.006.
consists of start-ups and entrepreneurs. [22] Statistica Global Mobile OS Market Share in Sales to End Users from 1 st Quarter

• The prevalence of Alternative and Complementary Medicine 2009–1 st Quarter 2016.; 2016. https://www.statista.com/statistics/266136/
global-market-share-held-by-smartphone-operating-systems/.
in Breast Cancer apps is assessed. [23] Deterding S, Dixon D, Khaled R, Nacke L, Sicart M, O’Hara K. Gamification: Using
• First Breast Cancer App Review to explore the presence any Game Design Elements in Non-Game Contexts. Proc 2011 Annu Conf Ext Abstr Hum
Factors Comput Syst (CHI 2011). 2011:2425–2428. 10.1145/1979742.1979575.
gamification or game elements.
[24] Deterding S, Dixon D, Khaled R, Nacke L. From game design elements to game-
fulness. Proc 15th Int Acad MindTrek Conf Envisioning Futur Media Environ?
MindTrek ’11. 2011:9–11. 10.1145/2181037.2181040.
[25] B. Ferguson, Games for wellness—impacting the lives of employees and the profits
Acknowledgements of employers, Games Health J. 1 (3) (2012) 177–179, http://dx.doi.org/10.1089/
g4h.2012.0023.
[26] J. Ferrara, Games for persuasion: argumentation procedurality, and the lie of ga-
Guido Giunti and Estefanía Guisado-Fernández gratefully acknowl- mification, Games Cult. 8 (4) (2013) 289–304, http://dx.doi.org/10.1177/
edge the grant number 676201 for the Connected Health Early-stage 1555412013496891.
researcher Support System (CHESS ITN) from the Horizon 2020 [27] C. Lister, J.H. West, B. Cannon, T. Sax, D. Brodegard, Just a fad? Gamification in
health and fitness apps, J. Med. Internet Res. 16 (8) (2014) e9, http://dx.doi.org/
Framework Programme of the European Commission. 10.2196/games.3413.
We would like to thank Santiago Hors-Fraile (PhD), Joaquin [28] Statista Number of apps available in leading app stores as of June 2016. http://
Chacón-Galvez, Prof. Minna Isomursu and Analia Baum MD for their www.statista.com/statistics/276623/number-of-apps-available-in-leading-app-
stores/. Published 2016. Accessed September 13, 2016.
cooperation and support. [29] Girardello A, Budde A, Wang B, Delchev I. 42matters. www.42matters.com.
Published 2016. Accessed February 26, 2016.
References [30] Giunti G, Guisado-Fernandez E, Caulfield B. Connected Health in Multiple Sclerosis:
a mobile applications review. IEEE Int Symp Comput Med Syst. 2017, (June).
http://ieeexplore.ieee.org/document/8104274/.
[1] World Health Organization, Global Status Report on Noncommunicable Diseases [31] B. Reeves, J.L. Read, Total Engagement: Using Games and Virtual Worlds to Change
2014, (2014) ISBN 9789241564854. the Way People Work and Businesses Compete 26 Harvard Bus Sch Press Books,
[2] J. Ferlay, I. Soerjomataram, R. Dikshit, et al., Cancer incidence and mortality 2009, http://dx.doi.org/10.1016/j.chb.2010.03.035 (6) 1.

8
G. Giunti et al. International Journal of Medical Informatics 110 (2018) 1–9

[32] M.N.K. Boulos, S.P. Yang, Exergames for health and fitness: the roles of GPS and 004.
geosocial apps, Int. J. Health Geogr. 12 (1) (2013) 18, http://dx.doi.org/10.1186/ [43] P.G. Forducey, R.L. Glueckauf, T.F. Bergquist, M.M. Maheu, M. Yutsis, Telehealth
1476-072X-12-18. for persons with severe functional disabilities and their caregivers: facilitating self-
[33] D. King, F. Greaves, C. Exeter, A. Darzi, Gamification: influencing health behaviours care management in the home setting, Psychol Serv. 9 (2) (2012) 144–162, http://
with games, JRSM 106 (3) (2013) 76–78, http://dx.doi.org/10.1177/ dx.doi.org/10.1037/a0028112.
0141076813480996. [44] FDA. Mobile Medical Applications. US Deapartment Heal Hum Serv Food Drug
[34] G. Giunti, D.H. Giunta, S. Hors-Fraile, M. Isomursu, D. Karoseviciute, Detecting Adm. 2015:44. http://www.fda.gov/downloads/MedicalDevices/
Gamification in Breast Cancer Apps: an automatic methodology for screening pur- DeviceRegulationandGuidance/GuidanceDocuments/UCM263366. pdf.
poses, Comp. Sci. (June) (2017) 22–24 http://ieeexplore.ieee.org/document/ [45] A.J. Barton, The regulation of mobile health applications, BMC Med. 10 (1) (2012)
8104273/. 46, http://dx.doi.org/10.1186/1741-7015-10-46.
[35] J.R. Landis, G.G. Koch, The measurement of observer agreement for categorical [46] A. Visvanathan, A. Hamilton, R.R.W. Brady, Smartphone apps in microbiology — is
data, Biometrics 33 (1) (1977) 159–174, http://dx.doi.org/10.2307/2529310. better regulation required? Clin. Microbiol. Infect. 18 (7) (2012) E218–E220,
[36] Scaife M, Rogers Y, Aldrich F, Davies M. Designing for or designing with? Informant http://dx.doi.org/10.1111/j.1469-0691.2012.03892.x.
design for interactive learning environments. Proc SIGCHI Conf Hum factors [47] M. Kuehnhausen, V.S. Frost, Trusting Smartphone Apps? To Install or Not to Install,
Comput Syst ? CHI ’97. 1997:343–350. 10.1145/258549.258789. That Is the Question, (2013), pp. 30–37.
[37] M.M.J. Muller, S. Kuhn, Participatory design, Commun. ACM. 36 (6) (1993) 24–28, [48] M. Gajecki, A.H. Berman, K. Sinadinovic, I. Rosendahl, C. Andersson, Mobile phone
http://dx.doi.org/10.1145/153571.255960. brief intervention applications for risky alcohol use among university students: a
[38] A. Dabbs, B. Myers, User-centered design and interactive health technologies for randomized controlled study, Addict. Sci. Clin. Pract. 9 (1) (2014) 11, http://dx.
patients, Comput. Inform. Nurs. 27 (3) (2009) 1–16, http://dx.doi.org/10.1097/ doi.org/10.1186/1940-0640-9-11.
NCN.0b013e31819f7c7c.User-Centered. [49] J. Torous, M. Keshavan, T. Gutheil, Promise and perils of digital psychiatry, Asian J.
[39] C. Sjoberg, T. Timpka, Participatory design of information systems in health care, J. Psychiatr. 10 (2014) 120–122, http://dx.doi.org/10.1016/j.ajp.2014.06.006.
Am. Med. Informatics Assoc. 5 (2) (1998) 177–183, http://dx.doi.org/10.1136/ [50] J.H. West, P.C. Hall, C.L. Hanson, M.D. Barnes, C. Giraud-Carrier, J. Barrett, There’s
jamia.1998.0050177. an app for that: content analysis of paid health and fitness apps, J. Med. Internet
[40] S. Pilemalm, T. Timpka, Third generation participatory design in health infor- Res. 14 (3) (2012) 1–12, http://dx.doi.org/10.2196/jmir.1977.
matics-Making user participation applicable to large-scale information system [51] Panzarino M. Apple’s App Store at the end of the app era. TechCrunch. https://
projects, J. Biomed. Inform. 41 (2) (2008) 327–339, http://dx.doi.org/10.1016/j. techcrunch.com/2016/06/12/apples-app-store-at-the-end-of-the-app-era/.
jbi.2007.09.004. Published 2016. Accessed June 8, 2017.
[41] J. Clemensen, S.B. Larsen, Participatory design in health sciences: using cooperative [52] Dillet R. The new iOS App Store lets devs choose whether or not to reset ratings
experimental methods in technology, Qual. Health Res. 17 (2007). when updating. TechCrunch. https://techcrunch.com/2017/06/07/ios-app-devel-
[42] C.M. Ruland, J. Starren, T.M. Vatne, Participatory design with children in the de- opers-will-be-able-to-keep-app-store-ratings-when-releasing-an-update/. Published
velopment of a support system for patient-centered care in pediatric oncology, J. 2017. Accessed June 8, 2017.
Biomed. Inform. 41 (4) (2008) 624–635, http://dx.doi.org/10.1016/j.jbi.2007.10.