Mobile health apps

This study addresses the burgeoning global shortage of healthcare workers and the consequential overburdening of medical
professionals, a challenge that is anticipated to intensify by 2030 [1]. It explores the adoption and perceptions of AI-powered
mobile medical applications (MMAs) by physicians in the Netherlands, investigating whether doctors discuss or recommend
these applications to patients and the frequency of their use in clinical practice. The research reveals a cautious but growing
acceptance of MMAs among healthcare providers. Medical mobile applications, with a substantial part of IA-driven
applications, are being recognized for their potential to alleviate workload. The findings suggest an emergent trust in AI-driven
health technologies, underscored by recommendations from peers, yet tempered by concerns over data security and patient
mental health, indicating a need for ongoing assessment and validation of these applications.

According to the World Health Organization (WHO), by 2030 the world will face a global shortage of 7 to 10
million healthcare workers, predominantly in developing and underdeveloped countries [1]. The doctors are
overworked, which results in general dissatisfaction and global strikes of people working in medical fields [2, 3].
With the rapid growth of technology, greater automatization in the field is one of the greatest factors supporting
prevention of the shortcoming. A part of medical innovation is supposed to be development and further adoption
in therapy of mobile medical applications (MMAs). That however is hampered by concerns from the public and
medical field.

Mobile Medical Applications (MMAs)

Mobile software technologies can improve the healthcare delivery and innovate the contact between patients
and doctors upon wider spread adoption and further development. Mobile medical application (MMA) is a
software application that can be executed on a mobile platform, or a web-based software application tailored
for a mobile platform, which is designed to be used for one or more medical purposes [4]. Purposes of MMAs
can be divided into four main categories: diagnostic, therapeutic, organizational and preventive. Diagnostic
applications, such as SkinVision [5], which via smartphone lens scans skin’s impurities in order to assess their
health risk, are meant to help diagnose conditions based on input data. The results from the application are
advised to be taken as a suggestion and not a final diagnosis. Therapeutic applications consist of applications
supporting longitudinal treatments, such as MyTherapy [6], which is a medication reminder and pill-tracking
application and mental support applications, such as Daylio [7] – a dairy and mood-tracking app. Preventive
applications focus mostly on health and wellness management and education. Organisational applications are
meant to support contact between patients and doctors. However, these are not as popularised yet as other
forms. A well-known example of such app could be MyChart which helps patients to manage their health records
as well as connects patients to the doctors [8].

Substantial part of MMAs that are currently being developed is AI-powered. AI-powered MMAs utilize artificial
intelligence algorithms to analyze data and provide insights or recommendations, while non-AI ones rely on pre-
programmed logic or user-inputted information without adaptive learning or predictive analysis. Furthermore,
AI-powered apps are characterized by their ability to constantly learn on vast amounts of data, refining their
algorithms over time. This evolution transpires within the predefined software constraints established by
developers. Non-AI apps operate within a static framework without the capability to assimilate or adapt to new
data patterns.

The global integration of MMAs, powered by both AI-based and traditional softwares, into healthcare is
currently facing several challenges, demanding attention from diverse stakeholders such as users, medical
professionals, policymakers, and insurers [9]. Those challenges often come from concerns and objections these
stakeholders have.


The research was executed using an online survey comprising of nine questions. The first part of the survey
consisted of 2 general questions aimed activate the participants’ recollections. That is an open question “Which
medical/healthcare applications have you used or are using?”, followed by a subsequent question to elucidate
the objective behind their usage – “What was the purpose of the application(s) you have been using?”.
The survey was designed to quickly conclude for respondents who haven’t discussed medical apps with a
doctor, while providing an avenue for those who have to offer detailed insights. Participants disclosed whether
they had discussed or been recommended a medical app by a doctor with a simple “Yes” or “No.” Those
responding “Yes” were prompted, in the following question, to identify the app used, selecting from a list of the
most downloaded medical applications on the Google Play Store, as shown in Figure 1, or specifying
another under “Other.”

The distinction between approaching a doctor with a MMA and being recommended one were distinguished
due to the versatile specifications of the applications. Diagnostic applications, as described above are
purpose-engineered to serve as initial consultative tools that, in the event of concerning results, need further
evaluation by a medical specialist.

The questions about application names were introduced in order to determine during analysis whether
application is fully or partially AI-powered. Based on that data a percentage of AI applications among the
traditional ones can be determined.


The incidence of doctors recommending mobile medical applications (MMAs) to patients in this study is lower
compared to the findings of Dahlhausen F.’s 2020 research in Germany. This could suggest a waning interest
in health applications among healthcare providers. However, this trend may also be attributed to the sampling
methodology, which predominantly involved younger participants. Physicians may presume younger individuals
are more capable of independently discovering appropriate applications. Conducting this research across a
broader demographic spectrum, particularly among less technologically adept adults and the elderly, would be
insightful. These groups might receive recommendations more frequently due to their potential unfamiliarity with
such technology and the consequent need for assistance in navigating these digital tools.
A noteworthy portion of the MMAs mentioned in the study are augmented by AI, yet it is crucial to
acknowledge that the majority are dedicated to menstrual cycle tracking, with “Flo Period & Pregnancy Tracker”
and “Clue” being the most cited. Apart from these, the only other AI-powered application that featured in
discussion with a healthcare professional was “Skinvision,” which is designed for the diagnosis of skin diseases
and it has not received any feedback from the doctor.

Concerns from the literature have been confirmed by the research conducted on patients sharing their
experiences with mobile health applications during doctor visits. “Data safety & security” is one of the primary
concerns of both patients and doctors, second only to the need for evidence-based accuracy in these tools.
The issue of “Worsening patients’ mental health” was less frequently mentioned by other scholars; however, it
is related to broader societal concerns about loneliness and the shift towards online interactions, which can lead
to feelings of isolation. Additionally, the pervasive nature of mobile health apps can lead to a heightened
awareness of health-related issues, as users receive constant updates and reminders on their devices.
Overall, the benefits outweigh the concerns of doctors, which gives a positive outlook for the future of mobile
medical applications. The findings reveal that the concerns regarding AI-driven and conventional MMAs are
shared by medical professionals. The benefits of IA-driven applications are seen in positive feedback about
specific ones from other doctors or patients. This pattern suggests an emerging trust in AI technologies, with
the act of recommendation serving as a pivotal factor in guiding doctors toward credible and effective
applications. Research shows that there remains a gap in reliable sources for medical practitioners to discern
which AI applications to implement in their practice, underscoring the need for a trusted framework to evaluate
and recommend such technologies.


[1] WHO, “Health workforce,”, Aug. 07, 2019.
[2] R. Tyssen, “Health problems and the use of health services among physicians: a review article with particular emphasis on Norwegian
studies,” Industrial Health, vol. 45, no. 5, pp. 599–610, Oct. 2007, doi:
[3] J. Wen, Y. Cheng, X. Hu, P. Yuan, T. Hao, and Y. Shi, “Workload, burnout, and medical mistakes among physicians in China: A cross-
sectional study,” BioScience Trends, vol. 10, no. 1, pp. 27–33, 2016, doi:
[4] C. for D. and R. Health, “Device Software Functions Including Mobile Medical Applications,” FDA, Sep. 09, 2020.
[5] Getting Started | SkinVision,”, May 19, 2021.
[6] “Medication Reminder and Pill Tracker App MyTherapy,”, 2020.
[7] “Daylio – Journal, Diary and Mood Tracker,” Daylio, 2021.
[8] “MyChart | Powered by Epic,”
[9] E. Bally and T. Cesuroglu, “Toward Integration of mHealth in Primary Care in the Netherlands: A Qualitative Analysis of Stakeholder
Perspectives,”, Jan. 15, 2020. Haggenmüller et al.,
“Digital Natives’ Preferences on Mobile Artificial Intelligence Apps for Skin Cancer Diagnostics: Survey Study,” JMIR mHealth and uHealth,
vol. 9, no. 8, p. e22909, Aug. 2021, doi:

Full article is currently under review.

Dr. Mark Vondenhoff
Senior Researcher
Mark Vondenhoff studied biology at Radboud University Nijmegen and has subsequently conducted fundamental biomedical research for many years. In particular in the field of the development of the immune system. Mark has a PhD in secondary lymphoid organ development and has worked as a postdoctoral researcher on immune cell development in human skin. He has been working at The Hague University of Applied Sciences since 2011, where he is senior lecturer in the Skin Therapy course. Through his work in the Data Science and the Oncological care research groups, Mark Vondenhoff conducts research in the field of suspicious skin lesions. The research he is interested in is mainly in the field of risk signaling of underlying pathologies upon recognition of clinical predictors in patients with a skin problem. In addition to research into the risk signaling of skin cancer, Mark Vondenhoff is also interested in other research in the field of (paramedical) skin care and skin cancer. To this end, he regularly assigns assignments for student teams. Both for student projects in the field of research with the aim of increasing knowledge and development projects in which students create practical solutions. He does this within graduation education of the Skin Therapy course. But also within the minor Skin Care Research in Practice and the minor Skin Care Innovation Project.