Abstract
Background
The growing adoption of digital health applications (apps) presents new opportunities for General Practitioners (GPs) to enhance care and empower patients. However, little is known about how Portuguese GPs incorporate these apps into their practice.
Objectives
To identify the distinguishing characteristics of Portuguese GPs recommending the use of digital health apps to patients, and to investigate the facilitators and barriers influencing this behaviour.
Methods
A cross-sectional study was conducted using an online questionnaire distributed to Portuguese GPs between July 2023 and January 2024. Univariate logistic regressions identified predictors of app recommendation. Wilcoxon rank-sum tests compared facilitators and barriers between groups.
Results
A total of 126 GPs responded (72.2% women; median age 36 years [IQR: 31.8–43.0]); 45.2% recommended digital health apps. The most common were for apps for physical activity (32.4%), nutrition (21.3%), and chronic disease management (21.3%). Among GPs recommending apps, 70.2% did 1–4 times monthly. Most GPs believed that apps could improve chronic disease self-management (97.6%) and reduce face-to-face consultations (74.6%). GPs recommending apps were more likely to personally use health and fitness apps (OR 3.03), clinical decision apps (OR 3.79), and to believe that apps reduce face-to-face consultations (OR 3.46). GPs not recommending apps more often rated scientific validity as ‘very important’ (84.1% vs 61.4%, p = 0.006).
Conclusion
Nearly half of Portuguese GPs surveyed recommended digital apps, highlighting their potential to support self-management and reduce face-to-face consultation. Broader adoption may depend on fostering greater physician confidence in app use by strengthening the scientific evidence of apps.
Keywords: Digital health applications, digital health, mHealth, mHealth recommendation, general practitioner
Introduction
In 2011, the World Health Organisation defined mHealth as ‘medical and public health practices supported by mobile devices, such as cell phones, patient monitoring devices, personal digital assistants (PDAs) and other wireless devices’ [1]. Since then, the development of tablets, smartphones, and wearable devices has led to the proliferation of digital applications as a key resource for telemedicine. By the end of 2022, 41,517 health applications (apps) were available on the App Store® and 54,546 on the Google Play Store® [2].
Digital health apps are mobile software programs that help users process and manage health-related data. They are designed for anyone looking to monitor, improve, or maintain personal or community health [3]. In several countries, digital health apps available from the app stores have also been adopted at the governmental level to enhance patient health literacy, health maintenance and chronic disease management capabilities [4–6]. For example, in Portugal, the Shared Services of the Ministry of Health (SPMS) launched the SNS24® mobile app, which allows users to access a range of digital services and information from the National Health Service (NHS) [7]. Since there is no list of officially approved digital health apps, Portuguese General Practitioners (GPs) cannot formally prescribe these apps but can only suggest them to patients. Given their essential role in guiding and encouraging patients to adopt health-related behaviours, GPs can leverage these apps to support and track the maintenance of healthy habits. To achieve this, GPs must know how to choose mobile apps that best suit their patients’ goals [8].
In chronic disease management, digital health apps facilitate the remote monitoring of conditions and strengthen the relationship between the doctor and the patient [9–13]. Similarly, their utility extends to mental health, where a systematic review by Sin et al. demonstrated their effectiveness in treating common mental conditions [14]. The European Association for the Study of Diabetes and the American Diabetes Association have even collaborated to draft safety and validation guidelines for diabetes-related applications [12]. Evidence further supports the effectiveness of digital health tools in driving behaviour changes. Afshin et al. concluded that these applications can improve physical activity levels and help users reduce weight [15]. When digital interventions utilise behaviour change techniques, like prompts and goal setting, they become more effective [16]. Additionally, integrating digital health apps with medical appointments has proven more effective for maintaining healthy behaviours than using each tool alone [17]. These emerging tools have redefined Primary Care by motivating GPs to promote shared health management with patients and to encourage patients to adopt health-related behaviours that improve how they manage their conditions [18].
Preliminary evidence from Australia and Germany suggests that while GPs do use health apps themselves, they are often hesitant to recommend them to patients due to a lack of knowledge or trust in these tools [19,20]. Key factors influencing their willingness to recommend apps include perceived clinical value, concerns about risk and responsibility and familiarity with using digital tools [21,22]. However, the body of research in this area regarding the facilitators and barriers that influence recommendations is generally scarce. This leaves a significant gap in knowledge regarding how GPs in other countries, like Portugal, approach digital health apps and what factors affect their behaviours.
Consequently, this study aimed to identify the distinguishing characteristics among Portuguese GPs who recommend digital health apps and those who do not, investigating the factors influencing their recommendation behaviour.
Methods
Study design
This online cross-sectional survey study was conducted among GPs from Portugal. The study is described per the STrengthening the Reporting of OBservational studies in Epidemiology (STROBE) guidelines for observational studies [23]. This study was approved by the Ethics Committee of the Faculty of Medicine of the University of Porto (101/CEFMUP/2023). At the beginning of the questionnaire, participants were presented with an explanation of the study, its objectives, the expected response time, a guarantee of confidentiality, and the option to withdraw from any section of the questionnaire. The process of filling out the questionnaire began with electronically consenting to accept the terms, with a positive response needed before the first question. If the answer was negative, the questionnaire ended for the participant.
Study participants
The sample was selected using convenience sampling. No a priori power calculation was conducted, as the study was exploratory. The primary aim was to collect as much data within feasibility constraints, rather than to achieve a predetermined sample size based on power considerations. Eligibility criteria included being a general and family medicine specialist or resident and practising medicine in Portugal. The organisational structure of general practice in Portugal is based on primary health care centres that operate under a chain of command. Family health units (FHU), categorised as Type A and B (with Type B having greater autonomy), are self-organised multi-professional teams that include GPs, nurses, managers and other professionals. These teams have the autonomy to establish their working processes and to negotiate performance goals with local health authorities [24]. This survey was designed to ensure the representation of physicians’ perceptions regardless of their general practice setting. Therefore, the type of the FHU did not influence participant recruitment.
The study was disseminated on social media platforms (on content pages aimed at doctors) and sent via institutional email to medical groups/communities. The Research Department of the Portuguese Association of General and Family Medicine promoted the dissemination of the study on its official communication channels, which were aimed at its members (General and Family Medicine specialists and residents).
Data collection
Data were collected through an online questionnaire implemented on the Google Forms platform (Appendix A), which was active from July 2023 to January 2024.
The questionnaire was designed after a bibliographical review of works on the same topic of study, and the questions were comprised of four sections (Table 1). To ensure the clarity and the relevance of the instrument, we asked three to five GPs to review the questionnaire and provide informal feedback.
Table 1.
Questionnaire sections and their descriptions.
| Sections of the questionnaire | Details |
|---|---|
| 1 | Demographic (i.e. gender, age, academic qualifications) and professional characteristics (i.e. years of medical practice, region where they practice, medical practice in urban or rural areas, organisational model in which they operate (Family Health Units: Type A, B or private practice), experience as General Practice trainer/academic supervisor). |
| 2 | Use of digital health apps (i.e. for personal use, clinical decision-making and recommendation to patients). |
| 3 | Recommendation of digital applications (i.e. type and frequency of recommendation). |
| 4 | Factors influencing the recommendation of apps and the ability to change clinical practice (i.e. facilitators and barriers to the recommendation of applications, ability to manage chronic disease by users and potential for reducing face-to-face consultations using applications). This is assessed using a five-point scale: ‘not important’, ‘slightly important’, ‘reasonably important’, ‘important’ and ‘very important’. |
Data analysis
Descriptive statistics were used to characterise the sample, app use, and factors influencing recommendation. The sample was categorised into two groups, according to the self-reported recommendation of apps (i.e. those who recommend and those who do not). The normality of continuous variables was tested using the Kolmogorov-Smirnov test, and the median and interquartile range were used for descriptive purposes for variables without normal distribution. Absolute and relative frequencies were used for categorical variables. Univariate logistic regression was used to explore associations between participants’ characteristics and the recommendation of apps. The selection of variables was informed by a set of predefined hypotheses developed prior to analysis. These included the assumption that factors such as GPs’ personal or professional use of health technologies, clinical experience, and perceptions of app utility may influence their recommendation behaviours. Differences in the median scores for facilitators and barriers to app recommendation between GPs who do and do not recommend apps were assessed using Wilcoxon rank-sum tests with continuity correction, appropriate for the ordinal nature of the data. Statistical data analysis was performed using the features of the IBM SPSS Statistics version 29.0 (IBM Corporation®, Armonk, NY, USA) program.
Results
General practitioners’ characteristics
Of the 126 respondents (Table 2), the majority were women (72.2%, n = 91) with a median age of 36 [IQR: 31.75–43.0]. Nearly half (48.4%, n = 61) had 5 to 15 years of medical practice. Most participants were affiliated with Type B FHUs (52.4%, n = 66), which are located in predominantly urban areas (66.7%, n = 84). Additionally, one-third of the family doctors were internship advisors (66.7%, n = 84). Geographically, the South region had the highest proportion of respondents (51.6%, n = 65).
Table 2.
Participants’ characteristics (n = 126).
| Groups | Total (n = 126) | % |
|---|---|---|
| Sex | ||
| Female | 91 | (72.2) |
| Male | 35 | (27.8) |
| Age (yrs), median [IQR] | 36 | [31.8, 43] |
| Academic degree | ||
| Degree/Integrated master | 78 | (61.9) |
| Master/PhD | 48 | (38.1) |
| Clinical practice experience | ||
| <5 years | 31 | (24.6) |
| 5-15 years | 61 | (48.4) |
| 16-25 years | 18 | (14.3) |
| 26-35 years | 11 | (8.7) |
| 36-45 years | 5 | (4.0) |
| Experience as GP trainer Region of the primary care centre |
42 | (33.3) |
| North | 34 | (27.0) |
| Centre | 26 | (20.6) |
| South | 65 | (51.6) |
| Urban/Rural Status of the primary care centre | ||
| Urban | 84 | (66.7) |
| Rural | 42 | (33.3) |
| Type of primary care centre | ||
| General Primary Health Care Centres | 32 | (25.4) |
| FHU A | 26 | (20.6) |
| FHU B | 66 | (52.4) |
| Private practice | 10 | (7.9) |
Abbreviations: n: Number; yrs: Years; IQR: Interquartile range; PhD: Doctor of Philosophy; GP: General Practitioner; FHU: Family health units. Data on region and type of primary care centre were missing for 1 and 2 responses, respectively.
Recommendation and use of applications
Most GPs reported using digital wellness and health applications in their personal lives (68.3%, n = 86) and to support their clinical practice and decision-making (88.1%, n = 111) (Table 3). A greater proportion of GPs who reported recommending apps also reported using health and fitness mobile apps (80.7%, n = 46) compared to those who did not recommend apps (58.0%, n = 40). Similarly, the use of clinical apps was more common among those who recommend (94.7%, n = 54), than among those who do not (82.6%, n = 57). The vast majority of physicians (97.6%, n = 123) believed that digital health applications could improve patients’ ability to self-manage chronic conditions. Furthermore, 92.9% (n = 117) indicated that these applications have the potential to reduce the frequency of face-to-face consultations over time.
Table 3.
The use and perception of using applications among GPs.
| GPs recommending apps (n = 57) |
GPs not recommending apps (n = 69) |
Total (n = 126) |
||||
|---|---|---|---|---|---|---|
| n | % | n | % | n | % | |
| Use of applications | ||||||
| Use of health and fitness mobile apps | 46 | (80.7) | 40 | (58.0) | 86 | (68.3) |
| Use of clinical apps | 54 | (94.7) | 57 | (82.6) | 111 | (88.1) |
| Perceptions of using applications | ||||||
| Perceived improvement of chronic disease self-management due to the use of apps | 57 | (100) | 66 | (95.7) | 123 | (97.6) |
| Perceived reduction of face-to-face consultations due to the use of apps | 51 | (89.5) | 66 | (95.7) | 117 | (92.9) |
Abbreviations: GPs: General Practitioners; n: Number.
Almost half of respondents (45.2%, n = 57) indicated they recommend health apps to patients. Physicians who recommended apps did so 1 to 4 times per month (70.2%, n = 40). The recommended applications included those focused on physical activity (77.2%, n = 44), nutrition (21.3%, n = 29) and chronic disease management e.g. for the management of Asthma, Chronic Obstructive Pulmonary Disease, Diabetes or Hypertension (21.3%, n = 29) (Table 4).
Table 4.
Frequency and type of health applications recommended (n = 57).
| n (%) | ||
|---|---|---|
| Frequency of recommendation | ||
| <1 time per month | 9 | (15.8) |
| 1 time per month | 20 | (35.1) |
| 2 to 4 times per month | 20 | (35.1) |
| >4 times per month | 8 | (14.0) |
| Type of apps recommended | ||
| Physical activity | 44 | (32.4) |
| Nutrition | 29 | (21.3) |
| Chronic disease management | 29 | (21.3) |
| Mental health | 23 | (16.9) |
| Menstrual cycle/contraception | 4 | (2.9) |
| Smoking | 2 | (1.4) |
| Travel | 1 | (0.74) |
| Obesity | 1 | (0.74) |
| Drugs dose/price | 2 | (1.4) |
| NHS | 1 | (0.74) |
Factors associated with app recommendation
Several factors were associated with a higher likelihood of recommending digital health apps (Table 5). GPs were more likely to recommend apps if they had personally used health and fitness apps (OR = 3.03; 95% CI 1.34–6.84) or clinical decision apps (OR = 3.79; 95% CI 1.01–14.17) Additionally, GPs who believed digital health apps could reduce face-to-face consultations had increased odds of recommending them (OR = 3.46; 95% CI 1.02–11.70).
Table 5.
Univariate analysis to explain the recommendation of digital health apps (n = 126).
| Characteristics | Crude OR | (95% CI) |
|---|---|---|
| Sex | ||
| Male | Reference | |
| Female | 0.71 | (0.32–1.55) |
| Age | 1.03 | (0.99–1.07) |
| Academic degree | ||
| Degree/Integrated Master | Reference | |
| Master/Doctorate | 0.82 | (0.13–5.05) |
| Clinical practice experience (years) | ||
| <5 | Reference | |
| 5–15 | 0.49 | (0.20–1.22) |
| 16–25 | 0.48 | (0.15–1.57) |
| 26–35 | 0.57 | (0.14–2.33) |
| 36–45 | 0.32 | (0.05–2.21) |
| Region | ||
| North | Reference | |
| Centre | 0.68 | (0.23–1.99) |
| South | 0.68 | (0.29–1.60) |
| Urban/Rural Status | ||
| Urban | Reference | |
| Rural | 1.16 | (0.55–2.43) |
| Type of primary care centre | ||
| General primary care centre | Reference | |
| FHU A | 0.89 | (0.30–2.65) |
| FHU B | 1.49 | (0.63–3.55) |
| Internship supervisor | ||
| No | Reference | |
| Yes | 1.54 | (0.73–3.24) |
| Use of health and fitness mobile apps | ||
| No | Reference | |
| Yes | 3.03 | (1.34–6.84)* |
| Use of clinical decision apps | ||
| No | Reference | |
| Yes | 3.79 | (1.01–14.17)* |
| Improving chronic disease self-management | ||
| No | Reference | |
| Yes | a | |
| Potential reduction of face-to-face consultations due to the use of apps | ||
| No | Reference | |
| Yes | 3.46 | (1.02–11.70)* |
Abbreviations: OR: Odds ratio; CI: Confidence interval; FHU: Family Health Unit; Reference: the category used as reference (i.e. to which other categories are compared), a; insufficient number of cases. *p < 0.05.
Facilitators and barriers influencing app recommendation
Facilitators
Factors influencing the decision to recommend digital health apps are detailed in (Table 6). The factors most commonly rated as ‘important’ and ‘very important’ included the app’s ease of use for patients (97.7%), patient access to necessary hardware (96.8%), and whether the app was free to use (95.2%). Other highly rated factors were patient access to the internet (95.3%), along with the GPs knowledge of digital health (93.7%) and interest in digital health (93.6%). Those who do not recommend health apps were significant more likely to rate proven scientific validity as “very important” than those who do (84.1% vs 61.4%, p = 0.006). Ratings for all other facilitators did not differ significantly between the two groups (Appendix B).
Table 6.
Factors that influence the recommendation of applications.
| Faciliators of recommendation | GPs recommending apps (n = 57) |
GPs not recommending apps (n = 69) |
Total (n = 126) |
||||
|---|---|---|---|---|---|---|---|
| n | (%) | n | (%) | n | % | ||
| Doctor’s knowledge of available applications | |||||||
| Very important | 37 | (64.9) | 43 | (62.3) | 80 | (63.5) | |
| Important | 18 | (31.6) | 20 | (29.0) | 38 | (30.2) | |
| Reasonably important | 2 | (3.5) | 6 | (8.7) | 8 | (6.3) | |
| Slightly important | 0 | (0.0) | 0 | (0.0) | 0 | (0.0) | |
| Not important | 0 | (0.0) | 0 | (0.0) | 0 | (0.0) | |
| Wilcoxon’s W (p-value) | 2051.5 (p = 0.625) | ||||||
| Doctor’s interest in digital health | |||||||
| Very important | 32 | (56.1) | 44 | (63.8) | 76 | (60. ) | |
| Important | 21 | (36.8) | 21 | (30.4) | 42 | (33. ) | |
| Reasonably important | 4 | (7.0) | 4 | (5.8) | 8 | (6. ) | |
| Slightly important | 0 | (0.0) | 0 | (0.0) | 0 | (0.0) | |
| Not important | 0 | (0.0) | 0 | (0.0) | 0 | (0.0) | |
| Wilcoxon’s W (p-value) | 1816.5 (p = 0.395) | ||||||
| Time available during appointments | |||||||
| Very important | 33 | (57.9) | 48 | (0.0) | 81 | (64.3) | |
| Important | 20 | (35.1) | 16 | (1.4) | 36 | (28.6) | |
| Reasonably important | 3 | (5.3) | 4 | (9.5) | 7 | (5.6) | |
| Slightly important | 1 | (4.8) | 1 | (23.2) | 2 | (1.6) | |
| Important | 0 | (0.0) | 0 | (69.6) | 0 | (0.0) | |
| Wilcoxon’s W (p-value) | 1754.5 (p = 0.219) | ||||||
| Application in Portuguese | |||||||
| Very important | 30 | (52.6) | 39 | (56.5) | 69 | (54.3) | |
| Important | 22 | (38.6) | 21 | (30.4) | 43 | (34.1) | |
| Reasonably important | 5 | (8.8) | 8 | (11.6) | 13 | (10.3) | |
| Slightly important | 0 | (0.0) | 1 | (1.4) | 1 | (0.8) | |
| Not important | 0 | (0.0) | (0.0) | (0.0) | 0 | (0.0) | |
| Wilcoxon’s W (p-value) | 1939.0 (p = 0.882) | ||||||
| The application is free | |||||||
| Very important | 39 | (68.4) | 49 | (71.0) | 88 | (69.8) | |
| Important | 16 | (28.1) | 16 | (23.2) | 32 | (25.4) | |
| Reasonably important | 2 | (3.5) | 4 | (5.8) | 6 | (4.8) | |
| Slightly important | 0 | (0.0) | 0 | (0.0) | 0 | (0.0) | |
| Not important | 0 | (0.0) | 0 | (0.0) | 0 | (0.0) | |
| Wilcoxon’s W (p-value) | 1931.5.0 (p = 0.833) | ||||||
| Ease of use by patient | |||||||
| Very important | 46 | (80.7) | 54 | (78.3) | 100 | (79.4) | |
| Important | 11 | (19.3) | 12 | (17.4) | 23 | (18.3) | |
| Reasonably important | 0 | (0.0) | 3 | (4.3) | 3 | (2.4) | |
| Slightly important | 0 | (0.0) | 0 | (0.0) | 0 | (0.0) | |
| Not important | 0 | (0.0) | 0 | (0.0) | 0 | (0.0) | |
| Wilcoxon’s W (p-value) | 2031.0 (p = 0.655) | ||||||
| Patient access to hardware | |||||||
| Very important | 40 | (70.2) | 51 | (73.9) | 91 | (72.2) | |
| Important | 15 | (26.3) | 16 | (23.2) | 31 | (24.6) | |
| Reasonably important | 2 | (3.5) | 2 | (2.9) | 4 | (3.2) | |
| Slightly important | 0 | (0.0) | 0 | (0.0) | 0 | (0.0) | |
| Not important | 0 | (0.0) | 0 | (0.0) | 0 | (0.0) | |
| Wilcoxon’s W (p-value) | 1892.0 (p = 0.642) | ||||||
| Patient access to the internet | |||||||
| Very important | 40 | (70.2) | 46 | (66.7) | 86 | (68.3) | |
| Important | 17 | (29.8) | 17 | (24.6) | 34 | (27.0) | |
| Reasonably important | 0 | (0.0) | 6 | (8.7) | 6 | (4.8) | |
| Slightly important | 0 | (0.0) | 0 | (0.0) | 0 | (0.0) | |
| Not important | 0 | (0.0) | 0 | (0.0) | 0 | (0.0) | |
| Wilcoxon’s W (p-value) | 2086.5 (p = 0.472) | ||||||
| Patient’s health literacy | |||||||
| Very important | 25 | (43.9) | 33 | (47.8) | 58 | (46.0) | |
| Important | 24 | (42.1) | 22 | (31.9) | 46 | (36.5) | |
| Reasonably important | 8 | (14.0) | 12 | (17.4) | 20 | (15.9) | |
| Slightly important | 0 | (0.0) | 1 | (1.4) | 1 | (0.8) | |
| Not important | 0 | (0.0) | 1 | (1.4) | 1 | (0.8) | |
| Wilcoxon’s W (p-value) | 1976.5 (p = 0.960) | ||||||
| Patient’s digital literacy | |||||||
| Very important | 30 | (52.6) | 31 | (44.9) | 61 | (48.4) | |
| Important | 24 | (42.1) | 32 | (46.4) | 56 | (44.4) | |
| Reasonably important | 3 | (5.3) | 6 | (8.7) | 9 | (7.1) | |
| Slightly important | 0 | (0.0) | 0 | (0.0) | 0 | (0.0) | |
| Not important | 0 | (0.0) | 0 | (0.0) | 0 | (0.0) | |
| Wilcoxon’s W (p-value) | 2142.0 (p = 0.337) | ||||||
| Patient’s interest in applications | |||||||
| Very important | 34 | (59.6) | 35 | (50.7) | 69 | (54.8) | |
| Important | 19 | (33.3) | 28 | (40.6) | 47 | (37.3) | |
| Reasonably important | 4 | (7.0) | 5 | (7.2) | 9 | (7.1) | |
| Slightly important | 0 | (0.0) | 0 | (0.0) | 0 | (0.0) | |
| Not important | 0 | (0.0) | 1 | (1.4) | 1 | (0.8) | |
| Wilcoxon’s W (p-value) | 2145.0 (p = 0.3243) | ||||||
| Proven scientific validity | |||||||
| Very important | 35 | (61.4) | 58 | (84.1) | 93 | (73.8) | |
| Important | 14 | (24.6) | 6 | (8.7) | 20 | (15.9) | |
| Reasonably important | 7 | (12.3) | 4 | (5.8) | 11 | (8.7) | |
| Slightly important | 1 | (1.8) | 1 | (1.4) | 2 | (1.6) | |
| Not important | 0 | (0.0) | 0 | (0.0) | 0 | (0.0) | |
| Wilcoxon’s W (p-value) | 1533.5 (p = 0.006) | ||||||
| Online data security | |||||||
| Very important | 38 | (66.7) | 53 | (76.8) | 91 | (72.2) | |
| Important | 10 | (17.5) | 6 | (8.7) | 16 | (12.7) | |
| Reasonably important | 7 | (12.3) | 9 | (13.0) | 16 | (12.7) | |
| Slightly important | 2 | (3.5) | 1 | (1.4) | 3 | (2.4) | |
| Not important | 0 | (0.0) | 0 | (0.0) | 0 | (0.0) | |
| Wilcoxon’s W (p-value) | 1784.5 (p = 0.2583) | ||||||
| Ease of access/extraction of data by the doctor | |||||||
| Very important | 25 | (43.9) | 32 | (46.4) | 57 | (45.2) | |
| Important | 16 | (28.1) | 21 | (30.4) | 37 | (29.4) | |
| Reasonably important | 10 | (17.5) | 11 | (15.9) | 21 | (16.7) | |
| Slightly important | 4 | (7.0) | 4 | (5.8) | 8 | (6.3) | |
| Not important | 2 | (3.5) | 1 | (1.4) | 3 | (2.4) | |
| Wilcoxon’s W (p-value) | 1867.0 (p = 0.6044) | ||||||
Abbreviations: n: Number. p-value in bold; p<0.05.
Barriers
Several barriers to recommending apps were also identified (Table 7). The patient’s low digital literacy (87.3%), patient interest in apps (85.7%) and a lack of patient hardware access (81.8%) were cited as ‘important’ and ‘very important’, alongside a lack of patient internet access (81.0%) and low health literacy (78.6%). Ratings for all the barriers to recommendation did not differ significantly between those who recommend digital apps and those who do not.
Table 7.
Barriers that influence the recommendation of applications.
| GPs recommending apps (n = 57) |
GPs not recommending apps (n = 69) |
Total (n = 126) |
||||
|---|---|---|---|---|---|---|
| Barriers to recommendation | n | (%) | n | (%) | n | (%) |
| Lack of patient’s internet access | ||||||
| Very important | 30 | (52.6) | 40 | (58.0) | 70 | (55.6) |
| Important | 17 | (29.8) | 15 | (21.7) | 32 | (25.4) |
| Reasonably important | 9 | (15.8) | 6 | (8.7) | 15 | (11.9) |
| Slightly important | 1 | (1.8) | 7 | (10.1) | 8 | (6.3) |
| Not important | 0 | (0.0) | 1 | (1.4) | 1 | (0.8) |
| Wilcoxon’s W (p-value) | 1939.0 (p = 0.883) | |||||
| Lack of patient’s hardware access | ||||||
| Very important | 30 | (52.6) | 40 | (58.0) | 70 | (55.6) |
| Important | 17 | (29.8) | 15 | (23.2) | 32 | (25.2) |
| Reasonably important | 7 | (12.3) | 7 | (10.1) | 14 | (11.1) |
| Slightly important | 3 | (5.3) | 5 | (7.2) | 8 | (6.3) |
| Not important | 0 | (0.0) | 1 | (1.4) | 1 | (0.8) |
| Wilcoxon’s W (p-value) | 1904.0 (p = 0.736) | |||||
| Patient’s interest in applications | ||||||
| Very important | 26 | (45.6) | 38 | (55.1) | 64 | (50.8) |
| Important | 22 | (38.6) | 22 | (31.9) | 44 | (34.9) |
| Reasonably important | 8 | (14.0) | 6 | (8.7) | 14 | (11.1) |
| Slightly important | 0 | (0.0) | 3 | (4.3) | 3 | (2.4) |
| Not important | 1 | (1.8) | 0 | (0.0) | 1 | (0.8) |
| Wilcoxon’s W (p-value) | 1788.0 (p = 0.337) | |||||
| Patient’s low health literacy | ||||||
| Very important | 27 | (47.4) | 29 | (42.0) | 55 | (43.7) |
| Important | 19 | (33.3) | 25 | (36.2) | 44 | (34.9) |
| Reasonably important | 9 | (47.4) | 11 | (15.9) | 20 | (15.9) |
| Slightly important | 1 | (1.8) | 4 | (5.8) | 5 | (4.0) |
| Not important | 1 | (1.8) | 0 | (0.0) | 1 | (0.8) |
| Wilcoxon’s W (p-value) | 2081.5 (p = 0.546) | |||||
| Patient’s low digital literacy | ||||||
| Very important | 30 | (52.6) | 37 | (53.6) | 67 | (53.2) |
| Important | 18 | (34.6) | 25 | (36.2) | 43 | (34.1) |
| Reasonably important | 8 | (14.0) | 5 | (7.2) | 13 | (10.3) |
| Slightly important | 0 | (0.0) | 2 | (2.9) | 2 | (1.6) |
| Not important | 1 | (1.8) | 0 | (0.0) | 1 | (0.8) |
| Wilcoxon’s W (p-value) | 1902.0 (p = 0.727) | |||||
| Concern about data confidentiality | ||||||
| Very important | 12 | (21.1) | 21 | (30.4) | 33 | (26.2) |
| Important | 16 | (28.1) | 17 | (24.6) | 33 | (26.2) |
| Reasonably important | 16 | (28.1) | 17 | (24.6) | 33 | (26.2) |
| Slightly important | 9 | (15.8) | 10 | (14.5) | 19 | (15.1) |
| Not important | 4 | (7.0) | 4 | (5.8) | 3 | (6.3) |
| Wilcoxon’s W (p-value) | 1783.0 (p = 0.356) | |||||
| Possibility of patient’s wrong self-management/self-medication | ||||||
| Very important | 16 | (28.1) | 19 | (27.5) | 35 | (27.6) |
| Important | 21 | (36.8) | 25 | (36.2) | 46 | (36.5) |
| Reasonably important | 13 | (22.8) | 19 | (27.5) | 32 | (25.4) |
| Slightly important | 5 | (8.8) | 3 | (4.3) | 8 | (6.3) |
| Not important | 2 | (3.5) | 3 | (4.3) | 5 | (4.0) |
| Wilcoxon’s W (p-value) | 1966.5 (p = 1.0) | |||||
| Lack of doctor’s training in application recommendation | ||||||
| Very important | 26 | (45.6) | 36 | (52.2) | 62 | (49.2) |
| Important | 19 | (33.3) | 17 | (24.6) | 36 | (28.6) |
| Reasonably important | 9 | (15.8) | 12 | (17.4) | 21 | (16.7) |
| Slightly important | 2 | (3.5) | 3 | (4.3) | 5 | (4.0) |
| Not important | 1 | (1.8) | 1 | (1.4) | 2 | (1.6) |
| Wilcoxon’s W (p-value) | 1887.0 (p = 0.675) | |||||
Abbreviations: n: Number. Data were missing for 1 response.
Discussion
Principal findings
Almost half of GPs surveyed in Portugal currently used digital health apps in clinical practice, with the most recommended apps focusing on physical activity, nutrition and chronic disease management. Doctors who use health and fitness apps are more likely to recommend them to patients. GPs who do not recommend apps report lack of scientific evidence as the most important facilitator behind their decision to recommend. No other significant differences in facilitator or barrier ratings were found between the two groups. Overall, the main factors influencing GPs’ recommendations for these apps are ease of use, patient access to free apps and compatible hardware. Additionally, a GP’s knowledge of and interest in digital health significantly influences their recommendation decisions. However, key barriers to recommendation include patients’ low digital literacy, lack of interest and limited access to necessary technology, such as hardware or internet connectivity.
Comparison with existing literature
The findings of this study align with but also challenge patterns observed in existing literature on the recommendation of digital health applications by family physicians. Previous research reports prescription rates ranging from 10% to 50% [19,20,25], consistent with the 45% observed in this study. Notably, Wanger et al. highlight that urban practice settings have a positive impact on prescribing behaviour, as highlighted by Wanger et al. [20]. However, our findings suggest a small but non-significant increase in the likelihood of recommendation among rural physicians. While no definitive conclusions can be drawn, environmental and infrastructure factors likely play a role in digital health adoption.
Demographics and experience reveal additional variations. While Della Vecchia et al. reported that men were likelier to prescribe health applications, this study concludes that there is no strong evidence to suggest a significant difference between male and female physicians [25]. While prior research indicates that younger physicians are more likely to recommend, our findings reveal a small positive association between recommendation and age [19,20,25]. While this association is weak and inconclusive, it is likely influenced by the low variance in age among participants rather than a true lack of association. This highlights the need for further investigation with larger sample sizes to better understand this relationship. These discrepancies may reflect contextual differences that warrant exploration in future research.
Personal use of digital technologies is another key factor influencing recommendation behaviour. Evidence demonstrates that personal use of apps and electronic health records enhances digital maturity and increases the likelihood of prescribing such tools [25,26]. Our results align with these conclusions, reinforcing the importance of personal digital engagement in driving clinical adoption.
Prior literature suggests that the most common applications used by patients are for mental health, mindfulness, nutrition, physical activity, women’s health, and chronic disease management [19,27,28]. While physicians in this study generally recommended apps one to four times per month, slightly less frequently than the five times per month reported in Catalonia, the similarity in app types may reflect global trends, the suitability of apps for these health areas, and their availability in the market [27].
Key factors influencing adoption remain consistent across studies. Ease of use, scientific validity, and data security emerged as critical determinants in our research, aligning with previous studies and the results of a systematic review conducted by Gagnon et al. [29]. These factors, along with app cost, design, and approval by reputable health entities, shape practices worldwide [19,20,22,25,30].
Physicians broadly view health applications as valuable for improving therapy adherence, supporting chronic disease management, and promoting health education [5]. While opinions on their transformative potential vary, as noted by Sarradon-Eck et al. this study revealed a cautious optimism among clinicians [20,22]. Most participants recognised the long-term potential of apps to reduce the number of face-to-face consultations, with some reservations about immediate impacts. The robust evaluation of these health applications should, therefore, be encouraged, as this is essential to reassure clinicians of their effectiveness and to cultivate greater confidence in their integration into primary care. Moreover, future research is required to examine studies across different counties, providing a more reliable understanding of behaviours during this critical period of digitalisation in clinical practice.
Strengths and limitations
To our knowledge, this is the first study to characterise the recommendation of digital health applications by Portuguese GPs, filling a critical gap in the literature. Dissemination was facilitated through collaboration with the Portuguese Association of General and Family Medicine (APMGF), a trusted organisation among GPs. Leveraging APMGF’s reach and credibility allowed the survey to target a diverse population of GPs practising across different regions and clinical settings in Portugal.
However, several limitations must be acknowledged. Convenience sampling was used, and a priori power calculation was not conducted, due to the exploratory nature and feasibility constraints of the study. These factors may affect the robustness of our findings, which should be interpreted with appropriate caution. A limitation of the study is the lack of a theoretical framework, such as the Theoretical Domains Framework, to guide the questionnaire design [31]. Integrating such a framework could have provided a more structured approach to capturing a broader range of digital health app recommendation determinants that may impact recommendation behaviour. Disseminating the survey via online platforms also introduces the potential for selection bias, as respondents who are frequent users of online technologies and digital applications may have been more likely to participate. Consequently, it is possible that the respondents have a personal interest in digital health and may have been disproportionately represented in our survey, potentially skewing the results. The survey did not differentiate between apps freely available for download and those officially approved by authorities, which could have provided additional insight into recommendation behaviour.
Furthermore, this national survey could not capture the local contextual factors, such as cultural and social influences, that could impact recommendation behaviour. The exclusion of these variables limits our ability to analyse all determinants comprehensively. Another limitation of this study is that only univariate analysis was conducted; future research would benefit from multivariate approaches to account for potential confounding factors and better isolate the independent effects of variables influencing recommendation behaviours. Plus, subsequent investigations should consider incorporating a broader range of contextual factors, such as reimbursement policies and recommendation durations, while using more representative sampling techniques to enhance the robustness and applicability of findings.
Conclusion
Nearly half of surveyed GPs in Portugal recommend health apps to their patients, with most acknowledging these tools’ significant role in promoting self-management. Doctors who personally use health and well-being apps are more likely to suggest them to their patients. Many GPs believe that health apps have the potential to revolutionise clinical practice in the long term by reducing the need for face-to-face consultations. However, since scientific evidence emerged as a key factor for GPs who do not currently recommend apps, further efforts are required to ensure these apps undergo a comprehensive evaluation to ensure their validity. Improving transparency and access to evidence may help encourage greater adoption among this group. Additionally, to fully embrace the digital transformation of primary care, it is essential to prioritise improving digital literacy and integrating user-centred design in the development of these technologies.
Supplementary Material
Acknowledgments
The authors acknowledge the Portuguese Association of General and Family Medicine Research Department for disseminating the study and questionnaire on its communication channels.
Funding Statement
ELE and ALN are supported by the Imperial College National Institute for Health Research (NIHR) Patient Safety Translational Research Centre (PSTRC), with infrastructure support from Imperial NIHR Biomedical Research Centre. ALN is additionally supported by the Northwest London National Institute for Health and Care Research Applied Research Collaboration (NWL NIHR ARC).
Disclosure statement
The authors have no conflicts of interest to declare.
References
- 1.World Health Orgnization . mHealth: new horizons for health through mobile technologies: second global survey on eHealth [Internet]. Switzerland: WHO Press; 2011. [Google Scholar]
- 2.Statista . 2023. Available from: https://www.statista.com/statistics/779910/health-apps-available-ios-worldwide/
- 3.Maaß L, Freye M, Pan C-C, et al. The definitions of health apps and medical apps from the perspective of public health and law: qualitative analysis of an interdisciplinary literature overview. JMIR Mhealth Uhealth. 2022;10(10):e37980. doi: 10.2196/37980. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Library NZHA. 2023. Available from: https://www.healthnavigator.org.nz/apps/.
- 5.Wangler J, Jansky M. Digital health applications in primary care—Experiences and observations of general practitioners with regard to the use of DiGA. Pravention Gesundheitsforderung; 2022. [Google Scholar]
- 6.RACoG P. What is the HANDI project? [Internet]. Available from: https://www.racgp.org.au/clinical-resources/clinical-guidelines/handi/about-handi/about-handi
- 7.SNd S. App SNS 24 [Internet]. Available from: https://www.sns24.gov.pt/guia/app-sns-24/#o-que-e-a-app-sns-24
- 8.Heidel A, Hagist C.. Potential benefits and risks resulting from the introduction of health apps and wearables into the german statutory health care system: scoping review. JMIR Mhealth Uhealth. 2020;8(9):e16444. doi: 10.2196/16444. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Vo V, Auroy L, Sarradon-Eck A.. Perceptions of mHealth apps: meta-ethnographic review of qualitative studies. JMIR Mhealth Uhealth. 2019;7(7):e13817. doi: 10.2196/13817. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Versteegh LA, Chang AB, Chirgwin S, et al. Multi-lingual “Asthma APP” improves health knowledge of asthma among Australian First Nations carers of children with asthma. Front Pediatr. 2022;10:925189. doi: 10.3389/fped.2022.925189. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Harada N, Harada S, Ito J, et al. Mobile health app for japanese adult patients with asthma: clinical observational study. J Med Internet Res. 2020;22(8):e19006. doi: 10.2196/19006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Fleming GA, Petrie JR, Bergenstal RM, et al. Diabetes digital app technology: benefits, challenges, and recommendations. A consensus report by the european association for the study of diabetes (EASD) and the American Diabetes Association (ADA). Diabetes Technol Work Group. Diabetes Care. 2020;43(1):250–260. doi: 10.2337/dci19-0062. [DOI] [PubMed] [Google Scholar]
- 13.Doyle-Delgado K, Chamberlain JJ.. Use of diabetes-related applications and digital health tools by people with diabetes and their health care providers. Clin Diabetes. 2020;38(5):449–461. doi: 10.2337/cd20-0046. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Sin J, Galeazzi G, McGregor E, et al. Digital interventions for screening and treating common mental disorders or symptoms of common mental illness in adults: systematic review and meta-analysis. J Med Internet Res. 2020;22(9):e20581. doi: 10.2196/20581. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Afshin A, Babalola D, McLean M, et al. Information technology and lifestyle: a systematic evaluation of internet and mobile interventions for improving diet, physical activity, obesity, tobacco, and alcohol use. J Am Heart Assoc. 2016;5(9):e003058. doi: 10.1161/JAHA.115.003058. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Mair JL, Salamanca-Sanabria A, Augsburger M, et al. Effective behavior change techniques in digital health interventions for the prevention or management of noncommunicable diseases: an umbrella review. Ann Behav Med. 2023;57(10):817–835. doi: 10.1093/abm/kaad041. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Schoeppe S, Alley S, Lippevelde W, et al. Efficacy of interventions that use apps to improve diet, physical activity and sedentary behaviour: a systematic review. Int J Behav Nutr Phys Act. 2016;13(1):127. doi: 10.1186/s12966-016-0454-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Mosa ASM, Yoo I, Sheets L.. A systematic review of healthcare applications for smartphones. BMC Med Inform Decis Mak. 2012;12(1):67. doi: 10.1186/1472-6947-12-67. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Byambasuren O, Beller E, Glasziou P.. Current knowledge and adoption of mobile health apps among Australian general practitioners: survey study. JMIR Mhealth Uhealth. 2019;7(6):e13199. doi: 10.2196/13199. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Wangler J, Jansky M.. The use of health apps in primary care—results from a survey amongst general practitioners in Germany. Wien Med Wochenschr. 2021;171(7-8):148–156. doi: 10.1007/s10354-021-00814-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.May S, Seifert F, Bruch D, et al. Insights into how mhealth applications could be introduced into standard hypertension care in germany: qualitative study with German cardiologists and general practitioners. JMIR Mhealth Uhealth. 2025;13:e56666. doi: 10.2196/56666. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Sarradon-Eck A, Bouchez T, Auroy L, et al. Attitudes of general practitioners toward prescription of mobile health apps: qualitative study. JMIR Mhealth Uhealth. 2021;9(3):e21795. doi: 10.2196/21795. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Elm E V, Altman DG, Egger M, et al. Strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. BMJ. 2007;335(7624):806–808. doi: 10.1136/bmj.39335.541782.AD. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Fialho AS, Oliveira MD, Sá AB.. Using discrete event simulation to compare the performance of family health unit and primary health care centre organizational models in Portugal. BMC Health Serv Res. 2011;11(1):274. doi: 10.1186/1472-6963-11-274. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Dv C, T L, C B, et al. Willingness of french general practitioners to prescribe mhealth apps and devices: quantitative study. JMIR MHealth UHealth. 2022;10(2):962924. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Teixeira F, Li E, Laranjo L, et al. Digital maturity and its determinants in General Practice: a cross-sectional study in 20 countries. Front Public Health. 2023;10:962924. doi: 10.3389/fpubh.2022.962924. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Ls F, Pb C, Ag N, et al. The prescription of mobile apps by primary care teams: a pilot project in catalonia. JMIR Mhealth Uhealth. 2018;6(6):e10701. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Peeters JM, Krijgsman JW, Brabers AE, et al. Use and uptake of ehealth in general practice: a cross-sectional survey and focus group study among health care users and general practitioners. JMIR Med Inform. 2016;4(2):e11. doi: 10.2196/medinform.4515. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Gagnon MP, Ngangue P, Payne-Gagnon J, et al. m-Health adoption by healthcare professionals: a systematic review. J Am Med Inform Assoc. 2016;23(1):212–220. doi: 10.1093/jamia/ocv052. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Zhang Y, Koch S.. Mobile health apps in Sweden: what do physicians recommend? Stud Health Technol Inf. 2015;210:793–797. [PubMed] [Google Scholar]
- 31.Dyson J, Cowdell F.. How is the Theoretical Domains Framework applied in designing interventions to support healthcare practitioner behaviour change? A systematic review. Int J Qual Health Care J Int Soc Qual Health Care. 2021;33(3):mzab106. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.