This study assessed the effects on type 2 diabetes self-management education provided in group courses with or without a supporting smartphone application (the DM2CUA app).
This open-label, cluster-randomized, controlled, multicenter pilot study involved three Austrian diabetes educational group courses. People with type 2 diabetes in the control group received a regular educational group course, whereas those in the intervention group received the same course plus the use of the DM2CUA app. The app prompted participants to carry out educational tasks that were discussed in the course. After the last lesson, the app provided participants with relevant messages for another 4 weeks. The primary outcome measure was the Diabetes Self-Management Questionnaire score assessed at four time points. Secondary outcome measures included scores on the Diabetes Distress Scale and the Health Education Impact Questionnaire and A1C levels.
Participants in the intervention group already had a higher level of diabetes self-management at the start, but the median score showed further improvement during the entire study period.
Findings from this pilot study suggest that the DM2CUA app may have a positive impact on diabetes self-management.
Type 2 diabetes is one of the most common chronic metabolic diseases with serious sequelae that pose major health policy challenges (1). In Austria, health expenditures for people with type 2 diabetes are approximately €4,000 more per year than for the average person, depending on the course of therapy and long-term consequences (2). The current prevalence of type 2 diabetes is 4.6% in Austria and is estimated to increase to 5.3% by 2030 (2,3).
In March 2017, the Austrian Federal Ministry of Health and Women published the Austrian Diabetes Strategy to counter structurally and strategically the problems that caring for people with type 2 diabetes already pose for the public health system. In particular, this strategy calls for “low-threshold, target group–specific, continuous” and adequate care services for people with type 2 diabetes, enabling them to deal with their diabetes as independently and competently as possible (4–6). Independence in therapy is particularly important in view of the increasing prevalence rates and burdens on the health care system. Current mobile health (mHealth) technologies offer the possibility to implement these requirements effectively and efficiently (7–9).
Better health competence is associated with better long-term glucose levels (i.e., A1C levels) (10,11). Thus, patient education offered in group training courses, providing adequate diabetes information and skills, plays a crucial role in the diabetes treatment and prevention process (4,6,12,13). Facilitating behavioral changes, increased motivation, and greater awareness of diabetes-related topics such as nutrition and exercise are some of the main goals of these courses (6).
Although it is known that active participation in these courses promotes individuals’ adherence to therapy (14–16), the primary didactic method implemented by health care professionals (HCPs) such as nurses, dietitians, and doctors is frontal instruction, while interactive participation is secondary (17). Still, several studies showed positive effects of patient education training on A1C levels, self-care, self-efficacy, and self-management (6,12,18,19). In addition, patient training positively influences health economics by reducing associated costs (5). However, these training programs do not address patients’ differing learning styles and do not seem to sustain behavioral changes (20,21).
Current mHealth technologies offer the possibility of implementing further training and information initiatives to help patients develop improved and more sustainable diabetes self-management abilities. Qualitative studies evaluating the use of smartphone apps in diabetes self-management have shown improved type 2 diabetes self-management and health (7,8). However, the sustainability of such app use (9) and the most effective use of mHealth technologies in patient education is unclear.
This study involved the use of a smartphone app called DM2CUA (an abbreviation of a German title that translates to “Diabetes Mellitus Type 2 Clever Support in Daily Life” in English). The app was developed to support group training courses by providing diabetes-relevant tasks during the training period and delivering diabetes-relevant tips over a 4-week period after the final course session. This article reports the results of a pilot study evaluating whether use of the DM2CUA app increases the effects of the training courses, especially with regard to promoting sustainable changes in diabetes self-management and adherence to the diabetes treatment plan.
Research Design and Methods
Study Design and Setting
This open-label, cluster-randomized, controlled, multicenter pilot study was conducted at three hospitals in the federal state of Salzburg, Austria, between March 2019 and July 2023. In nurse- and dietitian-led outpatient educational group courses (OEGCs), patients were educated about relevant topics (e.g., the basics of type 2 diabetes, nutrition, physical activity, secondary diseases, medications, foot care, and psychological and legal issues) in four to five lessons provided over a period of 4–5 weeks (22).
People with type 2 diabetes in the control group (CG) received a regular OEGC, whereas those in the intervention group (IG) received the regular OEGC plus use of the DM2CUA app. Figure 1 depicts the complete study design.
Study design and outcome measures in the respective phases from T0 (baseline) through T1 (immediately after the group course) and T2 (4 weeks after group course) to T3 (8 weeks after group course).
Study design and outcome measures in the respective phases from T0 (baseline) through T1 (immediately after the group course) and T2 (4 weeks after group course) to T3 (8 weeks after group course).
Ethical approval for this study was granted by the Province Ethics Committee of Salzburg (approved 24 January 2019 [reference 415-E/2438/5-2019]). The study was registered in the ISRCTN registry (ISRCTN-Nr. 29647256).
Recruitment, Randomization, and Blinding
Each OEGC was cluster-randomized as an IG or CG in advance, and each study center was allocated an equal number of IGs and CGs. The nature of the intervention allowed neither blinding nor randomization within an OEGC. Participation was voluntary, and each participant signed an informed consent form.
Participants and Sample Size
Eligible participants had to be diagnosed with type 2 diabetes and engaged in an OECG at a designated study center (Figure 2). Additional inclusion criteria were age ≥18 years, proficient in smartphone use, and capable of comprehending instructions provided within the app in German. Exclusion criteria included a diagnosis of type 1 diabetes or gestational diabetes mellitus and visual impairments adversely affecting smartphone use.
Type 2 diabetes group courses in the federal state of Salzburg. This map depicts the dimensions of Salzburg, with the degree of urbanization as defined by the European Commission (44) indicated. The locations and types of institutions that offered type 2 diabetes educational group courses during the study conception phase (2018), as well as the three study centers and the locations of the community type 2 diabetes OEGCs that HCPs offered for the first time during the recruitment phase (first half year of 2019) are shown.
Type 2 diabetes group courses in the federal state of Salzburg. This map depicts the dimensions of Salzburg, with the degree of urbanization as defined by the European Commission (44) indicated. The locations and types of institutions that offered type 2 diabetes educational group courses during the study conception phase (2018), as well as the three study centers and the locations of the community type 2 diabetes OEGCs that HCPs offered for the first time during the recruitment phase (first half year of 2019) are shown.
As described further below, Diabetes Self-Management Questionnaire (DSMQ) score was the primary outcome. Sample size calculation used the difference in respective DSMQ scores between patients with an A1C ≥9.0% against patients with an A1C ≤7.6% and ≤8.9% (15). With an α of 0.05, power at 0.8, and an expected dropout rate of 10%, the targeted number of participants was 73. The actual sample size was 15 participants. This discrepancy was the result of an unforeseeable major obstacle: the main public health insurance provider’s launch of type 2 diabetes OEGCs on the community level resulted in a substantially decreased number of courses in the three hospitals enrolled in this study. Details are provided in the first section of the Supplementary Material.
Intervention
The two-phase intervention program was developed based on a review of the relevant literature, from observations of eight regional OEGCs (five sessions per OEGC were attended as nonparticipating observers and notes were taken) (22), and five subsequent focus groups with people with type 2 diabetes, nurses with special training in diabetes care, dietitians, and diabetologists (22), with iterative feedback from HCPs from the study centers. The sessions for the CG and the IG were held at different time points because recruitment for the IG took longer.
Phase 1: supporting the group course
During the OEGCs, the DM2CUA app prompted IG participants to engage anonymously in six educational tasks: 1) documenting refrigerator contents and grocery shopping (by taking pictures), 2) keeping a nutritional photo diary (taking multiple pictures during 4 days), 3) assessing eating habits (reflecting on the influence of stress and emotions on eating behavior and multiple completions of a three-item survey over 4 days), 4) evaluating how exercise affects blood glucose (measuring blood glucose before and after an exercise), 5) setting an exercise goal (formulating and reviewing a goal for the week), and 6) addressing foot care and shoes (taking pictures of shoes).
All tasks were performed within the DM2CUA app, and the data collected were visualized via the DM2CUA website in the subsequent lesson and discussed with the respective HCPs. The tasks were timed in such a manner that they had to be completed either in preparation for or as follow-up of specific lesson topics as scheduled by the OEGC in the respective study center. More details about these tasks are provided in the second section of the Supplementary Material.
The three-part rationale behind these tasks was to 1) use the advantages of smartphones in a manner that allowed easy integration of patients’ individual living environment and associated challenges of managing type 2 diabetes, 2) make use of the time between lessons for a more intense level of engagement, and 3) support established OEGCs in shaping a more discursive environment in the lessons (by anonymously displaying the results of tasks using the auxiliary DM2CUA website) and address individual challenges of type 2 diabetes in a manner detached from specific patients.
Phase 2: after the group course
After the OEGC, participants received 42 messages on diabetes-relevant topics randomly distributed over 4 weeks, with a maximum of five messages per day. These messages related to different topics, including exercise (n = 13), nutrition (n = 10), self-management (n = 9), routine diabetes-related follow-up visits (n = 6), and foot care (n = 4). Exemplary messages are provided in the third section of the Supplementary Material. The rationale behind these messages was to reinforce and stabilize knowledge about and implementation of diabetes-specific health-promoting activities in daily life.
Technical Implementation
The DM2CUA app and website were developed at the Salzburg University of Applied Sciences and hosted on proprietary servers. The DM2CUA app was developed for the Android operating system and installed on the proprietary smartphones of the participants.
Rental devices were distributed to participants who rejected the installation on their own device or who owned smartphones with another operating system. Additionally, a DM2CUA website was provided, which allowed a separate login for each study center. All questionnaire data were collected via the DM2CUA app for both groups.
Data Collection and Outcome Measures
Outcome measures were collected for both groups at predefined time points: baseline (T0; the day of the first lesson), T1 (the day of the last lesson), T2 (4 weeks after the last lesson), and T3 (8 weeks after the last lesson) (Figure 1). Demographic and clinical data were gathered at baseline.
DSMQ scores at four time points (T0, T1, T2, and T3) were the primary outcome. The DSMQ is a self-reported questionnaire for people with type 2 diabetes assessing diabetes self-care activities associated with glycemic control (4,5). The questionnaire was applied in the German version and adapted to the 4-week period between the predefined time points (4).
Secondary outcomes included the Diabetes Distress Scale (DDS) (at T0, T1, T2, and T3), the Health Education Impact Questionnaire (heiQ) (at T0 and T1) (23,24), and A1C values (at T0, T1, T2, and T3). The DDS is a self-reported questionnaire for people with type 2 diabetes that measures diabetes distress (15,16,25,26), with scores ≥3 indicating clinically meaningful distress (25). The questionnaire was applied in the German version (26) and adapted to the 4-week period between the predefined time points. The heiQ is a self-reported survey that comprehensively evaluates patient education programs, with greater values indicating desirable outcomes (except for the emotional distress subscale) (23). The heiQ was applied in the German version (24). A1C is an objective outcome measure of patients’ glycemic control. Fresh samples of capillary blood from the fingertip were analyzed at the point of care by a biomedical analyst using the Eurolyser A1C test kit in combination with Eurolyser CUBE/Smart (Eurolyser Diagnostica GmbH, Salzburg, Austria), according to the manufacturer’s protocol.
Additionally, the IG gave feedback on the received messages in the second (post-group) phase of the intervention and rated the usability of the DM2CUA app via the System Usability Score (SUS) (15) at T3. Questions included in the SUS are provided in the fourth section of the Supplementary Material. We used the German version of the original English questionnaire and adapted the questions by replacing the word “system” with the word “app.” Finally, the researchers were verbally debriefed by the two HCPs involved in the IG (nurse and dietitian).
The participation rate for outcome measurements was between 75 and 100% for most readouts.
Statistical Analyses
Because of the limited number of enrolled participants (n = 15), none of the planned statistical tests could be performed. The following results are therefore purely descriptive.
Data Availability
Data are available in German language upon request. The authors confirm that all patient/personal identifiers have been removed or disguised so the patient/person(s) described are not identifiable and cannot be identified through the details.
Results
Demographics
The characteristics of the CG and IG participants are shown in Table 1.
Characteristics of CG and IG Participants
| Characteristic | CG (n = 7) | IG (n = 8) |
|---|---|---|
| Sex Female Male | 2 5 | 3 5 |
| Age, years | 58 (42–67) | 56 (32–62) |
| Diabetes duration, years | 0 (0–4) | 0.5 (0–3) |
| Drug treatment None Oral antidiabetic agents Insulin | 1 6 0 | 0 8 0 |
| Blood glucose measurements per week | 7 (3–24) | 6 (3–10) |
| Education group courses attended prior to the study No Yes | 4 3 | 5 3 |
| Part of Therapy Active program* Yes No Don't know | 3 2 2 | 4 2 2 |
| Characteristic | CG (n = 7) | IG (n = 8) |
|---|---|---|
| Sex Female Male | 2 5 | 3 5 |
| Age, years | 58 (42–67) | 56 (32–62) |
| Diabetes duration, years | 0 (0–4) | 0.5 (0–3) |
| Drug treatment None Oral antidiabetic agents Insulin | 1 6 0 | 0 8 0 |
| Blood glucose measurements per week | 7 (3–24) | 6 (3–10) |
| Education group courses attended prior to the study No Yes | 4 3 | 5 3 |
| Part of Therapy Active program* Yes No Don't know | 3 2 2 | 4 2 2 |
Data are n or median (range).
*A type 2 diabetes care and therapy program of the Austrian health insurance provider.
Primary Outcome (DSMQ Scores)
The DSMQ was performed to assess diabetes self-management, with higher scores indicating more desirable self-management behavior (15,16). For reference, people with A1C values ≤7.5% have mean DSMQ scores of 7.7 (15). The CG started with a median DSMQ score of 7.7 (at T0). However, at T3, the median DSMQ dropped noticeably below the initial level. The IG started with higher scores (median 8.2 at T0), but the medians showed further improvement until T3 (Figure 3).
DSMQ scores from T0 to T3 in the control and intervention groups. X indicates the mean scores.
DSMQ scores from T0 to T3 in the control and intervention groups. X indicates the mean scores.
Secondary Outcomes
DDS scores
The DDS was performed to evaluate the presence of stress resulting from type 2 diabetes, with lower scores indicating less diabetes-related stress. Scores <2 indicate little distress, and those >3 indicate high distress levels (26). Both groups displayed little distress at T0 (median score 1.5). In the IG, the median DDS score improved continuously until the end of the second intervention phase (T2) and deteriorated only slightly until T3. Although the median DDS score decreased during the OEGC (from T0 to T1) in the CG, it was higher again 8 weeks later (T3) (Figure 4).
DDS scores from T0 to T3 in the control and intervention groups. X indicates the mean scores.
DDS scores from T0 to T3 in the control and intervention groups. X indicates the mean scores.
A1C levels
A1C values were used as an indicator of diabetes-related health. In patients with well-managed type 2 diabetes, A1C values are <6% (27). The CG showed continuously falling median A1C values from 7.4% at T0 to 5.8% at T3. In the IG, the median A1C value was 5.8% at T0 and remained stable throughout the study (Figure 5). This finding was the result of the longer recruiting phase for the IG, leading to already improved blood glucose management by the start of the intervention.
A1C values from T0 to T3 in the control and intervention groups. X indicates the mean values.
A1C values from T0 to T3 in the control and intervention groups. X indicates the mean values.
heiQ scores
The heiQ was performed to evaluate the impact of the OEGCs. Both groups arbitrarily displayed good, above-benchmark mean baseline values in all eight constructs. In the IG above-benchmark standardized effect sizes from T0 to T1, changes could be observed in all but one construct (emotional distress) (28). In contrast, the CG displayed below-benchmark effects in five of the eight constructs. Arbitrary group comparisons showed greater standardized effect sizes from T0 to T1 in six of eight constructs in the IG than in the CG, indicating a more positive effect of health education through the support of the DM2CUA app (Table 2).
heiQ Scores
| heiQ Construct | CG | IG | Benchmark (30) | |||||
|---|---|---|---|---|---|---|---|---|
| T0 (n = 7), mean ± SD | T1 (n = 6), mean ± SD | T0 to T1, SES | T0 (n = 8), mean ± SD | T1 (n = 7), mean ± SD | T0 to T1, SES | T0, mean | T0 to T1, SES | |
| Positive and active engagement in life | 3.17 ± 0.41 | 3.40 ± 0.61 | 0.44 | 3.20 ± 0.56 | 3.46 ± 0.40 | 0.50 | 2.94 | 0.35 |
| Health-directed behavior | 2.93 ± 0.35 | 3.08 ± 0.62 | 0.23 | 2.93 ± 0.65 | 3.21 ± 0.67 | 0.44 | 2.84 | 0.37 |
| Skill and technique acquisition | 3.39 ± 0.37 | 3.42 ± 0.45 | 0.06 | 3.11 ± 0.40 | 3.32 ± 0.32 | 0.43 | 2.84 | 0.42 |
| Constructive attitudes and approaches | 3.43 ± 0.41 | 3.60 ± 0.42 | 0.32 | 3.60 ± 0.48 | 3.80 ± 0.28 | 0.38 | 3.04 | 0.21 |
| Self-monitoring and insight | 3.43 ± 0.42 | 3.50 ± 0.47 | 0.17 | 3.26 ± 0.33 | 3.46 ± 0.28 | 0.48 | 3.03 | 0.34 |
| Health service navigation | 3.37 ± 0.69 | 3.60 ± 0.49 | 0.48 | 3.26 ± 0.60 | 3.49 ± 0.46 | 0.48 | 3.10 | 0.19 |
| Social integration and support | 3.43 ± 0.43 | 3.53 ± 0.30 | 0.17 | 3.29 ± 0.63 | 3.51 ± 0.46 | 0.38 | 2.91 | 0.19 |
| Emotional distress | 1.89 ± 0.44 | 1.93 ± 0.38 | 0.06 | 1.66 ± 0.33 | 1.91 ± 0.40 | 0.40 | 2.34 | −0.21 |
| heiQ Construct | CG | IG | Benchmark (30) | |||||
|---|---|---|---|---|---|---|---|---|
| T0 (n = 7), mean ± SD | T1 (n = 6), mean ± SD | T0 to T1, SES | T0 (n = 8), mean ± SD | T1 (n = 7), mean ± SD | T0 to T1, SES | T0, mean | T0 to T1, SES | |
| Positive and active engagement in life | 3.17 ± 0.41 | 3.40 ± 0.61 | 0.44 | 3.20 ± 0.56 | 3.46 ± 0.40 | 0.50 | 2.94 | 0.35 |
| Health-directed behavior | 2.93 ± 0.35 | 3.08 ± 0.62 | 0.23 | 2.93 ± 0.65 | 3.21 ± 0.67 | 0.44 | 2.84 | 0.37 |
| Skill and technique acquisition | 3.39 ± 0.37 | 3.42 ± 0.45 | 0.06 | 3.11 ± 0.40 | 3.32 ± 0.32 | 0.43 | 2.84 | 0.42 |
| Constructive attitudes and approaches | 3.43 ± 0.41 | 3.60 ± 0.42 | 0.32 | 3.60 ± 0.48 | 3.80 ± 0.28 | 0.38 | 3.04 | 0.21 |
| Self-monitoring and insight | 3.43 ± 0.42 | 3.50 ± 0.47 | 0.17 | 3.26 ± 0.33 | 3.46 ± 0.28 | 0.48 | 3.03 | 0.34 |
| Health service navigation | 3.37 ± 0.69 | 3.60 ± 0.49 | 0.48 | 3.26 ± 0.60 | 3.49 ± 0.46 | 0.48 | 3.10 | 0.19 |
| Social integration and support | 3.43 ± 0.43 | 3.53 ± 0.30 | 0.17 | 3.29 ± 0.63 | 3.51 ± 0.46 | 0.38 | 2.91 | 0.19 |
| Emotional distress | 1.89 ± 0.44 | 1.93 ± 0.38 | 0.06 | 1.66 ± 0.33 | 1.91 ± 0.40 | 0.40 | 2.34 | −0.21 |
SES, standard effect size.
Debriefing by HCPs
The DM2CUA app could be integrated well into the existing OEGC process regarding time and organization. The sharing of personal photos in the IG via the tasks was beneficial, resulting in more pronounced self-reflection and increased cooperation in the IG than in the CG. Initial reservations of individual study participants concerning the app were overcome as the OEGC progressed. Three of the six tasks (task 2: nutritional photo diary, task 4: exercise and blood glucose, and task 5: exercise goal) were described as very suitable, with no or minor change requests. Task 1: refrigerator contents and grocery was not as suitable. Specific change requests were elaborated for task 3: eating habits and task 6: foot care and shoes. Further OEGC topics were mentioned for potential new task contents (e.g., stress, nutritional behaviors, and mental health screening).
Feedback from the IG
The IG participants retrospectively evaluated the post-group course phase as depicted in Figure 6. As shown in the shading of the figure, the content of the messages was understood by all participants and was mostly read in full. Only one participant indicated that the messages occasionally bothered him or her. The knowledge gain from the messages and the implementation of the messages into action were rated between occasionally and always. Additional positive feedback was that the messages made participants look further into the issues (rated on average “mostly”).
Retrospective evaluation of the app messages at T2 by the intervention group (n = 6). Shading indicates the number of responses from clear (no response) to dark gray (five responses).
Retrospective evaluation of the app messages at T2 by the intervention group (n = 6). Shading indicates the number of responses from clear (no response) to dark gray (five responses).
Additionally, IG participants reported that one to five messages per day would have been suitable (median 1.5 messages/day). Four of the six participants said that more information on nutrition would have been welcome. Half of the six participants reported that they would have liked more information on the topic of exercise. Only one of the six participants would have liked more information on self-management and control appointments. None of the six desired further information on foot care or additional topics.
SUS score
The SUS was performed to test the usability of the app and to assess the quality of the app itself (i.e., handling). The mean SUS score was 68, with a score range of 58–90. According to the general interpretation, this SUS score would be considered average, indicating the participants’ indifferent attitude toward the app, meaning that the handling of the app was considered suitable for the purpose but not above average (29).
Discussion
In this pilot study, an mHeath smartphone app was developed to support existing type 2 diabetes OEGCs to improve the effects of OEGCs for participants and lecturers.
It has been proven that existing type 2 diabetes OEGCs increase the health competence of people with type 2 diabetes through interactive participation of the participants (30–32). Based on an extensive analysis of the current situation in type 2 diabetes OEGCs in the province of Salzburg (22), the DM2CUA app was designed and developed to support group discussions about everyday situations and challenges. The DM2CUA app provided participants with educational tasks and diabetes-related tips. The tasks aimed to increase participants’ self-reflection and increase their active participation during OEGCs by intensifying their examination of their own lifestyle. The tips aimed to consolidate patients’ knowledge of their disease and how to deal with it.
Limitations and Strengths
This pilot study was limited by the small number of participants, leading to a solely descriptive evaluation and hence confined transferability. The target sample size of 73 participants could not be reached because of considerable challenges in recruitment. Furthermore, changes in Austrian health care policies led to the loss of two educational group course providers, leaving only one included study center. Conclusions about the general public can therefore only be drawn from these findings to a very limited extent.
Small sample sizes are a common issue in mHealth development but are problematic for a randomized controlled study design, as requested by the ethics committee. Newer study designs such as a sequential multiple-assignment randomized trial or a multiphase optimization strategy should overcome the difficulty of the fast-paced, iterative field of mHealth app development. These study designs were not developed to evaluate the benefits and effectiveness of mHealth interventions, but rather to facilitate the development and optimization of mHealth interventions. One of the goals is to generate data from the everyday lives of people, rather than from controlled experiments. However, such study designs still struggle with managing limited resources, an increased α error, and adequate washout periods (33–36).
Despite the pilot study’s limitations and without statistical proof, the effect of the OEGCs on self-management and diabetes distress seemed to improve with the intervention. Additionally, the participation rate was rated very good, and the intervention led to a greater sustainability of the goals achieved within the group course. Therefore, despite the small sample size, a trend toward improvement in self-management could be seen.
Considering the primary outcome—DSMQ scores—this study confirmed the positive effect of OEGCs (changes in median DSMQ scores from T0 to T1) (13,37–39), independent of the study group. However, 8 weeks after OEGC (T3), the median DSMQ score in the CG dropped significantly below the initial level (T0). In the IG, participants started with higher DSMQ scores, but the medians showed further improvement during the entire study period, including the 2-month washout after the last lesson, indicating a benefit of the intensified engagement with topics via the mHealth enhancement.
Regarding the secondary outcomes, similar patterns were observed in DDS and heiQ scores, as well as in the DSMQ scores. Those results are similar to findings of other studies (40,41).
Besides the advantage that the DM2CUA app is directly integrated into a group training setting, another strength of this study was the concurrent measurement of patient-centered outcomes such as DSMQ and DDS scores and a clinically meaningful outcome (A1C) (42). The feedback from both the HCPs and the participants underlined the positive impact of the DM2CUA app in terms of supporting the outcomes of the group courses as well as increasing patients’ daily reflection on topics to improve disease self-management.
Regarding A1C levels, the IG started with lower values than the CG. First, the IG started about 2 months later than initially communicated to potential study participants. This delay led to IG participants entering the pilot study with a well-adjusted A1C value, likely resulting from greater time between registration in the study and the first measurements.
The design and usability of the DM2CUA app was rated satisfactory. However, further work to simplify the handling of the app will be undertaken for future studies to create a more positive user experience.
Conclusion
Mobile app–assisted self-care interventions are bound to become effective tools to manage blood glucose by enabling remote health management, personalized recommendations, and communication with HCPs. To assess the feasibility of diabetes mHealth apps, key evaluation variables should include A1C levels and self-management practices embedded in a professional care setting.
The DM2CUA project provided first impressions of how a smartphone app adapted and integrated into the modules of a type 2 diabetes OEGC can support both patients and HCPs. This pilot study demonstrated the potential of combined online and offline care services, not only for people with type 2 diabetes, but also potentially for those with other civilization diseases (43) for which lifestyle modification and continuous management are vital. This study lays the foundation for additional randomized controlled trials to support patient-led self-management for various chronic diseases.
Acknowledgments
The authors are grateful for the cooperation with the participating study centers and their respective study physicians, nurses, and dietitians: Gemeinnützige Oberndorfer Krankenhausbetriebsgesellschaft mbH (Manuela Hofmann, MD), Diakonissen & Wehrle Privatklinik GmbH (Raimund Weitgasser, MD), and Gemeinnützige Salzburger Landeskliniken Betriebsgesellschaft mbH (Lars Stechemesser, MD). Raimund Weitgasser additionally supported the project team in obtaining ethics committee approval. The authors also thank all of the participants in the pilot study.
Funding
This work was supported by the Federal State of Salzburg, Austria (grant number 7040-028). The founding source was not involved in the study design; the collection, analysis, or interpretation of the data; the writing of the manuscript; or the decision to submit the article for publication.
Duality of Interest
No potential conflicts of interest relevant to this article were reported.
Author Contributions
All authors made substantial contributions to 1) the conception and design of the study, acquisition of data, or analysis and interpretation of data; 2) the drafting of the article or its critical revision for important intellectual content; and 3) final approval of the manuscript for submission. M.R. participated in conceptualization, methodology, visualization, investigation, formal analysis, writing, reviewing/editing, and project administration. J.B. participated in formal analysis, investigation, writing, and reviewing/editing. M.F.-N. participated in conceptualization, methodology, investigation, writing, and project administration. C.R. participated in methodology, investigation, and reviewing/editing. M.D. participated in conceptualization, methodology, investigation, software, reviewing/editing, and project administration. G.E. participated in investigation, software, and reviewing/editing. B.G. participated in conceptualization, reviewing/editing, and funding acquisition. G.J.O. participated in conceptualization, methodology, reviewing/editing, supervision, and funding acquisition. A.S. participated in formal analysis, investigation, and reviewing/editing. M.R. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
This article contains supplementary material online at https://doi.org/10.2337/figshare.26838754.
