Living with type 1 diabetes can be challenging and is associated with health complications such as poor mental well-being and poor quality of life, and it requires lifelong self-management and support to prevent these complications [1-3]. A life with type 1 diabetes can affect patients’ ability to self-manage, increasing their risk of diabetes complications [4,5]. Often, diabetes care and self-management education and support are provided through multidisciplinary teams [6-8]. However, the traditional calendar-based approach in outpatient care is often rigid, giving patients little opportunity to influence their care. Consequently, the calendar-based approach threatens health care service resource sustainability and presents several challenges; it is not given that outpatient care is appropriate or necessary at the time of the consultation [7,9,10]. Conversely, a patient-centered approach in outpatient care, where patients themselves influence the timing and nature of treatment based on patients preferences and needs, has been suggested as a solution to these problems [11]. Furthermore, individualization through patient-centered care can strengthen the patient’s autonomy and participation in their own treatment.

The increasing focus on individualizing care aligns with an increased recognition of how digital solutions in outpatient care can provide closer interaction with health care services [8]. Such digital outpatient care may include various forms of synchronous and asynchronous remote patient monitoring; telephone and video consultations; messaging services; connection to wireless tools, such as glucose sensors and insulin pumps; or other forms of telecommunication tools [12]. Allowing and facilitating a more active patient role, patient-reported outcomes (PROs) can be integrated into digital tools for follow-up care [8]. PROs offer direct feedback from patients themselves about their health, which, among other things, provides a broader perspective on clinical goals and serves as a conversational tool [13,14]. While PROs in clinical care have gained increasing attention in recent years, a scoping review found the use of PRO as part of routine treatment in diabetes outpatient settings to be scarce [15].

Although interest in and recognition of the opportunities for digital outpatient care have been acknowledged by international panels of health care professionals in recent years [8], few studies have investigated PROs in digital outpatient care [15]. A pioneering study from Denmark used the PRO measure DiabetesFlex before scheduled consultations delivered in a digital solution to allow patients to voice their need for consultations or cancel if not deemed necessary [16]. This resulted in reduced face-to-face visits and improved well-being and diabetes distress while providing safe diabetes management. In addition, DiabetesFlex altered the interactions between clinicians and patients as it shifted toward the patients’ perspectives on living with diabetes [17]. It increased patient involvement, strengthened their sense of responsibility, and tailored diabetes care to meet individual needs [16,17].

Despite previous research emphasizing the positive contribution of digital outpatient care solutions in diabetes outpatient care, there remains a lack of research to establish an evidence base. For instance, not everyone wanted to continue with DiabetesFlex after the end of the study period, despite their positive experiences [17]. Furthermore, there is a lack of knowledge on the characteristics of patients choosing digital or flexible diabetes follow-ups. Characteristics such as sociodemographics, clinical measures, and disease-specific measures can reveal whether users and nonusers of digital outpatient care differ and whether they possess specific traits or factors. Exploring such characteristics can provide insights into the actions and choices of patients with type 1 diabetes in their outpatient care [18-20].

This study aims to investigate which selected sociodemographic and disease-related patient characteristics are associated with opting for digital outpatient care among patients with type 1 diabetes.

This study used a cross-sectional design to investigate characteristics of patients with type 1 diabetes opting for digital outpatient care compared with those opting for usual care. The data used in this study is collected as part of a larger multimethod prospective observational study assessing the integration of digital outpatient care procedures for patients with type 1 diabetes and their clinicians [21]. In this paper, we explored the baseline characteristics of these participants, and the recruitment of participants and data collection took place from October 2022 to October 2023. This study is reported according to the Strengthening the Reporting of Observational studies in Epidemiology (STROBE) reporting guideline [22].

The study was carried out at the Endocrinological Outpatient Clinic at Akershus University Hospital. In November 2021, the clinic developed and introduced a voluntary digital mobile health (mHealth) supplement outpatient care program specifically tailored to patients with type 1 diabetes called DigiDiaS.

Data were collected through 3 primary sources: self-reported questionnaires, electronic medical records, and the national quality registry for diabetes, the Norwegian Organization for Quality Improvement of Laboratory Examinations (NOKLUS). Self-reported data and the digital consent form were collected using Nettskjema [24], a secure platform that stores data in the Services for Sensitive Data (TSD) platform for confidential handling. To deliver the digital consent form and self-reported data, the patients received a secure link via SMS text message or email, requiring them to log in with their national ID for secure access. Patients who preferred the consent form and self-reported data on paper received this and submitted these papers back to the clinic which was stored in a locked safety box. Data from medical records and the national diabetes quality registry (NOKLUS) were manually extracted into Microsoft Excel files, stored, and managed in TSD. To ensure the accuracy of the data extraction, a coresearcher cross-checked all extracted data against the original medical records and NOKLUS registry, as described in the protocol [15].

In addition, to compare our sample with the overall hospital sample comprising all patients with type 1 diabetes and the national population of people with type 1 diabetes, we received aggregated data directly from NOKLUS on the variables age, gender, education, years since diabetes duration, insulin delivery method, and glycated hemoglobin (HbA_1c), all provided in median and range. Comparing our sample with the national population is essential to identify and assess biases, ensuring that our findings are representative and applicable to the broader hospital sample and national population.

To explore the role of patients’ travel time from home to the outpatient clinic, an estimate of travel time was calculated based on the patients’ postal code. Google Maps was used to determine the travel time by car to the main premises of the hospital. Akershus University Hospital is a local hospital with a defined coverage area. Participants living outside this area were typically students or recent movers who had not switched to their local hospital or individuals who, for other reasons, did not receive outpatient care at their local hospital. Public transport is only available for parts of the coverage area and was therefore not used in the calculation.

As this research project involves human subjects, the project has been reported to the Norwegian Agency for Shared Services in Education and Research (SIKT; reference number: 456954) and approved by the local data protection officer at Akershus University Hospital (reference: 2022_125). The project was presented to the Regional Committees for Medical and Health Research Ethics (REK; reference number: 407539), and they deemed it to be outside their scope and defined as a quality improvement project.

All participants received oral and written information before providing informed consent to participate in this study and filling out the self-reported data. Digital consent forms, medical records data, and self-reported data were stored and managed in TSD. Participants provided their social security numbers for in the consent form but were assigned ID numbers for deidentification. The code key was stored in a separate file within TSD. Data received from NOKLUS for comparison with the overall hospital sample and the national population were aggregated and thus anonymous. The study participants were not compensated for participating in this study.

Categorical variables are described with counts and percentages; continuous variables are presented with median values and ranges. To compare DigiDiaS care and usual care, we used nonparametric chi-square tests for pairs of categorical variables and Mann-Whitney U tests for pairs of continuous data. Missing data were not imputed, and the counts and proportions of missing values for each variable are presented in tables. Responses such as “don’t know” or “not applicable” in self-reported questionnaires were treated as missing values. To assess internal consistency of the scales used in the study, Cronbach α was calculated for the validated questionnaires [45].

To assess the strength of the possible association between the selected patient- and disease-related variables and the outcome, univariate logistic regression was conducted. Variables that were statistically significant in univariate analyses were entered into a multiple regression model, and the final model was fitted using the backward conditional selection process. The results are expressed as odds ratios (ORs) with 95% CIs. All tests were 2-sided, and P values <.05 were considered statistically significant.

In addition, to compare our sample with the overall hospital sample comprising all patients with type 1 diabetes and the national population of people with type 1 diabetes, we calculated CIs for the median values in our study sample to compare with the median and min-max values of the hospital sample and national population.

All data were processed using Excel (Microsoft Office Professional Plus 2016) and SPSS (version 29.0.0.0).

Of the 398 patients invited to participate in the study, 241 consented to participate (Figure 1). Of the 237 patients included, 185 patients opted for DigiDiaS care, and 52 patients opted to continue with usual care. In the DigiDiaS care groups, 21 self-reported data were missing and 5 from the usual care group.

The usual care group (median 68, range 16-100) had a statistically significant higher well-being score than the DigiDiaS care group (P=.01) indicating better well-being, with the usual care group having a higher well-being score than DigiDiaS care (median 60, range 4-96; Table 6).

Furthermore, we assessed the correlation between well-being scores and HbA_1c levels using Spearman correlation, finding no statistically significant relationship in the study sample, within the DigiDiaS care group (P=.37), or the usual care group (P=.33).

If the median value or percentage for the hospital selection and national population falls within the CI of the study sample, there is no significant difference. However, if the median value or percentage falls outside the CI, the groups may be significantly different.

This cross-sectional study was designed to explore the characteristics of patients with type 1 diabetes who were given a choice between a novel, flexible digital mHealth supplement for outpatient care, referred to as DigiDiaS care, or continuing with their usual care.

Our main findings reveal, as expected, that the vast majority of the 238 participants included in the study opted for DigiDiaS care rather than usual care. When comparing the two groups, the DigiDiaS care group had a shorter duration of diabetes, more use of insulin pumps than insulin pens for insulin delivery, and a lower well-being score compared with the usual care group. Those who opted for DigiDiaS care did not differ from those who continued with their usual care in terms of sociodemographic variables, presence of late complications or comorbidities from diabetes, self-management, diabetes distress, or health literacy. Furthermore, our study sample did not differ from the outpatient sample or the national population on relevant variables.

Not surprisingly, most participants opted for DigiDiaS care, aligning well with their profile of being both younger and already users of technology for insulin administration. Using an insulin pump requires certain technical skills, which are also necessary to use the DigiDiaS care app. Using an insulin pump instead of an insulin pen was the strongest characteristic associated with the choice of DigiDiaS care over usual care. This aligns well with previous research findings on how familiarity with diabetes-related technology might drive the adoption of other digital health tools [46,47]. These findings indicate that patients already accustomed to using advanced diabetes management technologies are more likely to embrace new digital health solutions, such as DigiDiaS care. In general, patients in the usual care group had longer diabetes duration compared with those in the DigiDiaS care group, and they may have established a routine for their outpatient care, making them less inclined to try out something new. Thus, when introducing digital care solutions in diabetes management, those already using diabetes technology equipment may be more prone to exploring the possibilities within new ways of receiving outpatient care.

In our sample, those opting for DigiDiaS care had lower well-being scores than those in the usual care group. Living with a chronic condition requiring ongoing self-management, such as type 1 diabetes, can affect well-being, even though well-being can vary throughout life and with age [4]. This variation in well-being is also observed in the Norwegian general population, where well-being appears to increase with age from the early 40s to the mid-70s [48]. This might correspond with how our participants opting for DigiDiaS care had a median age of 48 years—that is, 11 years younger than those opting for usual care and reporting a higher well-being score. Compared with the Norwegian population of adults with type 1 diabetes, both groups in our sample had a median well-being score higher than the national average [49]. Thus, despite our DigiDiaS group presenting a lower well-being score compared with those continuing with usual care, all participants had a fair level of well-being without symptoms of depression interfering with their abilities to self-manage.

Although most of the recruited participants opted for DigiDiaS care, a small proportion of the participants chose to continue with usual care, thus declining the opportunity to explore digital follow-up. In our study, we did not investigate actual use but rather the willingness to use DigiDiaS care. To the best of our knowledge, DigiDiaS care is the first outpatient care model comprising specific components, such as a message service, PRO-based questionnaires, an information page, and video consultations for diabetes care [15]. The closest alternative to our knowledge is DiabetesFlex, and our study is a response to the call by Jensen et al [17] after interviewing patients with type 1 diabetes who received flexible PRO-based telehealth as part of their diabetes follow-up in Denmark, questioning whether this type of solution is acceptable for all patients. Our findings show that not all patients opt for digital outpatient care. A previous scoping review on reasons for nonuse of digital PRO identified factors, such as ability to use the PRO, engagement, emotional distress, and technical barriers [50]. Our identified differences in the use of insulin delivery equipment may represent differences in the technical skills required to use an insulin pump. Furthermore, the scoping review highlighted how participants with few symptoms or those in good health stopped using digital PRO solutions more frequently [50]. This aligns well with the differences in well-being among our groups, where those with higher well-being chose usual care.

In contrast to previous research highlighting the importance of socioeconomic background when offering digital health services, differences in such characteristics were not observed in our sample. Previous research highlights that health literacy and socioeconomically disadvantaged backgrounds may be less likely to be considered for digital PRO-based outpatient care [51]. Our findings show that there are no statistical differences in health literacy or socioeconomic status between the DigiDiaS care and usual care groups, suggesting that there is no reason not to consider such patients to use these solutions. However, this may be contextual, as the Norwegian population in general has a high socioeconomic status. Yet, further research can provide a deeper understanding of the specific characteristics that influence patients’ decisions to choose digital outpatient care.

Although we examined various characteristics from medical records and self-reported questionnaires, we did not investigate all potential influencing factors. For instance, while we assessed general health literacy to understand the ability to find, understand, evaluate, and apply health information, digital health literacy could provide deeper insights into patients’ ability to effectively use such solutions [52]. In addition, efforts should be made to explore whether the use of digital outpatient care solutions such as DigiDiaS care varies with factors, such as geography, travel time, and the availability of health care services. The introduction of new technology in health care services is a complex process [53], and patient acceptance of new solutions is influenced by how health care professionals adopt and support these technologies [54]. Further understanding and investigation of these dynamics may reveal factors for successful implementation in practice.

This study presents strengths and limitations that are important to consider when interpreting the findings. First, this study explores a fully implemented digital outpatient care service [21], hindering the use of established gold standard research designs. However, our nonrandomized study design was essential in capturing the participants’ preferences, allowing them to choose either DigiDiaS care or continuing with usual care. This approach enabled us to investigate characteristics based on the participants’ choice of follow-up, which would not have been possible with other study designs, such as randomization or a stepped wedge design. The resulting skewed distribution was, therefore, expected and reflects these treatment choices. To compensate for the skewed distribution, nonparametric tests were used; nevertheless, the results should be interpreted with caution due to the uneven distribution.

While a none-randomized design with consecutive sampling may introduce some bias, it provides valuable insights into real-world decision-making and patient preferences that are crucial for understanding the adoption and effectiveness of the DigiDiaS care model [21], especially considering the needs and choices of the minority group who continued with usual care. Acknowledging that some participants do not wish to use this type of outpatient care is important, and the reasons for this should be further investigated. Qualitative interviews may reveal factors that we may not have detected in this study.

Second, this study does not examine the usability of individual components in the patient app or the specific solution of the app, but rather its clinical implications. Our investigation does not exhaustively cover all factors that may influence the choice of digital outpatient care given the complexity of such decisions. However, by collecting a broad range of data from medical records and self-reported data from the PRO questionnaire, we have gathered a comprehensive dataset. Our cross-sectional study, incorporating both generic and disease-specific well-validated measures, allowed us to add depth to the medical records data [55]. The richness of PRO data lies in the patient perspective it provides, reflecting their experiences, symptoms, and the impact of the disease on their daily lives [56].

Third, the digital outpatient care solution was introduced to half of the participants in the DigiDiaS care group before they were included in this study. The extent to which this was used is unknown, as its use is mainly related to the patients’ need to contact the clinic outside of fixed appointments, preparation for consultation or information, and e-learning courses received in connection with consultations. The time before we started recruiting for this study was also an introductory period in the use of DigiDiaS care for the clinic. We do not see it as a limitation that participants had access to the app before the start of the study, as we investigated opting for use of DigiDiaS care or not in this paper. However, as previously addressed, experiences from the Danish DiabetesFlex indicated that not everyone wanted to continue with DiabetesFlex after the study period [17], and even though our participants were opting for DigiDiaS care, actual use should be further investigated.

This study reveals that patients with type 1 diabetes are opting to use a digital outpatient care solution when it is offered. Most of participants opted for DigiDiaS care, the novel mHealth solution, while a smaller proportion opted to continue with usual care. Those opting for DigiDiaS care are characterized by being already familiar with using diabetes-related technology, having a shorter diabetes duration and lower well-being than the usual care group.

Importantly, our findings indicate that factors such as demographic and clinical variables, as well as self-management, health literacy, and diabetes distress did not differ significantly between DigiDiaS care and usual care. Even though only a small proportion continued with usual care, the characteristics of this group should be further explored to understand how these individuals can also be included in digital outpatient care solutions. Further studies should also focus on how the implementation of digital solutions in outpatient care can affect how and if patients use them.