Digital interventions for adolescents with type 1 diabetes to promote health literacy? A qualitative study on physicians’ perspective in Germany

Inclusion criteria and recruitment

The population of interest for this study was physicians working with adolescents with T1D in Germany. An additional inclusion criterion was that they worked in a hospital setting, either inpatient, outpatient or both. Recruitment was carried out using purposive, criterion-based sampling (10 participants), as well as snowball sampling by asking participants to share the invitation with their colleagues (2 participants). Participants were recruited from various locations across Germany via the diabetes-kids.de20 website, which serves as a forum for the largest virtual community for German-speaking children and adolescents with T1D. A total of 70 clinics were contacted either through a collective mailbox email or directly to physician’s email addresses. Out of 137 emails sent, 17 responses were received, and 12 interviews were conducted. Five participants withdrew due to lack of time (see figure 1). The individual semistructured interviews took place between October 2022 and January 2023.

Request permission

Sample

A total of 12 physicians working with adolescents with T1D were interviewed, representing 9 different hospitals in Germany. Of these physicians, eight were female and four were male, with ages ranging from 33 to 58 years and a mean age of 47.9 years. All physicians had a minimum of 2 years of specialised experience in paediatric endocrinology, with an average experience of 15.25 years. Among the participants, 10 were employed in a specialised hospital setting that offered both inpatient and outpatient care, while the remaining 2 were exclusively inpatient physicians (see table 1).

Data analysis

The duration of the interviews ranged from 16 to 30 min, with a mean duration of 24 min. The authors discussed and determined that sampling saturation had been reached.21 Indeed, content saturation was achieved, with no significant new information emerging compared with the earlier interviews.21 The individual interviews were conducted in person (ANN, a female doctoral candidate experienced in qualitative interviews), by telephone or by video conference. The researcher had no familial or acquaintance connections to any of the 12 participants prior to data collection. The authors have provided a guideline for the interview questions, consisting of open-ended questions to ensure flexibility and adaptability for each interview. However, the guide was not made available to the interviewees. It comprises seven sections and is available in online supplemental material 1—interview guide. Prior to conducting the study, a pilot test was carried out with two physicians. The two interviews were conducted to improve the interview guidelines and were, therefore, not included in the current study. The thematic coding analysis was conducted iteratively using MAXQDA software. The analysis was then refined through an iterative process that involved comparing and analysing themes across cases. Two authors (ANN and NF) independently coded all the coding material and achieved a level of intercoder reliability of 80% to ensure data quality, consistency, validity and reliability of the results.17

Patient and public involvement

There was no patient or public involvement in the design of the study.

Informed consent and confidentiality, ethics approval

Participation in the study was voluntary. Prior to the interviews, the participants were informed verbally and in writing of the study’s objectives, the use of data and their rights as study participants. After being informed, the participants gave their consent to participate in the study. The study design and the interview guide had been submitted to the ethics committee of Medical School Hannover and were accepted on 18 October 2022. The interviews were audiorecorded and transcribed, and the recordings were subsequently erased. To ensure anonymity, any information that could potentially identify the participants was removed from the transcripts.

Peer support and acceptance of illness

All physicians agree that the connection with other individuals of the same age group who are affected by the same illness (‘peer-to-peer support’) offers a significant benefit on multiple levels, although it is unlikely to occur by chance. First, peer-to-peer support can provide useful information that the physicians may not be able to explain due to lack of time and resources.

My experience is that affected patients, families, and caregivers benefit greatly from online peer groups, simply because there is so much to learn and know about a condition like diabetes. As health care professionals, we cannot possibly convey and accompany all of the information, so peer groups serve as a very valuable complement. (E11:58)

Additionally, advice or messages delivered by peers are more likely to be heard and integrated compared with those from non-illness-afflicted adult physicians. Second to facilitating the transmission of information about the illness, other themes such as belonging to a group and/or psychological support through a sense of camaraderie are beneficial. Moreover, certain emotional or personal topics that may be challenging to address with adults can be more easily addressed between peers. However, physicians’ opinions regarding the adoption of such exchanges or peer groups are contradictory. Some physicians claim that young individuals do not want to or are not motivated to meet and connect with other patients, while others argue that it is requested by patients and well received.

The main debate revolves around the utilisation of digital interventions used for communication and connection among peers. In general, digital interventions for peer-to-peer exchanges are still promoted at a low threshold. From the physicians’ perspective, contacts with other young individuals would be beneficial but would mainly occur during analogue meetings, educational groups or in rehabilitation measures. In these groups, a certain dynamic can be established more easily and can lead to interesting discussions, sharing of experiences or exchange of contact information. The COVID-19 pandemic had a negative impact in this regard, as the formation of such groups was no longer possible.

This is truly different, and we are not there when they can exchange information peacefully and independently from us. We actively strive to achieve this peer-to-peer support. (E8:38)

Few physicians highlight online platforms or support group websites (deemed inappropriate for this age group) as a means to connect with other peers but acknowledge digital interventions such as social media for maintaining contact among peers, but for contacts that are initially established in person. Conversely, some argue that it is easier to connect with other young people through an internet platform. Two significant challenges, as identified by physicians, relate to the use of digital interventions in peer platforms: (a) a lack of awareness of existing options, which often prevents physicians from recommending specific websites to their patients:

I am not aware of any really good online forums. The largest one is diabeteskids.de, which is semi-moderated. However, primarily parents participate there, not adolescents, and it focuses on personal and established groups. Of course, such forums exist, but it is difficult to make recommendations because they are open and lack specific recommendations. (E7:52)

(b) a critical view of free platforms due to the lack of control over the accuracy of information shared on these sites. Given that anyone can register on a free site and disseminate false information or create misleading profiles for malicious purposes.

The downside, of course, is that you never know who may be present in such online platforms. Just like anywhere else in the world, there are always individuals who spread misinformation or false facts, and as medical professionals, we have to counteract those initially. A classic example is the rumor that has been circulating in online communities for the past four years, although I cannot pinpoint its origin, that high-dose vitamin D can cure type 1 diabetes. This is, of course, nonsense. However, when vulnerable families who struggle with accepting the illness and the fact that their child has to inject and be pricked multiple times a day come across such information and seek help and refuge there, it can be dangerous. Firstly, because insulin may be omitted, and secondly, because excessive intake of vitamin D can be harmful. (E11:58)

This topic is currently being discussed in some hospitals to prevent the negative effects mentioned above and better organise online support communities through analysing the existing platforms instead of build something new.

There are already several platforms […] It’s more about connecting these resources together […]. The question arises: shouldn't we do more research in that direction? Ultimately, it will go in that direction, utilizing what already exists. (E1:46)

Another proposed solution is to better educate and engage with young individuals to understand their online behaviour. This involves empowering them with the necessary skills to discern and select reliable websites or platforms.

We already emphasize in our trainings that not everything related to diabetes found digitally is true, accurate, or recommended. For example, when considering the vast market of health apps related to diabetes, there is a significant and, in my opinion, relatively unregulated market. The issue is also that it is highly dynamic, and we may not be able to fully keep track of it in its entirety. (E5:36)

Communication and interaction

Communication and interaction during outpatient care follow-up

In outpatient care settings, communication between physicians and the patient primarily occurs through email or telephone. Some physicians says that phone contacts are effective, while some consider them to be less efficient:

The official procedure would be for them to contact us at our phone number at the Specialized Pediatric Center (SPC), leave a message on the voicemail. Then the nursing staff listens to the message and leaves me a note in my compartment. But sometimes I only see it days later. (E10:34)

Some physicians concur that email communication for appointment scheduling and inquiries is effective, as it allows for direct access through smartphones, which are ubiquitous among young individuals. However, some physicians hold the view that not all adolescents use email for communication, or that emails may go unread due to various reasons:

A patient sending me an email? That doesn't happen. Maybe when they're 17 or 18. That they would contact me and make a request. Otherwise, I do communicate with parents via email as well. (E9:54)

Some physicians argue that it may be necessary to communicate with patients in a different manner, using other digital platforms that would be more appropriate, such as a messenger service.

Preferably, something like WhatsApp would be ideal. However, we are not using it. There must be a data-secure platform that we (.) that we use in the clinic. (E1:32)

Some physicians use their personal mobile phones to communicate with patients via WhatsApp, but this is limited to specific and occasional situations, and it is not the official channel. Most of the barriers encountered by physicians revolve around data protection issues. One interviewee mentioned the creation of a patient portal where communication between physicians and patients would be possible, along with other interactive features such as information transmission and link sharing. The project is currently being evaluated in terms of data protection measures.

A patient portal on a medical basis, similar to a Moodle platform or something similar, where I can upload data, download information, view resources, find links, but also communicate and receive push notifications or similar features. (E4:50)

Another mobile application project is underway in the transition programme to maintain contact between the patient and the case management. The idea behind the mobile application is to explore better ways of reaching out to young patients. However, technical issues arise regularly and require frequent updates, resulting in it not functioning as well as desired.

Motivation and interest

A challenge mentioned by physicians is the difficulty in motivating young adolescents with T1D with digital interventions. Physicians instead use various communication techniques such as positive language or motivational interviewing to promote a trusting relationship. Empathy is a term frequently used in interviews and involves attentive listening from the physicians and direct communication addressed to the young patients. This also entails communication on an equal footing, aiming to establish a physician–patient rapport where the patient feels comfortable and able to open up to the physician, particularly regarding topics they may not necessarily want to discuss with their parents. It is important to strike the right balance here to avoid a friendly relationship or the use of ‘youth language’. While the topics discussed during a consultation mainly revolve around medical questions, physicians take the time to show interest in the patient’s personal life and address other topics beyond the illness, such as school. Other motivation techniques are used by some physicians, such as using ‘superdays’ where the consultation starts by discussing the ‘good days’ before addressing less successful aspects. Shared decision-making plays a predominant role in the conduct of consultations and helps to prevent an ‘abuse of power’.

Instead of talking about therapy failures or missed goals or poor control or derailment, we try to name what is happening in more neutral but also scientific terms. Yes, I think word choice plays an important role in communication and it also affects how much motivation someone develops to carry out daily management. (E10:30)

Another concept mentioned in the interviews is the ‘Language Matters Movement’, which involves communicating without stigmatisation, recognising through language that living with T1D is complex and exhausting, and assigned the responsibility to physicians to convey this recognition.

Another approach to fostering this trust-based environment, as suggested by some physicians, is through maintaining consistent physician–patient relationships. In other words, ensuring that each patient consistently sees the same physician(s) and does not have to switch physicians for each consultation. In some cases, poor clinical outcomes may be attributed to psychological factors, such as stress, which can be better understood when the patient opens up during a patient-centred healthcare encounter and shares their current struggles in their personal life, family or school. The patient may feel more comfortable disclosing these sometimes very intimate difficulties only when a foundation of trust has been established, a trust nurtured only when the same person is consistently present during the encounters. However, logistical and organisational constraints often make it impractical to maintain consistent physician–patient relationships in most hospitals.

The use of direct and in-person communication techniques, as mentioned above, has been highlighted to increase motivation among young patients. According to physicians, digital interventions do not play a significant role in this context and cannot replace the trust-based relationship established through direct and in-person interactions.

Limitations

The strength of this study lies in its in-depth understanding of the physicians and their perceptions through semistructured interviews that explored inductive themes. However, there are limitations that should be considered. First, only physicians were included in the study, while it would have been desirable to include other professionals such as diabetologists, social workers, psychologists, nurses, etc, who work with adolescents with T1D. Additionally, the sampling of physicians was random across Germany, but some regions were not included, which may introduce bias to the results. The size of hospitals and the distinction between outpatient and inpatient settings were not considered, which may also influence the results. The participants’ responses may have been influenced by socially acceptable expectations of the interviewer. It is possible that the recruitment method used in this study, based on a snowball effect with a sample of only two participants where a chief physician solicited the participation of two of their employees, may have introduced sampling bias. The responses of the participants may have been influenced by their hierarchical relationship with their superior, potentially biasing the results. In conclusion, it is essential to thoroughly examine the experiences and perspectives of physicians on the use of digital interventions in HL, considering structural constraints and comparing the experiences of young adolescents with T1D for a comprehensive understanding of the issue.