Determinants of adherence to insulin and blood glucose monitoring among adolescents and young adults with type 1 diabetes in Qatar: a qualitative study

Ethical approval

The Institutional Review Board (IRB), Medical Research Centre Committee-Hamad Medical Corporation approved the study (17017/17).

Research design

Semi-structured, face-to-face individual interviews were used. The rationale for using this method was to obtain a detailed description and an in-depth understanding of diverse participants’ perspectives on subjects of interest,^60,61 and encourage free discussion, especially when disclosing sensitive issues.⁶² Data collection started in October 2017 and continued until January 2018, when 20 interviews were deemed sufficient to achieve data saturation,^51,63 and no further information could be added. This study was guided by the Consolidated Criteria for Reporting Qualitative Research (COREQ).⁶⁴

Participants and recruitment

Arab AYAs with T1D aged 17-24 years old receiving insulin therapy for at least two months prior to the study, and those using SMBG or CGM were eligible. AYAs were treated in outpatient diabetes clinics at Hamad General Hospital, the main public hospital in Doha, Qatar. The exclusion criteria were: gestational diabetes, secondary diabetes resulting from medical conditions or certain medications, such as corticosteroids; cognitive impairments, and drug and alcohol dependence. To ensure maximum variation sampling,⁶⁵ the physicians purposively selected eligible AYAs with T1D who were attending the clinic and had varied experiences across the indicated age group, i.e., AYAs with optimal, suboptimal and high blood glucose levels,^66,67 using HbA1c as an index of average blood glucose levels over the previous two-three months.⁶⁸ The aim was to obtain varied participants’ perceptions of their HbA1c levels and whether they linked these levels to their adherence behaviors. Data on HbA1c was collected from AYAs’ records. HbA1c levels were categorized into three levels based on the clinical practice guidelines’ recommendations.^23,69 The optimal level, defined as <7.5% (<58 mmol/mol), the suboptimal HbA1c level, defined as 7.5–9% (58–75 mmol/mol), and the high HbA1c level, defined as >9% (>75 mmol/mol) in adolescents. In young adults, the optimal HbA1c level is denoted <7% (<53 mmol/mol), the suboptimal HbA1c level is 7–7.9% (53–63 mmol/mol), and the high HA1c level is denoted ≥8% (≥ 64 mmol/mol). Purposive sampling continued in parallel during data collection. AYAs with T1D who agreed to participate were referred to the principal investigator (HB), who explained the study in detail and obtained written informed consent forms from young adults aged 18 to 24 years and assent forms from adolescents under 18 years old, which included their carers’ consents. Two persons declined to participate in the study, citing a busy schedule, but none of those who agreed to participate withdrew from the study.

Interview process and procedure

The first author (HB), trained in qualitative research and having no prior or post work relationship with the participants, used an open-ended and piloted interview guide during the interviews. ICM was used as the theoretical framework for developing the guide.⁵² The interview discussion topics are represented in Table 1. Experts in qualitative research methodology, in the area of behavior change theories, and in T1D, guided the development of the guide. The interviews were held in private at the outpatient diabetes clinics and took approximately 60 minutes. Participants were assured that all information would only be used for research purposes. Permission to audio-record and to take notes was granted before starting the interviews, which were then transcribed in Arabic before being translated into English. During the interview sessions, probing questions were used to elicit information extraction, such as (“Can you tell me more? Would you give me an example? Would you explain that further?” among others). To ensure that the participants’ answers were understood, the interviewer repeated the participants’ words and confirmed the summarized main points of their responses with them. Prior to interviewing, we collected information on demographics, diabetes, and insulin history.

Data analysis

Demographic and clinical characteristics

Demographic and clinical characteristics were described using descriptive statistics and frequencies.

Results were expressed as means (standard deviation [SD]) or percentage of total responses, such as the percentages of participants with optimal, suboptimal, or high HbA1c levels. Participants were classified as adherents or suboptimal adherents based on their responses during the interviews. Those who reported always or almost always administering insulin or testing blood glucose as recommended were classified as adherents; those who did not were classified as suboptimal adherents.

Thematic data analysis

The framework method was used to analyze the data thematically.^72,73 This method was chosen for the following reasons: first, it enables capturing different aspects of behavioral determinants; second, it guides systemic analysis through interrelated stages to describe as well as understanding the behaviors with their corresponding determinants. Third, it ensures transparency of data interpretations.⁷⁴ The analysis started with the familiarization stage, during which the coder (HB) read through all the transcripts and generated notes about common themes and other concepts from the data by considering each line and phrase. Firstly, a deductive theory-driven analysis was chosen. Based on this approach, a coding book was initially developed using the ICM described previously. The codebook was extensively reviewed by the authors HB, FS, and LM until agreement on the main codes and sub-codes was reached. All transcripts were coded manually in accordance with the coding book. The researchers HB, FS, and LM discussed the coding process to establish reliability.⁷⁵ The next stage was identifying the thematic framework (coding tree). During this process, codes were organized hierarchically based on how they related to one another, and additional emergent codes from the open discussions were added to the initial codebook and coding tree. An inductive approach was also used to generate meanings and to identify the patterns that could be grouped into themes and categories. During the final interpretation stage, mapping connections between the categories and the main themes aimed to look for similarities, differences, and patterns in the meaning of data. For example, the relationship between self-efficacy and coping plans; comparing participants’ perceptions about their diabetes management and their HbA1c values. Validity and reliability were established through member checking during the coding process and investigator triangulation during data analysis.⁶⁰

Demographic characteristics and medical status information

Out of the 20 interviewees, 70% were Qatari, 55% were females, and 65% had diabetes for longer than ten years. The mean age was 20.5 ± 2.35. Demographic and medical status information are shown in Table 2.

Distal information factors

AYAs with T1D most often asked healthcare professionals (physicians, diabetes educators, and dietitians) about information related to insulin and BGM. Many used general Internet websites as they indicated that diabetes-specialized websites in the Arabic language were deficient (Quote #1). They also sought information from friends without diabetes, parents, relatives, and trainers at the gym. Nearly all participants suggested that the information ideally should be simplified, updated, applicable to daily life, and refreshed. Female subjects wanted more information about pregnancy and child delivery. Others wanted to know more about cases that related to them to help them in the decisions they make on a daily basis (Quote #2). Others wanted more information about counting CHOs and what to do when blood glucose levels are very high.

Pre-motivational factors

Cognizance

Almost all participants reported being cognizant about whether they were adherents or not to injecting insulin and BGM as agreed with their HCPs. Quite a few participants indicated being adherent to IA on time. Still, some reported that they sometimes skipped insulin, either intentionally or unintentionally. A few others admitted that they were suboptimal adherents most of the time. Some AYAs, particularly females, deliberately omitted insulin doses to induce hyperglycemia for inpatient admission to avoid school exams or family conflicts (Quote #3). Very few females skipped or reduced insulin doses to control their weight (Quote #4). Irrespective of adherence to insulin, many participants were adherent to carbohydrate counting, apart from a few participants who either needed more education or more motivation to use the required information. Very few participants described themselves as adherent to healthcare providers’ recommendations in terms of the frequency of SMBG and CGM. Two participants said that they would check their blood glucose before doing physical activity. The majority of participants reported testing randomly or rarely (Quotes #5 and #6).

The awareness of diabetes management varied between participants. Many of them, irrespective of whether they were adherents or suboptimal adherents, believed that their diabetes was unmanaged. Some reported that their T1D was fairly managed (Quote #7). Some suboptimal adherent AYAs perceived that their diabetes was well managed, but their perceptions were not consistent with their actual HbA1c levels; thus, they did not feel that they needed to adjust their adherence behaviors (Quotes #8 and #9).

The main changes few suboptimal adherents wished to make were to manage their time effectively, care more about their health, be able to make diabetes decisions in general and manage their HbA1c. Relating to IA, some suboptimal adherents wished not to forget to administer or skip doses. Also, to be competent in counting CHO, while other suboptimal adherents wished to perform BGM more frequently as recommended. A number of suboptimal adherents indicated that there was no need to change their adherence behavior at all. They stated that they did not accept the disease or the treatment; others believed that bringing their glucose levels near normal levels was impossible (Quotes #10 and #11).

Knowledge

Regarding knowledge, all adherents reported knowing how to calculate CHO and use the corrective dose, as well as how to obtain the necessary information from HCPs (Quote #12). However, a few suboptimal adherents stated that they lacked the necessary knowledge to count CHO content, adjust insulin dose accordingly, and base decisions on BGM results (Quote #13), and a couple others reported lacking a certain level of knowledge about long-term complications of diabetes.

Risk perception

Nearly all adherents and many suboptimal adherents reported being aware of the risks of developing short- and long-term complications resulting from suboptimal adherence and acknowledged that these complications could have a severe impact on their health, like prolonged hyperglycemia, coma, blindness, and leg amputation (Quote #14). A few adherents, on the other hand, believed that adherence would delay severe long-term complications like renal failure; however, less severe complications like microalbunurea and short-term complications like hypoglycemia are unavoidable (Quote #15). Nonetheless, some suboptimal adherents believed that whatever they did, the complications were unavoidable (Quote #16). Many suboptimal adherents were more likely to be worried about the complications if they avoided thinking about them. Others believed that the complications would not occur as long as they manage their diabetes well (Quotes #17 and 18). Some suboptimal adherent adolescents believed they still had years to develop complications since they were still young, or according to their religious beliefs, everything that happens to them is destined to happen (Quotes #19 and 20).

Regarding BGM, very few adherents thought they were less likely to have complications because of regular testing. They could easily detect and act on hypo-or hyperglycemia (Quote #21). A number of suboptimal adherents, however, did not associate the risks of their suboptimal adherence with a level of diabetes management (e.g., detecting variations in blood glucose levels and linking these to possible causes and acting on them to bring blood glucose levels to a near normal level) or with diabetes-related complications (e.g., detecting unawareness hypoglycemia). These participants checked only when they felt the symptoms of hypo-or hyper-glycemia. Some others lacked the awareness to check even when they were experiencing hypoglycemia (Quotes #22 and 23) (Table 3).

Motivational factors

Attitude

Participants discussed the advantages and disadvantages of adhering to insulin from a physical, practical, and psychological perspective. All interviewees reported similar physical advantages, like managing, normalizing, or decreasing blood glucose and subsequently avoiding symptoms of hyperglycemia, e.g., fatigue, dizziness, and frequent urination (Quotes #24). The main practical benefit identified by many adherents was improved quality of life for not being hyperglycemic all of the time. Adherents described specific benefits related to the practicality of using injectable devices; while some found injectable pens satisfactory and provided flexibility; others found insulin pumps to be more flexible (Quotes #25 and 26). Nearly all adhering participants agreed on the psychological advantages of adherence, like feeling relieved, not worried, not stressed, under control, and feeling like a normal person (Quote #27).

All interviewees reported similar physical disadvantages, like skin irritation and discoloration, muscle deformity, and hypoglycemia (Quote #28). Few students talked about the disadvantages of experiencing hypoglycemia the night before or while taking standardized tests, which have adversely affected their exam performances. They requested consideration of options to retake the test and to refund the test fees in case of inability to take the exam due to their diabetes (Quote #29). Weight gain was of particular inconvenience, especially among pump users because of the ease of use of corrective doses and the difficulty of losing weight (Quote #30). Suboptimal adherents talked about the interference of insulin in their daily routine and found the process of injecting and CHO counting as tiring and time-consuming as the main practical disadvantages (Quote #31). Some suboptimal adherents found the pump inconvenient because of life-interference, pain at the injection site, hassle, and the time it takes to prepare the pump (Quotes #32). All suboptimal adherents discussed the psychosocial disadvantages relevant to diabetes care, including negative emotional reactions such as anxiety, depression, among others; ideals for weight; intolerance of regularity; fear of needles and testing, fear of hypoglycemia, which resulted in omitting insulin to avoid hypoglycemia; feeling the burden of counting CHO in their meals and synchronizing insulin dose with CHO content (Quotes # 33). Few female subjects mentioned their concerns about future marital relationships and the side effects of insulin on pregnancy and having babies. One male talked about his concern about erectile dysfunction.

Regarding BGM, very few adherents appreciated the advantages of providing assurance and safety through detecting hypoglycemia/hyperglycemia patterns and blood glucose levels (Quote #34). CGM’s specific practical advantages were the ease and convenience of using sensors (Quote #35). Overall, many suboptimal adherents doubted the role of monitoring in managing diabetes (Quote #36) and described physical, practical, and psychological disadvantages. The physical disadvantages included pain, while skin discoloration, irritation, and callus were particularly relevant among female suboptimal adherents, such that they stopped testing altogether (Quote #37). Some non-Qatari individuals mentioned the cost. Practical disadvantages include frequency of testing, as it needs considerable time management and planning efforts; accuracy and practicality of use through the ability to interpret and act on results; and access to a glucometer. A few suboptimal adherents found CGM inaccurate and described preparing the sensor as disturbing. Suboptimal adherents illustrated psychological disadvantages linked to fear of pricking and negative feelings.

Social influence

To many adherents, family social influence was perceived as positive, especially when they were younger. However, as they grew older, they wanted to take more control over their diabetes, IA, and BGM. Many of them also appreciated the support they got from their healthcare team, especially the diabetes educators (Quotes #38 and #39). Several suboptimal adherents accepted advice around diabetes from their close friends but not from their families; they felt family interactions were stressful and conflictual, which elicited guilt and overprotective emotions (Quotes #40 and #41).

While some adherents informed their friends about their disease and IA to help them when they were experiencing hypoglycemia events, suboptimal adherents did the opposite because they did not want to be treated differently or with pity. Some felt that the interference of their friends had embarrassed or labelled them. Others justified that their friends were less supportive because they did not know much about diabetes and its complications. To avoid stigma, some skipped insulin or testing in public places (Quotes #42 and #43).

One of the emerged themes was social modelling, where some adolescents used traditional homeopathic medicines such as ginger, cinnamon, and pomegranate peel drinks or even bee stings to substitute insulin, just as their older relatives with diabetes did (Quote #44).

Self-efficacy

The situations which facilitated adherence for some adherents were using the pump and seeing positive results; also, when not surrounded by people and when supported by the healthcare system, when injecting or testing. Unlike suboptimal adherents, the daily IA and BGM of adherents became life routines with the long duration of diabetes (Quotes #45 and #46).

Suboptimal adherents, particularly those using insulin pens and glucometers, highlighted a range of situations in which IA and BGM adherence were perceived to be difficult. These included, first, practical situations in which AYAs found it unpractical to take the insulin pen or the glucometer with them when they are outdoors (Quotes #47 and #48); second, social and cultural situations such as injecting or testing in public places because of people’s curiosity and being judgmental or due to shyness or preference for maintaining a relationship over being adherent (Quotes #49 and #50); third, physical external environmental situations such as traveling, when engaged in social events, being busy. The internal physical situations were related to forgetfulness and sleep issues like feeling sleepy, having a different sleeping pattern, or sleeping (Quotes #51 and #52); fourth, psychological situations, e.g., being in a bad mood, low motivation, feeling restricted and occupied. All suboptimal adherents reported being overwhelmed by their emotions due to the disease’s long duration and the fact that it is incurable, which has negatively impacted their acceptance and adherence (Quotes #53 - #55) (Table 4).

Post-motivational factors

Preparatory, action and coping planning

The majority of participants, irrespective of adherence, did not report having made preparatory or action plans to manage their T1D, adhere to IA and/or BGM, or cope with problems. A few suboptimal adherents considered increasing the frequency of SMBG, decreasing insulin dose, and not using the corrective dose on a regular basis, but they lacked structured plans (Quote # 56). Few others saw planning as an additional emotional strain because they would be preoccupied with their plans (Quote #57).

Concerning coping planning, only one newly diagnosed male used reminders to help him adhere to IA and SMBG (Quote #58). The vast majority of suboptimal adherents did nothing to cope with the difficult situations they encountered in IA, while some used maladaptive coping strategies by drinking herbal tea when they were in a bad mood to use insulin (Quotes #59 and #60). Further, setting goals, maintaining adherence, and the actual skills required (matching insulin dose with CHO content and interpreting BGM results and the ability to act on them) were reported by many suboptimal adherents as areas where they needed improvement in. Routine intolerance and decommitment, having low desire, lack of intention and self-control, distraction, confusion about what to do, not feeling supported as they used to be when they were younger, and not seeing the results, were reported as common barriers to coping with adherence. (Quotes #61- #63). Among the very few facilitating factors reported by adherents were that they kept themselves motivated and had a great deal of self-discipline (Quote #65) (Table 5).

Information factors

Nearly all participants wanted information provided to them to apply to their daily lives, and be refreshed more frequently. Consistent with other results,⁷⁶ female subjects wanted more information about pregnancy and child delivery. Few suboptimal adherents sought information from older relatives who recommended homeopathic medicines instead of insulin or from generic websites. A previous study among Qatari participants showed that 95% of them used Google as a search engine to look for health information, and the younger adults were using the Internet significantly more than older adults.⁷⁷ Evidence supports that lack of information⁷⁸ or irrelevant and false information can negatively impact diabetes management. International guidelines recommend that the information provided to patients should be structured, continuous, and accessible in a language that the patient understands and fulfils his/her needs.^79,80 Therefore, it is necessary to focus on improving information-seeking behaviors, understanding how the patient processes information, and translating this understanding into action so as to improve knowledge and facilitate positive self-care behaviors.

Pre-motivational factors: Cognizance, knowledge, and risk perception

Concerning cognizance, this study showed that almost all adherents and suboptimal adherents reported being aware of their behaviors. Yet, a few suboptimal adherents did not have enough awareness that they needed to change their behaviors due to misjudgment about their level of diabetes management, misbelief about achieving target levels, or lack of acceptance of the disease and the treatment. Prior findings support the hypothesis that AYAs who accepted their disease were more adherent than those who did not,^81,82 while those who did not achieve their target blood glucose levels felt powerless and under an impossible burden.⁸³ It is documented that knowing how the patient perceives his or her illness improves adherence.^84,85 Therefore, increased awareness, monitoring, and support for AYAs with T1D, especially those with problems in psychological adaptation, are needed within different T1D care settings.

Regarding knowledge, our study confirms recent findings on the importance of periodically assessing the knowledge and readiness for education of people with diabetes in clinic visits to meet the demanding nature of diabetes.⁸⁶ This is necessary because learning is more likely to be achieved when it is presented from the patient’s perspective.

Concerning risk perceptions, our findings showed that few adherents perceived insulin adherence would delay long-term complications (like renal failure) but not short-term complications (like microalbuminuria). Researchers have found contradictory results relating risk perceptions of AYAs with T1D and adherence to DSM activities, emphasizing the need to identify short- and long-term complications when assessing risk perceptions of diabetes in AYAs.^87,88 Additionally, in this study, suboptimal adherents indicated that they were not aware of the risks associated with suboptimal adherence. Some groups, reported to deliberately omit insulin doses to induce hyperglycemia to get exam exemptions or avoid family problems. Previous research showed that female adolescents who experienced family problems were more likely to develop psychiatric problems and medication suboptimal adherence.⁸⁹ Previous cases of deliberate insulin misuse to induce dysglycemia (hypoglycemia or hyperglycemia) have been observed in AYAs with T1D. Their motives varied, such as mental disorders like major depression or factitious disorder (FD) and/or suicidal ideation.^{31,34,90–92} Induced hypoglycemia was also used to justify eating sweets and carbohydrate-rich meals.³¹ Factitious dysglycemia should be considered a possible cause of brittle diabetes and requires specialist consultation.^90,92 However, the literature on FDs from the Arab region is limited,^93,94 with only one report on Iraqi women with type 1 and type 2 diabetes.⁹² More information is needed from the Arab region to clarify FD’s clinical profile and implications for T1D. Our findings support results from earlier studies^95,96 indicating that some AYAs with T1D were not sufficiently aware of the consequences of their illness-specific risk behaviors, which could have a negative impact on optimal management and outcomes, such as severe hyperglycemia, diabetic ketoacidosis, and long-term diabetes complications. Further, similar to previous findings,⁹⁷ other suboptimal adherents believed that they were less vulnerable to complications because of inherent beliefs that they wouldn’t suffer any negative outcomes or because complications would occur at an older age. Others, and in line with Garrett and colleagues’ study (2014)⁹⁸ felt that complications were inevitable despite their best efforts. Moreover, a few suboptimal adherents perceived the susceptibility and severity of risks associated with IA suboptimal adherence as ongoing threats that were worrying and depressing, and they avoided thinking about them. These threats have previously been identified to affect AYAs’ emotional well-being and adherence,^99–101 and those who perceived their illness to be of greater severity had higher glucose levels than the targets.^102,103 Yet, perceptions of risk and severity may encourage adherence.^84,104,105 A meta-analysis has shown that communicating threat messages has an effect on behavior only when efficacy is high.¹⁰⁶ Indeed, our results have addressed, similarly to other recommendations,^85,107,108 the importance of teaching AYAs about the potential long-term complications in a way that is not depressing or discouraging.

The findings indicated that the majority of suboptimal adherents did not perceive being at risk because of suboptimal adherence to BGM, and thus they rarely tested themselves. Previously, knowledge gaps regarding SMGB were identified.^39,109,110 A prior study found that people with diabetes who were less knowledgeable about the importance of glycemic management in developing diabetes vascular complications were less adherent to SMBG.¹¹¹ In summary, understanding awareness factors is a prerequisite for understanding individuals’ engagement or disengagement in adherence behavior and should be continually evaluated and addressed in every health encounter.

Motivational factors: attitude, social influence and self-efficacy

The majority of suboptimal adherents reported not believing that adhering to the frequency of BGM as recommended would be advantageous. Similar findings are also reported in other studies.^112,113 It is also known that the psychological impact of a chronic illness alters an individual’s attitude and affects medication adherence.¹¹⁴ While national and international guidelines recommend that periodic psychological screening should generally be a routine part of diabetes management,^115–118 it is clear that these recommendations are not yet being followed up sufficiently. It is important to develop communication strategies to convince suboptimal adherent individuals to adhere to the recommended frequency of BGM. This study also highlights AYAs’ concerns about diabetes and its impact on pregnancy, marital life, and sexual dysfunction (SD). Previous research has discovered that AYAs with T1D have more conflicting marital relationships than their counterparts without diabetes.^58,89 Studies from the Arab world^119,120 discussed the implications of diabetes on sexual life, usually considering SD in men and people with T2D. AYAs with T1D have been given less attention, underlining the need for more exploration on this subject and in relation to insulin adherence in order to better understand these factors.

In our study, AYAs reported hypoglycemia before or during standardized tests to influence their concentration and performance, requesting the option to retake the test or be refunded. A systematic review reported that 23–39% of older children (>11 years) reported that they did not have the opportunity to repeat school exams again when they experienced hypoglycemic events before or during an exam.¹²¹ Another systematic meta-review¹²² has provided evidence that students with chronic illness often demonstrate worse school experiences and outcomes than students without chronic illness. While some countries have expressly defined legislative protections to support students with T1D^123–125 other countries have not. Additionally, these specific laws relating to the requested reasonable accommodations during exams vary from country to country. Therefore, international and national policies should provide unified guidance to ensure that AYAs with T1D are not placed at a considerable disadvantage compared to students without diabetes. There is a need for further research into the perceptions and needs of students with T1D relating to exams in the Arab world.

In this study, adherents indicated that having good support from HCPs had motivated their IA adherence. They added that the availability of social media as a method of communication between them and the diabetes healthcare team had empowered them in their daily decision-making. Prior findings indicated that the patient-provider relationship strongly impacts treatment adherence in AYAs with T1D.^14,126 AYAs with T1D also wanted to become more autonomous in their self-care decisions and behaviors with balanced family interactions, suggesting that the transition of diabetes responsibilities from pediatric to adult should be considered and evaluated for AYAs and their families in order to avoid family conflicts. Previous findings demonstrated that negative family reactions to self-care behaviors were found to be related to suboptimal adherence and suboptimal blood glucose levels.^102,127 Noser et al. (2017) found that diabetes-specific family conflict negatively influenced SMBG adherence in young adults with T1D even if they possessed high self-efficacy levels.¹²⁸ The national clinical guidelines¹¹⁶ emphasize the delivery of diabetes self-management education (DSME) at transition points in care. However, more information on the details of the processes involved in the transition preparation period and on how to fully implement these processes is needed to be addressed in further research. Our results revealed that key people in the interviewees’ lives and cultural factors affected their perception of the meaning of illness and its treatment. In a previous study, some adults with T1D commented that their family and friends provided incorrect information.¹²⁹ Hence, the sociocultural environment should be evaluated closely.

Our findings demonstrated that adherents showed higher levels of confidence in their ability to follow medical recommendations and expected more meaningful positive consequences for adherence. On the other hand, difficulty administering insulin and/or performing BGM as recommended has been reported in suboptimal adherents as a result of low self-efficacy. Self-efficacy was previously identified as a significant predictor of relatively high adherence rates.^42,130 Therefore, self-efficacy promotion is important to help overcome the various barriers and, in turn, to improve adherence in Arab AYAs with T1D.

Post-motivational factors: action planning: preparatory and coping planning

Our results revealed that action planning was lacking for nearly all participants, whether adherent or not. Some suboptimal adherents indicated that planning aggravates emotional strains. Previous studies highlighted the action planning construct as a prominent predictor of adherence to asmathtic¹³¹ and antiretroviral¹³² medication in AYAs. Additionally, goal-setting was shown to improve SMBG.³⁸ It is therefore important that health communication programs to foster T1D adherence focus on adequate action planning and goal setting for people with T1D in Qatar.

Strengths and limitations

The findings from this study can serve to design an educational program in which these psychosocial determinants can be addressed in a patient-centered approach. The strengths of this research are that, firstly, the face-to-face semi-structured approach with a researcher who did not have a relationship before and after the work helped AYAs communicate openly about sensitive issues such as SD, which would be generally difficult to address,¹³³ particularly since it is influenced by different tribal and social attitudes in the Arab region.¹³⁴ Secondly, while a sample of 12 interviews demonstrated achieving saturation previously with a similar study nature and analytical approach,^135–137 we examined the depth and richness of collected information using a saturation grid,¹³⁸ and interviewing continued until saturation was deemed achieved with a sample of 20 interviews. Therefore, findings from this study may be transferable to similar groups. Thirdly, adopting a framework analysis approach offered a systematic structure to easily manage, analyze, and identify themes.¹³⁹

This study also has some limitations. Data collection was self-reported; therefore, reporting bias is possible. Although information about the magnitude of such bias is unavailable in most epidemiologic studies,¹⁴⁰ there is a reasonably reliable self-report when questions are asked in a non-judgmental manner.⁶⁶ Second, although this study has highlighted specific determinants related to the different insulin delivery devices and BGM systems, more research is needed in this area to draw further comparisons and conclusions. Third, the views expressed in the interviews are less representative of the adherents’ sample, particularly for BGM. Therefore, this study may not capture other important facilitators and experiences, highlighting the need for further research among adherents.

Conclusions

First, concerning information needs, most respondents reported that they needed more information relevant to their daily lives and more Arabic language websites that provide simple diabetes, insulin, and CHO counting information. Second, concerning awareness, some suboptimal adherents were not optimally aware of the need to change their behaviors, lacked the knowledge required to make decisions on insulin dose adjustments, and underestimated T1D risks. Third, concerning their motivation, many suboptimal adherents reported a negative attitude towards adherence, which resulted from several perceived disadvantages that outweighed the advantages of adherence. They also reported a lack of social support and a low sense of self-efficacy. Fourth, most respondents lacked specific plans to prepare for and cope with adherence. In conclusion, increased efforts are needed for people with T1D in Qatar to strengthen awareness, knowledge, and perceived risks of suboptimal adherence, as well as to realize a positive attitude, strengthen social support and self-efficacy, and enhance appropriate action planning. A comprehensive approach that takes into account the broader social context is needed in order to minimize conflicts in families and stigma in the sociocultural environment.

Recommendations

Previous systematic reviews and meta-analyses^141,142 on the effectiveness of DSM interventions in young adults with T1D indicated that a limited number of studies have incorporated behavioral theories or outcomes in their research, and there has been a noticeable underutilization of behavioral techniques. Additionally, these interventions were not adequately designed with the active participation of young adults. Consequently, the outcomes of these interventions did not lead to significant improvements in glycemic targets or self-management behaviors. Furthermore, self-management interventions designed for older adults should not be assumed to be equally effective for AYAs. This underscores the immediate demand for the development of more effective interventions tailored to this unique population of AYAs living with diabetes. Therefore, we recommend that for future DSM interventions to be effective, they should integrate the identified socio-cognitive factors throughout the awareness, motivation, and action phases, alongside informational factors. This should be complemented by precise operational measures for these factors. The development of these interventions should be a collaborative effort with AYAs, ensuring alignment with their unique needs and preferences, ultimately enhancing their overall efficacy.

Underlying data

The qualitative data that support the findings of this research are not publicly available due to concerns that the data could potentially reveal participants’ identities. However, data are available to researchers upon request from the corresponding author (H.Burno@maastrichtuniversity.nl/mailto:hanoneh111@gmail.com) and with permission from the Institutional Review Board-Hamad Medical Cooperation for further academic research.

Extended data

Figshare: Appendix 1: The interview guide for “Determinants of adherence to insulin and blood glucose monitoring among adolescents and young adults with type 1 diabetes in Qatar: a qualitative study”, https://doi.org/10.6084/m9.figshare.20368068.v2.¹⁴³

Data are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).