This was an analytical cross-sectional study conducted in the pediatric department and the diabetes unit of Treichville Teaching Hospital.

The diabetes unit operates from Monday to Friday, 7:30 AM to 4:30 PM. It holds four main consultation days: Monday through Thursday for scheduled appointments for both adult and pediatric patients, while Friday is dedicated to therapeutic education sessions covering diabetes, its complications, nutrition, and lifestyle measures. The unit’s activities include medical consultations (an average of 30 patients per day for three physicians scheduled daily), patient observation (day hospital care for parenteral treatments and other minor complications), and therapeutic education sessions (covering recommended physical activities and their duration, diet, the different types of insulin used, injection techniques, understanding the disease, possible complications, and the follow-up program). The unit also organizes training for physicians from other healthcare facilities, as well as diabetes screening and awareness campaigns.

Children followed-up are examined at least once every three months by the physician responsible for pediatric care, focusing on clinical management and therapeutic education. For children who have started treatment, follow-up visits take place weekly during the first three months. When glycemic control is good and the treatment is well managed, appointments are scheduled every three months.

Children are also followed daily in the pediatric ward, undergo quarterly check-ups, and are admitted to the pediatric emergency unit when complications arise. An organized system is in place within these care units to ensure appointment compliance, clinical and biological monitoring, assessment of diabetes-related knowledge, and therapeutic education. Additionally, all patients receive free insulin, glucometers, glucose test strips, and follow-up notebooks through the “Changing Diabetes in Children” (CDiC) program.

Data collection took place over a two-month period, from March 1 to April 30, 2025.

The study population consisted of children living with type 1 diabetes who were followed at Treichville Teaching Hospital.

Included were children diagnosed with type 1 diabetes at least three months prior to the start of the study, whose parents provided informed consent. Older children also gave oral assent to participate.

Children with other chronic conditions (e.g., psychiatric disorders, intellectual disabilities, chronic kidney disease, cancer) were excluded.

Sample size was calculated using the Schwartz formula:

Z = 1.96; margin of error (e) = 10%; prevalence (p) = 27% (based on Boiro et al. [9] ). This yielded a minimum required sample size of 71 participants.

After obtaining research approvals, all medical records of eligible children were reviewed. Patients were then contacted by the diabetes unit coordinator via telephone to schedule appointments according to their availability, with transportation costs reimbursed. Interviews were conducted either with the child alone or accompanied by parents, in a private office, ensuring confidentiality, lasting approximately 15 minutes on average. All questions were administered by the principal investigator.

Socio-demographic characteristics (age, gender, family structure, parental occupation, caregiver identity and education level, and socio-economic status).

Diabetes-related characteristics (age at diagnosis, disease duration, circumstances of diagnosis, person responsible for blood glucose monitoring and insulin administration, daily glucose monitoring frequency, insulin therapy regimen, pharmacy refill practices during supply shortages, type of physical activity).

Level of diabetes knowledge among children and/or parents.

Degree of treatment adherence and factors affecting it.

Data were collected using Google Forms and stored in Google Sheets, then exported to Excel 2021 for descriptive analysis. Univariate analysis was performed using XLSTAT 2019. Associations between qualitative variables were assessed using the chi-square test and Fisher’s exact test. Comparisons between qualitative and quantitative variables used ANOVA; when ANOVA assumptions were not met, the Kruskal-Wallis test was applied. A p-value < 0.05 was considered statistically significant.

For odds ratio (OR) calculations, the outcome variable—treatment adherence—was dichotomized: good/moderate adherence = good adherence; poor adherence = poor adherence. Similarly, diabetes knowledge was categorized as: sufficient knowledge (good/moderate) vs. insufficient knowledge.

Adherence to insulin therapy was assessed using the 4-item Morisky Green Levine scale, with yes/no responses scored 0 or 1.

The items were as follows:

Patients were classified into three groups based on total score: good adherence (4), moderate adherence (2 - 3), and poor adherence (0 - 1). This scale has demonstrated good reliability in diabetic children aged 6 - 18 years, with a Cronbach’s alpha of 0.857 [12] .

Acceptance of the disease: awareness of one’s physical condition after discussion with the staff, self-assessment, and active acceptance (recognizing oneself as ill, therefore having an insulin deficiency). This involves acknowledging one’s state of health and being willing to accept therapeutic treatment.

The study protocol was approved by the Medical and Scientific Directorate of Treichville Teaching Hospital. Data collection adhered strictly to ethical and professional standards.

The children and parents enrolled in the Changing Diabetes in Children (CDiC) program for follow-up and management had already provided written informed consent at the time of enrollment. It had been specified that the data collected would be used as a scientific database (to better understand the disease and raise public awareness). Thus, for the purposes of this study, only the oral consent of parents and of children over the age of 10 was obtained before the start of the interview, in compliance with the recommendations of the Helsinki Declaration.

Interviews were conducted in a confidential setting. Participant anonymity was maintained, and data were used solely for scientific purposes.

The relatively small sample size may reduce the power of the results, and recruitment from a single center could limit representativeness, potentially introducing selection bias. The assessment of knowledge was conducted using a non-standardized tool that may be incomplete and lacked cross-cultural validation. Additionally, the Morisky Green Levine adherence scale, although a reliable self-reporting method, remains subjective and less precise than other documented methods. It also does not account for two other essential aspects of diabetes management: diet and physical activity. Lastly, the small number of parents in the sample limits the interpretation of factors such as parental involvement.

Despite these limitations, the findings raise important points for discussion.

The mean age was 13.9 years, with a sex ratio of 1, indicating an even distribution between males and females. Adolescents aged 13 to 18 years were the most represented group. These results are consistent with those from Massi Idrissa et al. in Abidjan (2022), Boiro et al. in Senegal, and Ngwiri et al. in Kenya, reflecting a common overrepresentation of adolescents in pediatric diabetes studies across Sub-Saharan Africa [9] [13] [14] .

Most participants lived in two-parent households, a figure comparable to Massi Idrissa et al. [13] but higher than that reported by Niba et al. in Cameroon [15] . The predominance of two-parent families in our study may reflect local cultural specificities and may offer an advantage in treatment adherence.

Therapeutic adherence in this population often involved an identified caregiver. In our study, the majority of diabetic children managed their care independently. This differs from Niba et al., where mothers were the primary caregivers. This discrepancy, with a high proportion of self-managing children in our study, may be due to the relatively high average age (13.9 years) or the fact that the questionnaire in Niba’s study was completed by parents rather than the children themselves [15] .

Most caregivers in our study had a secondary education level. These findings are consistent with Kyokunzire et al. in Uganda, where 66% of caregivers had completed secondary school [16] . The predominance of adolescents in our sample partly explains this, as their parents may be better educated. This higher level of education likely contributes to a better understanding of disease management and may enhance treatment adherence.

Furthermore, most children came from families with a modest socio-economic status. This differs from Niba et al., where 69.7% of participants had low socio-economic status. This discrepancy may be due to the geographic setting of our study, conducted in an urban area, compared to the rural setting of Niba’s study [ 15] .

The mean age at diagnosis in our study was 10.9 ± 0.85 years, with an average disease duration of 3.9 ± 0.6 years. The most common presentation was the classic polyuria-polydipsia syndrome. These results are similar to those from Ngwiri et al. and Noyerie (Nantes, France, 2022) [14] [17] , suggesting that diabetes onset often precedes adolescence and may reflect challenges in early diagnosis or limited awareness of early symptoms.

A family history of diabetes was found in one-third of participants. This varied across studies: 57% in Uganda, 40% in northwestern Cameroon, and 29.6% in Thi-Qar, Iraq [15] - [18] . Such variability may be attributed to ethnic-genetic differences in susceptibility to type 1 diabetes.

Insulin therapy and self-monitoring of blood glucose were primarily managed by the children themselves in more than six out of seven cases. However, fewer than three in seven performed 2 - 3 daily blood glucose checks. These findings align with Djonou et al. in Cameroon, where 51.6% of patients reported 2 - 3 daily checks [19] . The similarity could be due to comparable methodologies, where children were interviewed without parental presence, limiting reporting bias.

Regarding insulin regimens, 79.7% of participants received 2 - 3 injections per day, with 3-injection regimens being most common (67.5%). These findings echo Djonou et al., where over two-thirds followed regimens with at least 3 injections per day. This preference may be due to ease of application, improving adherence among young patients [19] .

Therapeutic adherence and glycemic control were reflected in HbA1c levels. Our data revealed suboptimal control, with a mean HbA1c of 10.55%. Despite a modest improvement of 0.59% compared to the initial diagnosis, the reduction was insufficient: 90% of patients had HbA1c > 7%. Similar results were found by Kyokunzire et al. (80.5%) and Darwish in Egypt (92%) [16] [20] . The better outcomes reported by Noyerie (58% of poorly controlled patients at 1 year) may stem from their longitudinal study design, more structured management, and more favorable socio-economic context in Europe [17] . Ideally, the mean HbA1c should be maintained below 7%; however, achieving this target remains challenging across all settings and socio-economic contexts. Therapeutic strategies should therefore account for the unique physiological and psychosocial characteristics of adolescence.

The evaluation of diabetes knowledge, as part of therapeutic education, showed concerning levels among young patients. Only 5.4% had good knowledge, while 39.1% had poor and 55.4% average knowledge. These findings partially match those of Geneti et al. in Ethiopia, who reported 17.1% with good knowledge, and confirm the predominance of poor-to-average levels (50.5% and 32.3%) [21] .

Owusu et al. in Ghana added further insight, showing that knowledge of self-monitoring and hyperglycemia management was often satisfactory, suggesting that the gaps may lie more in theoretical understanding than in practical skills [22] .

Therapeutic adherence in children with type 1 diabetes is essential for treatment success and complication prevention. This study identified several factors significantly associated with good adherence, many of which are supported by the literature.

Better adherence was observed among children aged 0 - 6, while adherence was poorer in those aged 7 - 12 and 13 - 18, similar to findings by Habteyohans et al. in Egypt. That study noted that preschoolers were less likely to have poor glycemic control, likely due to parental supervision, which ensured optimal adherence (85%) [23] . The 7 - 12 age group may be particularly vulnerable as they begin to gain autonomy without yet being mature enough to manage diabetes independently.

Socio-economic status was associated with adherence (p > 0.05), with a trend toward better adherence among modest-income families compared to low-income ones. This may reflect greater household stability or more consistent access to transport and medical appointments. These results align with Niba et al. in Cameroon [15] . The provision of free insulin and glucose monitoring supplies through the CDiC project may have mitigated the typical impact of socio-economic hardship on adherence.

No correlation was found between caregiver education level and adherence. This may be due to the fact that 63.5% of children managed their care independently. This contrasts with findings from Dehiol et al. in Iraq, where maternal education was a key determinant of adherence [18] . This highlights the influence of cultural and contextual factors.

Children who performed 2 - 3 daily glucose checks had better adherence than those performing only one, echoing results from Niba et al., where 80% of patients with at least 3 checks per day were more adherent [15] . Thus, the frequency of glucose monitoring appears to be a predictive factor of adherence.

Glycated hemoglobin (HbA1c), as a reflection of adherence, showed that children with HbA1c < 8% were 42.4 times more likely to be adherent. This confirms the validity of using the Morisky-Green-Levine scale in our study. These findings support the conclusions of Elhenawy et al. in Egypt, who validated this tool for measuring insulin adherence in children [12] .

Knowledge level was also associated with good adherence, as shown in our study and in that of Geneti et al., where well-informed adolescents were nine times more likely to adhere to self-management protocols [21] . This may be because informed patients better understand the consequences of non-adherence, adapt more easily to treatment changes, and have greater confidence in managing their condition.

Other factors such as psycho-emotional well-being, disease acceptance, belief in treatment efficacy, and social support were also linked to adherence. This was similarly observed by Gonzalez et al., who emphasized that emotional factors (e.g., disease acceptance) and social context (e.g., family support) were key predictors of adherence in pediatric diabetes [12] [24] . Specific and feasible psychosocial interventions (e.g., structured peer-support programs, family therapy, or motivational interviewing) could be implemented in this context based on these findings to improve treatment adherence and compliance, particularly among adolescents. Baron et al. reported that adherence could be enhanced through therapeutic education and support for both the patient and their family. The goal is to help individuals acquire and maintain the skills necessary to optimally manage life with diabetes, while progressively developing self-care autonomy. This involves establishing realistic, shared treatment and adherence objectives that are discussed and negotiated between healthcare providers and patients—constituting what is referred to as a therapeutic alliance [25] .