A Real-World Study of Electronic Prescriptions and Medication Adherence in Bulgaria

Abstract

Background: The implementation of electronic prescriptions in Bulgaria since 2023 provides new opportunities for real-world data analysis of prescribing practices and medication adherence particularly for chronic diseases such as diabetes. Objective: To evaluate prescribing patterns medication adherence and National Health Insurance Fund (NHIF) coverage in patients with diabetes across three major Bulgarian regions using 6-month electronic prescription (e-prescription) data from January to June 2024. Materials and Methods: Data were collected from community pharmacies in Sofia, Plovdiv, and Dobrich and Stara Zagora. Prescriptions were categorized by diabetes type (type 1 or type 2), therapy class (traditional vs. innovative), refill frequency, and NHIF reimbursement. Adherence was estimated using refill frequency and the Proportion of Days Covered (PDC) method. Results: A total of 1071 prescriptions were analyzed. In Sofia, type 1 patients (n = 13) had high rates of adherence (mean: 6.5 refills; 54% refill-based adherence), while type 2 patients (n = 138) showed more variability (mean: 4.55 refills; 27.5% with PDC ≥ 80%). Innovative therapies (e.g., DPP-4, SGLT2 inhibitors) were more frequently used in Sofia, reflected by higher NHIF reimbursement costs. In contrast, Dobrich, Stara Zagora, and Plovdiv predominantly relied on traditional medications (e.g., metformin, sulfonylureas) with lower adherence and limited use of innovative therapies. Conclusions: E-prescription data provides valuable insights into real-world diabetes management and shows substantial regional differences in adherence and treatment intensity. In the bigger cities there is better adherence and access to modern therapies, while smaller regions are slower in adopting these treatments. These findings support the need for national policies to ensure equitable access to innovative therapies and enhanced digital support for chronic disease management. From clinical practice point of view adherence follow-up is a key for treatment success and healthcare decision-making.

1. Introduction

World health organization claims that the number of people living with diabetes rose from 200 million in 1990 to 830 million in 2022. A total of 59% of diabetics aged 30 years and over were not taking medication regularly in 2022. Many countries do not cover the cost of diabetic medicines therapy [1]. As of 2024, the prevalence of diabetes in Bulgaria has reached alarming levels. According to leading endocrinologists, approximately 770,000 to 800,000 people in the country are currently living with diabetes. This marks a significant increase compared to previous years and reflects a broader global trend, with diabetes cases in Bulgaria rising by 199% between 2007 and 2021. The primary cause of this increase is obesity, which is now recognized as the leading cause of type 2 diabetes. Urban populations, particularly men, are disproportionately affected, and unhealthy dietary habits—especially among children—are a growing concern [2].

A 2024 multicenter study conducted across 16 Bulgarian regions further supports these findings. The study, involving 936 adults aged 20–79, found that the overall prevalence of diabetes in Bulgaria is 16.55%, with 8.22% diagnosed and 8.33% undiagnosed cases. Additionally, 20.6% of the population was found to have prediabetes, including 8.2% with impaired glucose tolerance and 12.4% with impaired fasting glycemia. The data also revealed significant gender and age differences: men accounted for 63.9% of diabetes cases, compared to 36.1% among women. The frequency of the condition increased sharply with age, from 9% in the 20–44 age group to 55.5% in those aged 60–79. These findings highlight the urgent need for improved screening and early detection mechanisms, as the rate of undiagnosed diabetes remains critically high [3]. Globally, the economic impact of diabetes is expected to rise significantly—from USD 1.3 trillion in 2015 to an estimated USD 2.1 trillion by 2030 [4]. While global prevalence estimates now exceed 800 million people living with diabetes, national data provide important context. In Bulgaria, for example, a national diabetes register constructed from over 262 million outpatient records identified 483,836 diabetic patients between 2010 and 2016 [5]. This highlights the scale of the disease burden even in smaller populations and underscores the importance of robust data systems for tracking and managing diabetes.

Effective management of diabetes, particularly type 2 diabetes, relies heavily on long-term adherence to prescribed therapies. However, adherence remains a significant challenge in clinical practice. Studies show that in developed countries, only about 50% of patients with chronic conditions such as diabetes adhere to their prescribed treatment regimens. This rate is often even lower in developing countries due to limited access to healthcare resources [6].

Clinical significance in diabetes adherence and full coverage is in ensuring adequate control of the diseases and prevention of its complications. Adherence is defined as the extent to which a patient’s behavior—taking medication, following a diet, and adopting lifestyle changes—corresponds with agreed recommendations from a healthcare provider. In diabetes care, poor adherence is associated with suboptimal glycemic control, increased risk of complications, and higher healthcare costs [7]. A systematic review found that only 56.2% of patients with type 2 diabetes continued their treatment one year after initiation [6].

Medication dosage form also influences adherence, for example, oral medications like metformin are associated with higher adherence and persistence rates compared to injectable therapies such as insulin 1. Factors contributing to poor adherence include adverse drug effects, complex dosing regimens, psychological resistance to injections, and lack of patient education [6].

Improving adherence is therefore a critical goal in diabetes management. Interventions such as simplifying treatment regimens, enhancing patient–provider communication, and using digital reminders have shown promise in increasing adherence and improving clinical outcomes [6].

Digital health interventions, including electronic prescribing, have been introduced to address these adherence gaps. Since 2023, Bulgaria has implemented mandatory electronic prescriptions (e-prescriptions), enabling real-time tracking of medication use [8,9]. This transition provides a unique opportunity to assess real-world patient behavior across different regions and explore patterns in therapy adoption and prescription renewal habits. In parallel, international evidence shows that digital tools such as app-based reminders, smart pill dispensers, and digital prescription systems improve adherence outcomes significantly [10]. A systematic review and meta-analysis of 26 randomized controlled trials demonstrated that mobile health interventions—often integrated with e-prescription systems—significantly enhanced medication adherence, particularly when they included interactive features and real-time feedback mechanisms [11]. Another review emphasized that mobile applications linked to e-prescribing platforms improved adherence by offering reminders, educational content, and two-way communication between patients and providers [12]. Moreover, electronic prescribing systems themselves have been shown to reduce medication errors and support long-term adherence, especially in chronic disease management [12].

The aim of this study is to perform a comparative analysis of antidiabetic medication prescribing patterns, medication adherence, and NHIF coverage in patients with diabetes across three different types of cities in Bulgaria (Sofia, Plovdiv, and Dobrich and Stara Zagora) using 6-month electronic prescription (e-prescription) data from January to June 2024. As a secondary goal we aimed to evaluate the usefulness and information capability of electronic prescription records in the cities under consideration.

2. Materials and Methods

2.1. Study Design

We performed a retrospective observational study based on 6-month e-prescription records (January–June 2024) [13]. Prescriptions for type 1 and type 2 diabetes were extracted by ICD-10 codes (E10, E11) and analyzed by region. Anonymized patient data was extracted from the national e-prescription system in Bulgaria.

2.2. Settings

We analyzed all antidiabetic prescriptions issued over a six-month period (January–June 2024) across three different types of cities: Sofia (capital), Plovdiv (large regional city), and the combined area of Dobrich and Stara Zagora (smaller towns). Prescriptions were categorized by diabetes type (type 1 or type 2) and linked to a unique patient identifier for the purposes of refill tracking. For Sofia, detailed data was available for both type 1 and type 2 patients. Selection of cities was based only on geographic distribution and willingness to provide information from the electronic databases.

2.3. Observed Variables

Data fields included: date of issuance, date of fulfillment, medication name and formulation, quantity dispensed, number of prescribed items, and reimbursement cost covered by the NHIF. Therapies were classified as traditional (e.g., Metformin, Sulfonylureas, Human Insulins) or innovative (e.g., DPP-4 inhibitors, SGLT2 inhibitors, GLP-1 receptor agonists, insulin analogs). The main outcomes evaluated were: (1) average number of prescriptions per patient in the six-month period; (2) NHIF reimbursement per patient; and (3) regional differences in therapy type and refill behavior.

2.4. Outcomes Measures and Calculations

Refill frequency was used to estimate adherence as well as PDC, while NHIF cost data allowed for financial impact analysis. Patients with multiple prescription events were assumed to maintain therapy adherence.

We adapted a refill frequency adherence which was originally used in pharmacoepidemiologic research to classify patients based on the prescriptions filled during a 6-month period. Patients with 5 or more refills were considered adherent, 2–4 refills as partially adherent, and 1–2 refills as non-adherent.

PDC is widely recognized as a standard measure for evaluating medication adherence in chronic disease management and has been endorsed by the Pharmacy Quality Alliance as a preferred quality indicator. It reflects the percentage of days within a specified observation period during which a patient has access to their prescribed medication. The PDC is typically calculated as follows:

PDC = (Total days covered during the period of interest/Total number of days in the period) × 100. A threshold of ≥80% is commonly used to define adequate adherence for most chronic medications [14].

2.5. Qualitative Analysis

To compare therapeutic strategies across regions, we classified prescription patterns by therapy intensity level. This classification reflects the diversity and complexity of prescribed antidiabetic agents. Regions prescribing a mix of traditional agents (e.g., metformin, sulfonylureas) and innovative therapies such as DPP-4 inhibitors, SGLT2 inhibitors, or fixed-dose combinations were categorized as having moderate therapy intensity. Regions with primarily traditional regimens and limited or no use of modern agents were labeled low intensity, while those prescribing exclusively basic oral monotherapies were considered very low intensity. This structured, non-standardized classification has been applied in prior real-world pharmacoepidemiologic studies to illustrate variability in treatment practices [15].

2.6. Sampling and Risk of Bias

We attempt to collect information for as many prescriptions as possible. The target was to reach 384 patients, which were calculated as representative enough for the diabetic patients of 500,000. Due to immaturity of electronic system, we encounter lots of technical problems, especially in Plovdiv that might create generalizability risk of bias.

2.7. Statistical Analysis

In addition to descriptive statistics (mean, SD, median, 95% CI), independent-sample t-tests were performed to compare refill frequency and NHIF reimbursement between regions. Effect sizes were expressed as Cohen’s d. A significance level of p < 0.05 was used for all tests.

2.8. Ethical Approval

The study is approved by the Ethics Committee for Scientific Research at the Medical University of Sofia (KENIMUS) under contract number: D-111/2024.3.

3. Results

A total of 1132 prescriptions were analyzed across three different types of Bulgarian cities: Sofia (capital), Plovdiv (big region city), and Dobrich and Stara Zagora (smaller cities). Data were analyzed separately by diabetes type and city to assess refill behavior, therapy patterns, medication adherence as PDC, and NHIF reimbursement. NHIF costs reflected the complexity of regimens. In Sofia, 732 prescriptions were reviewed, of which 89 were for type 1 diabetes and 643 for type 2. Patients with type 1 diabetes mostly received insulin analogs (basal-bolus regimens), with some using premixed insulins. Type 2 diabetes therapy was more heterogeneous, including metformin, sulfonylureas, and a significant proportion of prescriptions involving DPP-4 inhibitors, SGLT2 inhibitors, and fixed-dose combinations—indicating moderate prescriptions with innovative therapies. The average number of prescriptions per patient in Sofia was 4.55 over six months. PDC-based adherence showed that 23.1% of type 1 patients and 27.5% of type 2 patients were adherent (PDC ≥ 0.80), with overall mean PDCs of 0.544 and 0.464, respectively. Refill-based adherence (defined as ≥5 fills) showed slightly higher adherence rates: 54% in type 1 and 41% in type 2. NHIF reimbursement data revealed higher average cost per patient in Sofia, consistent with broader use of newer drug classes and combination therapies.

In Dobrich and Stara Zagora, 300 prescriptions were analyzed. Most therapies were traditional—metformin and sulfonylureas—with no prescriptions involving GLP-1 receptor agonists or SGLT2 inhibitors. The average number of prescriptions per patient was 1.33 (SD = 0.57), and the median NHIF reimbursement per patient was 5.40 BGN, indicating minimal adoption of high-cost therapies.

In Plovdiv, 100 prescriptions were extracted and analyzed. All were presumed to involve type 2 diabetes, as therapy patterns exclusively included oral antidiabetic agents. The average number of medications per prescription was 1.05 (SD = 0.22), with low variance, suggesting limited therapeutic diversity dominated by single-drug regimens. NHIF cost data were not available for this region.

In Sofia, type 1 diabetes patients received an average of 6.5 prescriptions over the 6 months, including basal-bolus insulin analogs and some premixed insulins. Type 2 patients (n = 138) averaged 4.55 prescriptions over the 6 months, with therapy patterns including metformin, DPP-4 inhibitors, and SGLT2 inhibitors. Near 10% of patients received fixed-dose combinations.

Dobrich and Stara Zagora showed lower adoption of innovative therapies. Type 1 patients had refill rates between 1 and 6 per patient over the observed period, with traditional human insulin used most often. Type 2 diabetes patients averaged 4.1 refills per patient over the 6 months period, with primary use of metformin and sulfonylureas. No prescriptions included GLP-1 receptor agonists or SGLT2 inhibitors.

In Plovdiv prescriptions for Type 1 diabetes therapy included premixed insulins with an average refill of 6.0 per patient over the observed period. Type 2 patients received mostly metformin and sulfonylureas, with no innovative therapies recorded.

The Regional Comparison

Table 1. The regional comparison table.

Region	Number of Prescriptions Collected	Type 1—Avg. Refills	Type 1—Max Refills	Type 2—Avg. Refills	Type 2—Max Refills	Innovative Therapy Use
Sofia	732	6.5	13	4.55	18	Moderate (DPP-4, SGLT2 present)
Dobrich and Stara Zagora	300	3.9	6	4.1	12	Low (mostly metformin/sulfonylureas)
Plovdiv	100	6.0	8	3.8	9	Very Low (only traditional therapy)

summarizes these values by average refills, maximums, and use of modern therapies.

Among patients with type 1 diabetes in Sofia, there is a concentration of higher refill counts, suggesting more regular access to prescribed therapies and likely stronger adherence. In contrast, individuals in Dobrich and Stara Zagora exhibit a wider variation in refill patterns, with a significant proportion receiving fewer prescriptions, which may reflect gaps in continuity of care or access.

For type 2 diabetes, refill frequencies are generally lower in both regions compared to type 1 but still show a modestly higher average in Sofia. This may be indicative of more proactive or structured follow-up practices in the capital, possibly due to better healthcare infrastructure or more frequent patient monitoring.

Analysis of therapy class distribution among patients with type 2 diabetes revealed regional differences. In Dobrich and Stara Zagora, 44.4% of patients were treated with metformin and 41.7% with sulfonylureas, while only 2.8% received a DPP-4 inhibitor and 2.8% a GLP-1 receptor agonist. In contrast, in Sofia, 11.6% of patients received a DPP-4 inhibitor, and 1.4% were prescribed a GLP-1 receptor agonist. Use of metformin and sulfonylureas was lower in Sofia (6.5% and 2.9%, respectively), suggesting broader adoption of innovative therapies. This shows the differences in prescription patterns and access to modern antidiabetic treatments among the different types of cities.

When examining reimbursement costs covered by the NHIF, the data shows that prescriptions in Sofia tend to involve higher-cost medications, consistent with broader use of combination therapies and more innovative treatment options. On the other hand, prescriptions from Dobrich and Stara Zagora are concentrated around lower reimbursement values, suggesting reliance on traditional monotherapies such as metformin and sulfonylureas, with limited uptake of newer drug classes. Reimbursement data for Plovdiv was not available.

In

Table 2. NHIF cost per prescription by region and diabetes types.

Region	Diabetes Type	Mean Cost (BGN)	SD (BGN)
Sofia	Type 1	114.26	59.88
	Type 2	47.51	92.99
Dobrich and Stara Zagora	Type 2	20.30	45.80

, NHIF cost per prescription by region and diabetes types are presented. In Sofia, prescriptions for type 1 diabetes had the highest mean reimbursement at 114.26 BGN (SD = 59.88), reflecting the use of more complex regiments such as insulin analogs. For type 2 diabetes, Sofia prescriptions averaged 47.51 BGN (SD = 92.99), indicating wide variability due to the mix of traditional and innovative therapies (e.g., DPP-4, SGLT2-inhibitors).

In Dobrich and Stara Zagora, the average NHIF reimbursement for type 2 diabetes prescriptions was significantly lower at 20.30 BGN (SD = 45.80), which aligns with the observed use of more traditional therapies like metformin and sulfonylureas rather than newer higher-cost medications.

This finding supports the treatment pattern in both access and funding in Bulgaria.

Independent-sample t-tests confirmed significant regional differences. For type 2 diabetes, refill frequency was significantly higher in Sofia compared to Dobrich and Stara Zagora (t(≈139) = 10.3, p < 0.001, d = 1.27), and in Plovdiv compared to Dobrich and Stara Zagora (t(≈323) = −55.8, p < 0.001, d = −4.99). A smaller but statistically significant difference was also observed between Sofia and Plovdiv (t(≈137) = 2.05, p = 0.043, d = 0.25). NHIF reimbursement per patient was significantly higher in Sofia compared to Dobrich and Stara Zagora (t(≈159) = 4.38, p < 0.001, d = 0.59). Detailed results are shown in

Table 3. Independent-sample t-tests comparing refill frequency and NHIF reimbursement between regions (Type 2 diabetes patients).

Outcome	Comparison	Group 1 (Mean ± SD, n)	Group 2 (Mean ± SD, n)	t (df)	p	Cohen’s d
Refill frequency	Sofia vs. Dobrich and Stara Zagora	4.55 ± 4.02 (138)	1.33 ± 0.57 (225)	10.3 (≈139)	<0.001	1.27
Refill frequency	Sofia vs. Plovdiv	4.55 ± 4.02 (138)	3.80 ± 0.22 (100)	2.05 (≈137)	0.043	0.25
Refill frequency	Dobrich and Stara Zagora vs. Plovdiv	1.33 ± 0.57 (225)	3.80 ± 0.22 (100)	−55.8 (≈323)	<0.001	−4.99
NHIF reimbursement (BGN)	Sofia vs. Dobrich and Stara Zagora	47.51 ± 92.99 (138)	12.30 ± 21.40 (225)	4.38 (≈159)	<0.001	0.59

.

4. Discussion

4.1. Summary and Interpretation of Key Findings

This real-world study analyzed antidiabetic prescribing patterns and medication adherence using Bulgaria’s national e-prescription system across three different types of cities. Key findings reveal that type 1 diabetes patients in Sofia demonstrated the highest adherence, averaging 6.5 prescriptions over six months, with refill-based adherence of more than 50%. Type 2 patients in the capital showed more heterogeneous behavior (a mean of 4.55 prescriptions; PDC~46%), indicating broader variations in treatment complexity. In contrast, patients from Dobrich, Stara Zagora, and Plovdiv received mainly traditional therapies with minimal adoption of innovative agents and lower refill frequencies, showing limited treatment intensity and potential access barriers.

These findings highlight how urban areas like Sofia provide more structured care, better access to innovative therapies, and possibly more proactive follow-up than the other regional cities.

Bulgaria delays introducing electronic prescriptions in comparison with other European countries. There were technical- and human-related barriers in introducing it. Therefore, the clinical and regulatory significance of our study is to show the authorities and healthcare providers the impact of structured data on prescribing and the importance of pharmacists as a key safeguard of medication therapy management. Our study contributes to the broader discussion in the country on health information technology developments and usage in advancing medication safety and health outcomes for chronic diseases [16,17,18].

The statistical analysis reinforces the descriptive findings. Sofia patients had significantly higher refill frequency than Dobrich and Stara Zagora, and slightly higher than Plovdiv, indicating better therapy continuity in the capital. Plovdiv also showed markedly better adherence than Dobrich and Stara Zagora. NHIF reimbursement per patient was substantially greater in Sofia than in Dobrich and Stara Zagora, reflecting unequal access to innovative therapies. These quantitative results confirm the systemic disparities in diabetes care provision across Bulgaria.

4.2. Role of E-Prescriptions and Comparison with Other Studies

The introduction of electronic prescription in Bulgaria in 2023 marked a major shift toward digitization in healthcare. E-prescriptions are now mandatory for reimbursable medications, enabling easier medication tracking, reducing paper errors, and enabling large-scale, real-world data analysis. Internationally, e-prescriptions are associated with improved medication safety, fewer prescribing errors, and enhanced adherence monitoring [8,9].

In our study, using e-prescription records helped us objectively track how often people refilled their medications in different regions. However, we need to be careful when interpreting these results. While refill rates and PDC are useful indicators, they do not tell us if patients took the medications. Also, our data only includes certain community pharmacy networks. So, if someone filled their prescription at a pharmacy outside our sample, they might appear non-adherent, even if they were adherent. This is especially important in areas with many pharmacies or where healthcare services are spread out.

Compared to studies from Central and Eastern Europe, our findings are consistent: rural and smaller-city patients have more limited access to advanced therapies and tend to receive fewer prescriptions, often due to cost restrictions, provider familiarity, and infrastructure limitations [19,20]. For example, in Romania, only 14% of eligible patients received SGLT2 inhibitors despite guideline recommendations—aligning with our finding that no patients in Dobrich or Plovdiv received such therapies. This is especially important given the growing real-world evidence confirming clinical effectiveness of SGLT2 inhibitors in type 2 diabetes management. They support cardiometabolic health and patients tend to show better adherence to them [21].

Globally, refill-based adherence for chronic medications like antidiabetics is around 50–60% [7,11] and is typically lower in regions with fewer resources. Sofia’s results suggest better alignment with international standards, while the other cities are falling behind. This supports the idea that infrastructure, digital tools, and physician access to training significantly affects therapy delivery.

Digital tools such as telemonitoring, mobile apps, and integrated refill alerts—if added on top of e-prescription platforms—can substantially improve adherence by up to 25% [10,12]. Their integration into Bulgaria’s system could be an affordable and practical way to reduce the gap between care in cities and rural areas.

Additionally, adherence and persistence differ by medication type. Oral therapies like metformin and DPP-4 inhibitors often show higher adherence than injectable agents due to simpler regimens and fewer psychological barriers [6]. This may partly explain the stronger adherence trends seen in Sofia, where a mix of oral agents and fixed-dose combinations was more common.

These patterns are not unique only to Bulgaria. According to the Central and Eastern European Diabetes Forum, infrastructure, physicians access to training, and clear health policies are key factors determining treatment equity across Europe [22,23].

4.3. Strengths and Limitations

This study has several important strengths:

• It is one of the first real-world analyses using national e-prescription data in Bulgaria for diabetes care.
• It includes multiple types of cities, offering a comparative perspective on how geography and infrastructure impact medication, access, and adherence.
• It captures reimbursement data, allowing not just clinical but also economic insights into therapy intensity and public spending.

However, several limitations must be acknowledged:

• The study is not nationally representative; additional cities and rural municipalities were not included. It is necessary to further structure the electronic prescriptions database at the national level and to provide access to it to be able to analyze more economic variables in detail. As pointed out in the methodology, generalizability is still an expected issue. The fact that the study proved the results of many other studies in the field is a positive sign of its validity.
• Reimbursed cost data were unavailable for Plovdiv due to technical reasons and immaturity of the electronic system. This gap negatively affects the ability to draw economic comparisons across all regions. Further broader analysis should be organized when technical problems are solved and information becomes available from all regions.
• As noted, adherence was estimated using indirect methods—refill frequency and PDC—without connection and data of actual clinical or behavioral outcomes. Refill frequency and PDC are indirect measures for medication adherence. Their values depend on repeating the same pharmacy for prescriptions fulfillment. We assume that patients are vising the same pharmacies, which might not be the case and therefore it is a limitation of the study. As noted above national database and access to all data will benefit the results of adherence evaluation.
• Pharmacy data may be missing some information—if patients refilled prescriptions in different pharmacies not captured in our data, potentially underestimating adherence. Finally, we lack patient-level data (e.g., age, comorbidities), which limits deeper risk differences. This limits us to relate prescribing and adherence to demographic and clinical information about the patients, that could be considered as limitation to the study. At the initial point of the analysis, such information was not available, but we will further deepen the analysis in future to include as many variables as possible.
• One limitation of the statistical analysis is the lack of reimbursement data for Plovdiv, which prevented a full three-way comparison across all regions for NHIF costs. Therefore, the reimbursement analysis was restricted to Sofia and Dobrich and Stara Zagora. Future studies should ensure complete data coverage across all regions to allow for more comprehensive economic comparisons.

4.4. Perspective for Clinical Practice and Future Directions

Electronic prescriptions in Bulgaria were initially met with disapproval by doctors. They dismissed them as technical complications that would not improve their work. Gradually, doctors realized the benefits of being able to track not only prescriptions but also their fulfillment in the long term, which would allow for improved patient care. Electronic prescriptions are a tool for patients’ communication and follow-up which is essential for better healthcare services.

Further research should focus on national-level data that brings together prescription records, clinical information, and social factors. Qualitative studies with patients and providers could uncover barriers to innovative therapy adoption, especially in smaller regions. Future analyses should also explore how e-prescription systems can be integrated with mHealth solutions, like reminders and teleconsultations, to personalize adherence support.

Additionally, cross-national comparisons with other EU countries may reveal policy best practices for equitable access to medical care. Longitudinal analyses will be key to understanding whether e-prescriptions improve long-term outcomes like glycemic control, complication rates, and hospitalization.

5. Conclusions

The statistical comparisons confirmed that refill frequency was significantly higher in Sofia and Plovdiv compared to Dobrich and Stara Zagora, with Sofia also slightly exceeding Plovdiv. NHIF reimbursement was substantially greater in Sofia than in Dobrich and Stara Zagora. Within Sofia, the mean NHIF reimbursement per prescription was also significantly higher for type 1 than for type 2 diabetes, consistent with the greater complexity and cost of insulin-based regimens. These findings quantitatively reinforce the descriptive evidence and highlight the persistence of regional inequalities in diabetes care across Bulgaria.

This study highlights significant regional differences in diabetes medication adherence and prescribing patterns across Bulgaria. The findings indicate that capital centers benefit from higher adherence rates—particularly among patients with type 1 diabetes—and greater access to innovative therapies, whereas smaller regions continue to rely more heavily on traditional treatment regimens. These differences reflect systemic challenges, including unequal access to advanced medications, variations in clinical practice, and potential gaps in healthcare provider training. From a clinical practice perspective, adherence follow-up is key to treatment success and healthcare decision-making. Further studies should be undertaken as the national e-prescription database matures, and the resulting insights should be presented to physicians to support improved healthcare provision.