Abstract
Background:
Diabetes mellitus imposes a substantial economic burden on patients and healthcare systems, particularly in low- and middle-income countries.
Objective:
To evaluate longitudinal changes in direct and indirect costs of diabetes care and identify predictors of cost reduction following an educational and empowerment-based intervention.
Methods:
A prospective longitudinal controlled interventional study (quasi-experimental) was conducted over 24 months among adults with type 2 diabetes mellitus at a tertiary care hospital. Patient-level data on direct medical expenses, ancillary costs, and clinical parameters were analyzed using mixed-effects and regression models.
Results:
A significant reduction in direct cost and ancillary fees, such as travel and parking, was observed from baseline to 24 months (p<0.001). Mixed-effects modeling confirmed significant temporal reductions in costs at each follow-up (β = -3,117.05 at 24 months; 95% CI: -3,224.01 to −3,010.09). Furthermore, clusters of moderate, gradual, and high responders were identified. Multivariable regression showed that baseline cost (β = 0.582, p<0.001) and HbA1c (β = −611.53, p < 0.001) as independent predictors of cost reduction.
Conclusion:
Education and empowerment lead to reductions in direct and indirect costs over two years, suggesting improved cost efficiency in diabetes mellitus management.
Introduction
Diabetes mellitus is characterized by uncontrolled hyperglycemia affects millions people worldwide, especially in lower-middle-income countries (LMICs) and contributes to micro- and macrovascular complications (, ). It has been observed that type 2 diabetes mellitus (T2DM) has increased globally, regardless of income or geographic location, over the past 30 years (). According to the World Health Organization (WHO), diabetes accounts for 1.5 million annual deaths across countries (). Recent studies from all over the world estimated that 10.5% of people had diabetes mellitus, with forecasts of 12.2% by the year 2045 (–). According to a 2021 study by the International Diabetes Federation (IDF), 537 million people internationally had diabetes, a number expected to increase to 643 million by 2030 and 783 million by 2045 (). These data highlight the economic burden of diabetes treatment and the challenges of affordability for patients (). The economic impact of diabetes in North India is staggering, with annual treatment costs exceeding ₹50,000–₹1,00,000 per patient (). Complications double or triple expenses, for example, diabetic foot ulcers cost ₹1.5–3 lakhs per hospitalization (). Although Ayushman Bharat-PMJAY covers hospitalization costs, outpatient care (medicines, tests, and doctor consultations) remains largely out-of-pocket (OOP). A recent study found that in Uttar Pradesh, 70% of diabetes care expenses are OOP, pushing 5-7% of households into poverty annually ().
In low- and middle-income countries, the healthcare system faces significant obstacles in diabetes management, with fragmented care delivery, urban-rural disparities, and inadequate insurance coverage. These systemic gaps lead to delayed diagnoses, higher complication rates, and increased healthcare expenditures among low-income households. The present work aimed to assess the impact of education and empowerment on reducing direct and indirect costs of T2DM management.
Methodology
Study design
This prospective longitudinal controlled interventional study (quasi-experimental) recruited 320 participants from a single tertiary care center in North India with the following inclusion criteria. (i) participants aged 25–65 years, irrespective of gender; (ii) confirmed diagnosis of T2DM for less than 6 months through standardized glycaemic criteria including either fasting plasma glucose ≥126 mg/dL (7.0 mmol/L), 2-hour postprandial glucose ≥200 mg/dL (11.1 mmol/L) following 75-g OGTT, HbA1c ≥6.5% (48 mmol/mol) via NGSP-certified assay, or random plasma glucose ≥200 mg/dL (11.1 mmol/L); (iii) absence of insulin therapy at baseline (excluding latent autoimmune diabetes and advanced T2DM cases); and (iv) willing to give written consent. Exclusion criteria followed: (i) diagnoses of type 1 or gestational diabetes, or T2DM duration exceeding 6 months; (ii) current insulin therapy; (iii) severe renal impairment (eGFR <30 mL/min/1.73 m2); (iv) active oncological disease or treatment; (v) psychiatric comorbidities; (vi) pregnancy/lactation; (vii) substance use disorders; (viii) established microvascular complications; and (ix) non-attendance at scheduled study interventions.
Participants were divided into two groups: (i) an intervention group that received structured education and empowerment sessions, and (ii) a control group that continued to receive routine diabetes care in accordance with institutional clinical practice. Allocation was non-randomized and based on patient willingness and availability to participate in the structured education program. Patients who consented to participate in the education and empowerment program formed the intervention group. Meanwhile, those receiving standard clinical care without participation in the structured sessions served as the control group. The study therefore followed a quasi-experimental non-randomized design, with both groups followed prospectively for 24 months. This rigorous selection framework ensured methodological consistency while accounting for potential confounding variables in evaluating cost-effectiveness.
Baseline socio-demographic and economic data were collected using a standardized questionnaire covering age, gender, education level, occupation, monthly family income, rural/urban residence, health insurance status, and distance from the study center. Socio-economic status (SES) was classified using the modified Kuppuswamy scale: low (<₹20,000/year), middle (₹20,000-₹50,000/year), or high (>₹50,000/year). These variables enabled assessment of intervention effects across diverse economic strata and healthcare access patterns.
Education and empowerment implementation strategy
The intervention protocol for enrolled participants included a comprehensive educational curriculum, with weekly, structured 2-hour sessions. The educational modules included several aspects of diabetes management, such as diabetes awareness, individualized counselling, basic educational aspects of diabetes pathophysiology, psychosocial aspects of diabetes, possible complications of diabetes, self-monitoring of blood glucose, behavioural motivational techniques, methodologies of dietary preparation, physical activity regimens, and quality of life enhancement, as per a previously published protocol (). The educational program emphasized the importance of social support systems and was conducted by certified diabetes educators using several instructional techniques, including linguistically appropriate printed materials, digital animations, audio-visual didactic tools, individualized instruction, short educational films, knowledge assessment quizzes, macronutrient identification exercises, and practical sessions.
Diabetes educators utilized concurrent empowerment sessions with a patient-centred focus on cognitive, biophysical, psychological, and social aspects. This paradigm focused on individual value systems, personal beliefs, and perspectives, and was strengths-based rather than deficit-oriented counselling. Collaborative goal setting for glycaemic targets was established through shared decision-making, with flexibility to accommodate individual behavioural patterns and achieve mutual consensus. Personal accountability was encouraged among participants through regular session attendance, and the educators used facilitatory skills such as problem exploration, emotional expression, alternative solution development, consequence exploration, and support for autonomous decision-making. Motivational reinforcement was sustained to ensure long-term adherence to optimal glycaemic control strategies.
Data collection
Data were collected at baseline, 6, 12, 18, and 24 months, with primary outcomes including the cost of treatment (medication, lab tests, equipment, and travel expenses) and clinical efficacy (HbA1c levels). Each component was analysed individually to assess the economic impact, while glycaemic trends were monitored to assess the treatment effectiveness. All the participants completed the follow-up at baseline, 6, 12, 18, and 24 months. No dropouts were noted in the study. A standardized survey was used to collect comprehensive baseline and endline data including socio-economic status, socio-demographic factors, history of diabetes (both personal and familial), comorbidities, patterns of compliance, and non-compliance factors. After completion of the educational and empowering process, glycaemic control was monitored at regular intervals (every quarter) by measuring HbA1c levels. This approach enabled a comprehensive evaluation of costs, both direct and indirect, concerning health outcomes resulting from these interventions. This methodology enabled precise quantification of resource utilization and clinical effectiveness. The study therefore achieved 100% follow-up retention across the 24-month study period.
Direct costs included antidiabetic medications (oral hypoglycemics and insulin preparations), laboratory investigations (HbA1c, lipid profile, renal function tests, and complete blood counts performed quarterly), physician consultations and specialist referrals (endocrinologist, ophthalmologist, podiatrist), and diabetes supplies (glucometers, test strips, lancets, syringes). Indirect costs comprised travel expenses (public transport fares and private vehicle fuel costs to the study center), parking charges at hospital premises, accompanying person expenses (one attendant per visit covering travel costs plus daily wage loss), and productivity losses.
Statistical analysis
The results are presented using appropriate descriptive statistics: normally distributed continuous variables as means ± standard deviations with 95% confidence intervals. The data were analyzed using Python (pandas, statsmodels, scikit-learn). A linear mixed-effects model (LMM) was used to assess temporal trends in cost trends. K-means clustering was used to identify heterogeneous cost trajectories. Multivariable OLS regression was used to determine the predictors of total cost reduction (baseline to 24 months). Confidence intervals were generated using bootstrapping to assess coefficient stability. Statistical significance was defined as p < 0.05 for two-tailed tests. Group differences across time were analysed using linear mixed-effects models with fixed effects for group, time, and the group × time interaction, allowing assessment of differential temporal trends between the intervention and control groups. As complete data were available for all participants across all follow-up time points, no imputation or missing-data handling procedures were required. Linear mixed-effects models were fitted using maximum likelihood estimation with robust standard errors (`cost ~ group×time + (1 + time | patient_id)`). Fixed effects tested intervention impact and temporal trends; random intercepts accounted for baseline patient heterogeneity (ICC reported); random slopes permitted patient-specific cost trajectories. Complete variance components, intraclass correlation coefficient (ICC), and model fit statistics (AIC/BIC) were presented for transparency analyses conducted using Python statsmodels MixedLM. Although a Difference-in-Differences (DiD) analysis was considered, linear mixed-effects modeling (LMM) was preferred due to the study’s five repeated measurements (baseline, 6, 12, 18, and 24 months), complete participant retention, and the presence of non-parallel trends in the outcomes over time. LMM allowed modeling of individual patient trajectories using random effects, accommodated clustered data (ICC = 0.37), and captured time-varying intervention effects through group×time interactions. Sensitivity analyses using the DiD approach produced consistent findings (interaction β = −385.2, p = 0.018; Appendix Table A1), supporting the robustness of the LMM approach.
Results
Table 1 presents comprehensive baseline data showing well balanced groups: mean age ~49 years, 58% male, 47% low SES (<₹10,000/month), 56% rural residents, and only 14% insured. No significant between-group differences (all p>0.05) confirm comparability. A total of 320 adults with type 2 diabetes mellitus were enrolled at baseline, and all participants completed the 24-month follow-up assessments, resulting in a complete dataset across all study time points. The intervention group demonstrated a significant decline in direct costs compared with the control group over the 24-month follow-up period (p < 0.001), while Cohen’s d quantified the standardized between-group effect size. Larger negative d values at later time points indicate a substantial cost advantage associated with the intervention. Each time point shows a significant decrease in direct costs, indicating sustained economic benefits following education and empowerment interventions among patients with T2DM (Table 2; Figure 1).
Table 1
| Characteristic | Intervention (n=160) | Control (n=160) | Total (n=320) | P-value |
|---|---|---|---|---|
| Age (years), mean ± SD | 48.2 ± 9.4 | 49.1 ± 10.1 | 48.6 ± 9.8 | 0.42 |
| Gender, Male, n (%) | 92 (57.5%) | 94 (58.8%) | 186 (58.1%) | 0.78 |
| Education, n (%) | 0.31 | |||
| - Illiterate | 28 (17.5%) | 32 (20.0%) | 60 (18.8%) | |
| - Primary | 52 (32.5%) | 48 (30.0%) | 100 (31.3%) | |
| - Secondary | 56 (35.0%) | 54 (33.8%) | 110 (34.4%) | |
| - Graduate+ | 24 (15.0%) | 26 (16.2%) | 50 (15.6%) | |
| Monthly income (₹), n (%) | 0.38 | |||
| - <10,000 (Low SES) | 72 (45.0%) | 76 (47.5%) | 148 (46.3%) | |
| - 10,000-30,000 (Middle) | 72 (45.0%) | 68 (42.5%) | 140 (43.8%) | |
| - >30,000 (High SES) | 16 (10.0%) | 16 (10.0%) | 32 (10.0%) | |
| Rural residence, n (%) | 88 (55.0%) | 92 (57.5%) | 180 (56.3%) | 0.67 |
| Distance from center (km), median (IQR) | 28 (15-45) | 30 (18-48) | 29 (16-46) | 0.55 |
| Health insurance, n (%) | 24 (15.0%) | 20 (12.5%) | 44 (13.8%) | 0.52 |
Baseline socio-economic and demographic characteristics (n=320).
*p-values from χ²/Fisher’s exact test (categorical) and independent t-test/Mann-Whitney U test (continuous) comparing intervention vs. control groups.
Table 2
| Timepoint (months) | Intervention group mean ± SD (₹) | Control group mean ± SD (₹) | Absolute difference (₹) | Percent change (cases vs controls) | Cohen’s d (standardized effect) |
|---|---|---|---|---|---|
| Baseline | 4243 ± 1 205 | 4123 ± 1 150 | +120 | +2.91% | +0.10 |
| 6 | 3566 ± 980 | 3845 ± 1 050 | −279 | −7.26% | −0.28 |
| 12 | 2993 ± 650 | 3410 ± 920 | −417 | −12.23% | −0.52 |
| 18 | 2380 ± 420 | 3030 ± 850 | −650 | −21.45% | −0.97 |
| 24 | 1930 ± 310 | 2625 ± 800 | −695 | −26.48% | −1.15 |
Temporal trend of direct costs in intervention and control groups over 24 months, showing a significant reduction in the cost.
Figure 1
Descriptive inspection revealed marked declines in mean patient-level costs over time. The LMM demonstrated strong, statistically significant reductions relative to baseline: estimated baseline mean was ₹5,160.15 (95% CI ₹4,983.51–₹5,336.79). Estimated mean differences versus baseline were −₹1,476.70 at 6 months (95% CI −₹1,583.66 to −₹1,369.74; p < 0.001), −₹2,209.49 at 12 months (95% CI −₹2,316.45 to −₹2,102.53; p < 0.001), −₹2,671.00 at 18 months (95% CI −₹2,777.96 to −₹2,564.04; p < 0.001), and −₹3,117.05 at 24 months (95% CI −₹3,224.01 to −₹3,010.09; p < 0.001). These results indicate a substantial and progressive reduction in expenditure across the two-year follow-up.
In our study, the cost-trajectory clustering approach revealed three major clusters within the study population. Cluster 0 comprised 95 patients with moderate baseline costs and a steady reduction. Cluster 1 included 115 T2DM patients who presented with a gradual, stable decline in direct costs, while Cluster 2 included 109 T2DM patients who showed a sharp, high-magnitude cost reduction (Figure 2). In multivariable regression analysis, a higher baseline cost (β = 0.582, p < 0.001) and higher baseline HbA1c (β = −611.53 per % HbA1c, p < 0.001) independently predicted greater cost reduction over 24 months. Baseline HbA1c was also significantly associated with cost reduction (β = −611.53 per % HbA1c; bootstrap 95% CI −811.17 to −403.31; p < 0.001), suggesting greater clinical burden at entry predicted larger subsequent cost declines. Age, travel, and parking costs were not significant predictors in the study population. Bootstrap confidence intervals further confirmed the robustness of these associations. Age and ancillary costs (travel and parking) were not significant predictors. These findings were consistent with estimates derived from mixed effects modeling. Table 3 and Figure 3 show a substantial decrease in ancillary (indirect) cost in the study population (p<0.001).
Figure 2
Table 3
| Predictor | β estimate | Std. error | T value | P value | Bootstrap 95% CI | Interpretation |
|---|---|---|---|---|---|---|
| Intercept | 4597.56 | 984.80 | 4.67 | <0.001 | 2594.8 – 6308.2 | Reference baseline level |
| Baseline Cost | 0.582 | 0.037 | 15.71 | <0.001 | 0.39 – 0.76 | Larger baseline cost predicts larger absolute reduction |
| Age | 2.79 | 11.09 | 0.25 | 0.801 | −17.8 – 23.7 | Age is not significantly associated with change |
| HbA1c (Baseline) | −611.53 | 114.42 | −5.35 | <0.001 | −811.2 – −403.3 | Higher HbA1c predicts greater cost reduction |
| Travel cost | −0.38 | 0.41 | −0.92 | 0.356 | −1.53 – 0.18 | Non-significant negative trend |
| Parking fees | 0.23 | 1.37 | 0.17 | 0.865 | −1.61 – 2.51 | No meaningful effect |
Indirect cost of travel and parking charges showed financial cost savings in the study population.
Figure 3
The results of the present study showed progressively increasing statistical significance, thereby confirming sustained effects beyond six months in T2DM patients who received education and empowerment (Table 4). Furthermore, the subgroup comparison of odds ratios for glycaemic control showed statistically significant improvement after education and empowerment in T2DM patients (Table 5; Figure 4). Table 6 presents standardized effect sizes derived from mean cost differences using Cohen’s conventional thresholds, and large effects were observed after one year of the intervention (Figure 5). A mixed-effects model was fitted for participants’ group, time point, and their interaction, and which identified a negative interaction coefficient indicating a steeper decline in cost over time in the education-and-empowerment group compared to the control group (Table 7). The complete mixed-effects model demonstrated substantial patient-level heterogeneity, with random intercepts explaining 37% of total variance (ICC = 0.37) and random slopes indicating moderate patient-specific trajectory variation (ρ=0.42). Model fit was excellent (AIC = 24,932.4). The significant group-by-24-month interaction (-420.1, 95% CI -781 to -59, p=0.022) confirms the intervention group’s steeper cost trajectory compared to controls, robust despite individual variability in baseline costs and temporal patterns. Difference-in-Differences sensitivity analysis: Aggregated pre- and post-intervention Difference-in-Differences analysis supported the primary findings of the LMM, demonstrating a significantly greater reduction in annualized costs in the intervention group compared with controls (−2,095.04 vs. −695.02, respectively). The estimated DiD effect was −1,400.02 (p = 0.018), indicating a 2.6-fold greater reduction in cost in the intervention arm. Table 8 presents the linear mixed-effects model for changes in the direct cost pattern, showing a statistically significant reduction in direct costs for patients with T2DM after receiving education and empowerment. We assessed the regression coefficients using bootstrap confidence intervals (Table 9) and found a significant association between baseline cost and HbA1c, thereby highlighting the robustness of these predictors.
Table 4
| Timepoint (months) | Reported OR | 95% CI | log (OR) | SE (log OR) | z-Statistic | Two-tailed p |
|---|---|---|---|---|---|---|
| Baseline | 1.12 | 0.75 – 1.68 | 0.113 | 0.206 | 0.55 | 0.582 |
| 6 | 2.31 | 1.52 – 3.52 | 0.837 | 0.214 | 3.91 | 9.3 × 10-5 |
| 12 | 4.25 | 2.71 – 6.67 | 1.447 | 0.230 | 6.30 | 3.0 × 10-¹0 |
| 18 | 6.50 | 3.89 – 10.9 | 1.872 | 0.263 | 7.12 | 1.1 × 10-¹² |
| 24 | 9.75 | 5.42 – 17.5 | 2.277 | 0.299 | 7.62 | 2.6 × 10-¹4 |
Odds ratio depicting the likelihood of achieving cost reduction.
Table 5
| Timepoint (months) | HbA1c < 7% OR (95% CI) | P value | HbA1c ≥ 7% OR (95% CI) | P value |
|---|---|---|---|---|
| Baseline | 1.05 (0.70–1.56) | 0.82 | 1.09 (0.73–1.63) | 0.74 |
| 6 | 2.18 (1.36–3.49) | 0.0016 | 2.25 (1.33–3.80) | 0.0027 |
| 12 | 3.92 (2.34–6.55) | < 0.001 | 4.09 (2.48–6.73) | < 0.001 |
| 18 | 6.01 (3.43–10.53) | < 0.001 | 6.34 (3.55–11.31) | < 0.001 |
| 24 | 9.11 (5.04–16.45) | < 0.001 | 9.46 (5.24–17.08) | < 0.001 |
Subgroup odds ratio analysis of study participants based on glycemic control.
Figure 4
Table 6
| Effect magnitude category | Definition (Cohen’s d) | Observed timepoints | Interpretation |
|---|---|---|---|
| Small | 0.2 ≤ | d | < 0.5 |
| Medium | 0.5 ≤ | d | < 0.8 |
| Large | d | ≥ 0.8 |
Standardized effect sizes for direct cost reduction in the study participants who received education and empowerment.
Figure 5
Table 7
| Fixed effects | Estimate | SE | 95% CI | z | P-value |
|---|---|---|---|---|---|
| Intercept (Control, Baseline) | 4,120.85 | 85.2 | 3,953–4,287 | 48.3 | <0.001 |
| Intervention Group | -120.13 | 130.4 | -377 to 137 | -0.92 | 0.357 |
| Time: 6 months | -280.11 | 110.2 | -496 to -64 | -2.54 | 0.011 |
| Time: 12 months | -417.23 | 115.8 | -644 to -190 | -3.60 | <0.001 |
| Time: 18 months | -650.44 | 122.1 | -890 to -411 | -5.33 | <0.001 |
| Time: 24 months | -695.02 | 128.9 | -948 to -442 | -5.39 | <0.001 |
| Group×Time: 24 months | -420.1 | 183.2 | -781 to -59 | -2.29 | 0.022 |
| Random Effects | Variance | SD | ICC/ρ | ||
| Patient Intercept | 1,245.3 | 35.29 | 0.37 | ||
| Patient Time-Slope | 892.7 | 29.88 | 0.42 | ||
| Residual | 2,108.4 | 45.92 | – | ||
Mixed-effects model evaluating group, time, and group × time interaction effects on cost outcomes.
Model specification: cost ~ group×time + (1 + time | patient_id) using maximum likelihood estimation. Bold = primary intervention effect of interest. ICC = intraclass correlation coefficient (37% patient-level clustering). Random slopes capture heterogeneous patient trajectories (patient-time correlation: ρ = 0.42). Significant Group×24-month interaction confirms a steeper cost decline in the intervention arm—analysis via Python’s statsmodels MixedLM.
Model fit: Log-likelihood = -12,456.2; AIC = 24,932.4; BIC = 25,089.1.
Table 8
| Fixed effect | Estimate (₹) | 95% CI (Lower–Upper) | z Value | p Value | Interpretation |
|---|---|---|---|---|---|
| Intercept (Baseline) | 5160.15 | 4983.51 – 5336.79 | 57.26 | <0.001 | Mean baseline cost per participant |
| Month 6 | −1476.70 | −1583.66 – −1369.74 | −27.06 | <0.001 | Cost decreased significantly by 6 months |
| Month 12 | −2209.49 | −2316.45 – −2102.53 | −40.49 | <0.001 | Continued decline from baseline |
| Month 18 | −2671.00 | −2777.96 – −2564.04 | −48.94 | <0.001 | Persistent reduction over time |
| Month 24 | −3117.05 | −3224.01 – −3010.09 | −57.12 | <0.001 | Largest and most sustained cost reduction |
Linear mixed-effects model for change in direct medical costs over time.
Table 9
| Parameter | Estimate | 2.5th percentile | 97.5th percentile | Significance |
|---|---|---|---|---|
| Intercept | 4597.56 | 2594.84 | 6308.20 | Significant |
| Baseline Cost | 0.582 | 0.393 | 0.761 | Significant |
| Age | 2.79 | −17.77 | 23.73 | Non-significant |
| HbA1c (Baseline) | −611.53 | −811.17 | −403.31 | Significant |
| Travel Cost | −0.38 | −1.53 | 0.18 | Non-significant |
| Parking Fees | 0.23 | −1.61 | 2.51 | Non-significant |
Bootstrap confidence intervals for regression coefficients.
Discussion
This two-year, prospective, longitudinal, controlled, interventional study (quasi-experimental) demonstrated a robust and significant decline in diabetes-related expenditure, including both direct and indirect costs, following the implementation of education and empowerment approaches. Direct costs decreased by 60% compared to the baseline, while indirect costs declined substantially. The findings of our study align with the observations of previously published studies by Grover et al. (2022) and Kumar et al. (2023), which reported similar economic benefits following educational interventions for chronic disease management (, ). The results are consistent with previous findings, in which structured diabetes self-management education and empowerment (DSMEE) have been shown to reduce healthcare expenditure significantly. Previously published meta-analyses showed that DSMEE improved glycaemic control while reducing healthcare expenditure (–). Chrvala et al. (2016) and Powers et al. (2020) also demonstrated the beneficial role of education in sustained glycaemic control; Pal et al. (2013) also reported similar findings in the Indian context ().
Regression analysis in our study further demonstrated that baseline cost and HbA1c were major predictors of cost change, concluding that patients with poor glycaemic control and higher medical-related expenditure derived the greatest economic benefits. Furthermore, Gilmer et al. (2005), Zhang et al. (2010), and Li et al. (2013) demonstrated that a 1% reduction in HbA1c is associated with a 13% reduction in expenditure for care for patients with T2DM (–). Similar observations were made by Stratton et al. (2000) and Nichols et al. (2007) (, ).
Our study demonstrates sustained reductions in diabetes-related costs over time, as evidenced by odds ratios rising from 2.31 at six months to 9.75 at 24 months. Such behavioural changes among patients changes the empowerment interventions, often need time to translate into tangible cost outcomes, which is corroborated by a previously published study of Duncan et al. (2011) and Steinsbekk et al. (2012) who reported that long-term interventions have a significant effect on glycaemic control and healthcare costs (, ). Subgroup analysis of the study group showed a substantial reduction in costs among patients with both poor and good baseline glycaemic control; these findings are supported by previous studies by Khunti et al. (2018) and Sturt et al. (2015) (, ).
Ancillary costs, including travel and parking, also showed a non-significant decrease in T2DM patients, likely due to a lower frequency of hospital visits driven by better glycaemic control and fewer diabetes-related complications. Grover et al. (2019) and Joshi & Anjana (2021) reported that educational and digital interventions significantly reduced the travel-related burden in Indian patients (, –). Education and empowerment contribute to substantial reductions in direct and indirect costs and to improved patient-oriented outcomes, thereby reducing the overall healthcare burden on patients. From a health systems perspective, the observed cost reductions suggest that structured education and empowerment interventions could complement existing public-sector initiatives such as the NP-NCD. Integration of such interventions within primary and secondary care settings may enhance cost efficiency while supporting long-term glycaemic control, particularly in resource-constrained environments. Further long-term follow-up studies are needed to assess the scalability of such interventions at the national level through comprehensive evaluation (–).
Conclusion
This study demonstrated sustained reductions in direct and indirect costs through education and empowerment interventions. Over time, higher odds ratios indicated a positive impact of the intervention on glycaemic control and cost reduction. Although the findings support the cost-effectiveness of structured self-management interventions, their overall impact will depend on their integration into existing healthcare frameworks. Integrating self-management interventions into existing tertiary and primary healthcare frameworks of tertiary and primary healthcare could improve cost-effectiveness, efficiency, and empowerment in different communities.
Statements
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Ethics statement
The studies involving humans were approved by Ethics Committee, Aligarh Muslim University. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.
Author contributions
SR: Conceptualization, Writing – original draft. SK: Validation, Conceptualization, Project administration, Writing – review & editing. HA: Writing – review & editing, Conceptualization, Project administration, Resources. DK: Methodology, Conceptualization, Supervision, Writing – review & editing. AR: Conceptualization, Validation, Data curation, Resources, Visualization, Formal analysis, Writing – review & editing.
Funding
The author(s) declared that financial support was not received for this work and/or its publication.
Acknowledgments
The authors are thankful to the staff of Rajiv Gandhi Center for Diabetes and Endocrinology, J.N. Medical College, Aligarh Muslim University, for their support at every step.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The author AR declared that they were an editorial board member of Frontiers, at the time of submission. This had no impact on the peer review process and the final decision.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
1
World Health Organization. Diabetes (2022). Available online at: https://www.who.int/news-room/fact-sheets/detail/diabetes (Accessed March 7, 2026).
2
MaglianoDJBoykoEJIDF Diabetes Atlas 10th edition scientific committee. Chapter 3, global pictur. In: Idf Diabetes Atlas, 10th edition. International Diabetes Federation, Brussels (2021).
3
FloodDSeiglieJADunnMTschidaSTheilmannMMarcusMEet al. The state of diabetes treatment coverage in 55 low-income and middle-income countries: a cross-sectional study of nationally representative, individual-level data in 680,102 adults. Lancet Health Longev. (2021) 2:e340–51. doi: 10.1016/s2666-7568(21)00089-1
4
AkhtarSNasirJAAbbasTSarwarA. Diabetes in Pakistan: A systematic review and meta-analysis. Pak J Med Sci. (2019) 35:1173–8. doi: 10.12669/pjms.35.4.194
5
GillaniAHAzizMMMasoodISaqibAYangCChangJet al. Direct and indirect cost of diabetes care among patients with type 2 diabetes in private clinics: a multicenter study in Punjab, Pakistan. Expert Rev Pharmacoecon Outcomes Res. (2018) 18:647–53. doi: 10.1080/14737167.2018.1503953
6
SunHSaeediPKarurangaSPinkepankMOgurtsovaKDuncanBBet al. IDF Diabetes Atlas: Global, regional, and country-level diabetes prevalence estimates for 2021 and projections for 2045. Diabetes Res Clin Pract. (2022) 183:109119. doi: 10.1016/j.diabres.2021.109119
7
AlsaediYEAlmalkiAAAlqurashiRDAltwairqiRSAlmalkiDMAlshehriKMet al. Assessment of type II diabetes patients' Caregivers' Burnout level: A cross-sectional study in taif, Saudi Arabia. Diabetes Metab Syndr Obes. (2022) 15:1091–9. doi: 10.2147/DMSO.S357340
8
ButtMDOngSCRafiqAKalamMNSajjadAAbdullahMet al. A systematic review of the economic burden of diabetes mellitus: contrasting perspectives from high and low middle-income countries. J Pharm Policy Pract. (2024) 17:2322107. doi: 10.1080/20523211.2024.2322107
9
RaghavSKumarSAshrafHKhannaP. Economic burden of diabetes care in a tertiary care hospital-based population: A prospective observational study. Int J Environ Sci. (2025) 11:3591–9.
10
RaghavSKumarSAshrafHKhannaP. Cost-effectiveness of the 3E model in diabetes management: a machine learning approach to assess long-term economic impact. Front Public Health. (2025) 13:1571546. doi: 10.3389/fpubh.2025.1571546
11
GroverSAvasthiABhansaliAChakrabartiSKulharaP. Cost of ambulatory care of diabetes mellitus: a study from north India. Postgrad Med J. (2005) 81:391–5. doi: 10.1136/pgmj.2004.024299
12
KumarANagpalJBhartiaA. Direct cost of ambulatory care of type 2 diabetes in the middle- and high-income group populace of Delhi: the DEDICOM survey. J Assoc Phys India. (2008) 56:667–74.
13
ChrvalaCASherrDLipmanRD. Diabetes self-management education for adults with type 2 diabetes mellitus: A systematic review of the effect on glycemic control. Patient Educ Couns. (2016) 99:926–43. doi: 10.1016/j.pec.2015.11.003
14
PowersMABardsleyJKCypressMFunnellMMHarmsDHess-FischlAet al. Diabetes self-management education and support in adults with type 2 diabetes: a consensus report of the American Diabetes Association, the Association of Diabetes Care & Education Specialists, the Academy of Nutrition and Dietetics, the American Academy of Family Physicians, the American Academy of PAs, the American Association of Nurse Practitioners, and the American Pharmacists Association. Diabetes Care. (2020) 43(7):1636–49. doi: 10.2337/dci20-0023
15
NorrisSLLauJSmithSJSchmidCHEngelgauMM. Self-management education for adults with type 2 diabetes: A meta-analysis of the effect on glycemic control. Diabetes Care. (2002) 25:1159–71. doi: 10.2337/diacare.25.7.1159
16
SteinsbekkARyggLLisuloMRiseMBFretheimA. Group based diabetes self-management education compared to routine treatment for people with type 2 diabetes mellitus: A systematic review with meta-analysis. BMC Health Serv Res. (2012) 12:213. doi: 10.1186/1472-6963-12-213
17
KerrDAhnDWakiKWangJBreznenBKlonoffDC. Digital interventions for self-management of type 2 diabetes mellitus: systematic literature review and meta-analysis. J Med Internet Res. (2024) 26:e55757. doi: 10.2196/55757
18
GilmerTPO'ConnorPJRushWACrainALWhitebirdRRHansonAMet al. Predictors of health care costs in adults with diabetes. Diabetes Care. (2005) 28(1):59–64. doi: 10.2337/diacare.28.1.59
19
ZhangYDallTMMannSEChenYMartinJMooreVet al. The economic costs of undiagnosed diabetes. Popul Health Manag. (2010) 13(2):95–101. doi: 10.1089/pop.2009.12202
20
LiRZhangPBarkerLEChowdhuryFMZhangX. Cost-effectiveness of interventions to prevent and control diabetes mellitus: A systematic review. Diabetes Care. (2013) 36:922–30.
21
StrattonIMAdlerAINeilHAWMatthewsDRManleySECullCAet al. Association of glycaemia with macrovascular and microvascular complications of type 2 diabetes (UKPDS 35): prospective observational study. BMJ. (2000) 321(7258):405–12. doi: 10.1136/bmj.321.7258.405
22
NicholsGABrownJB. The impact of glycemic control on hospitalization costs. Diabetes Care. (2007) 30:1878–83. doi: 10.2337/dc06-2197
23
DuncanIAhmedTLiQEStetsonBRuggieroLBurtonKet al. Assessing the value of the diabetes educator. Diabetes Educ. (2011) 37(5):638–57. doi: 10.1177/0145721711416256
24
SteinsbekkARyggLLisuloMRiseMBFretheimA. Group based diabetes self-management education versus routine treatment. BMC Health Serv Res. (2012) 12:213. doi: 10.1186/1472-6963-12-213
25
KhuntiKDaviesMJKalraS. Self-management of type 2 diabetes in low- and middle-income countries. Lancet Diabetes Endocrinol. (2018) 6:885–7.
26
SturtJDennickKHesslerDHunterBMOliverJ. The efficacy of a diabetes self-management education intervention on glycemic control: A meta-analysis. Diabetes Med. (2015) 32:759–69. doi: 10.1111/dme.12685
27
JoshiSRAnjanaRM. The economic burden of diabetes in India: A review. Int J Diabetes Dev Ctries. (2021) 41:385–93.
28
KumpatlaSKothandanHTharkarSViswanathanV. The costs of treating long-term diabetes complications in India. Indian J Endocrinol Metab. (2017) 21:92–7. doi: 10.4103/2230-8210.196013
29
AnjanaRMPradeepaRDeepaMDattaMSudhaVUnnikrishnanRet al. Prevalence of diabetes and prediabetes (impaired fasting glucose and/or impaired glucose tolerance) in urban and rural India: phase I results of the Indian Council of Medical Research–India Diabetes (ICMR–INDIAB) Study. Diabetologia. (2011) 54:3022–7. doi: 10.1007/s00125-011-2291-5
30
TeljeurCMoranPSWalsheSSmithSMCianciFMurphyLet al. Economic evaluation of chronic disease self-management for people with diabetes: a systematic review. Diabet Med. (2017) 34(8):1040–9. doi: 10.1111/dme.13281
31
PolonskyWHFisherLHesslerD. Impact of self-management education on quality of life and healthcare utilization. Patient Educ Couns. (2017) 100:1659–66. doi: 10.1016/j.pec.2017.04.003
32
BarberJThompsonS. Analysis of cost data in randomized trials. BMJ. (2004) 329:1197–200.
33
LiRZhangPBarkerLE. Cost-effectiveness of diabetes prevention programs. Diabetes Care. (2019) 42:2013–21.
34
MohanVSandeepSDeepaMShahBVargheseC. Epidemiology of type 2 diabetes: Indian scenario. Indian J Med Res. (2022) 155:17–30.
35
HermanWHYeWGriffinSJSimmonsRK. Economic analysis of diabetes prevention and control interventions. Lancet Diabetes Endocrinol. (2015) 3:835–45.
36
de JongLALiXEmamipourSvan der WerfSPostmaMJvan DijkPRet al. Evaluating the cost-utility of continuous glucose monitoring in individuals with type 1 diabetes: a systematic review of the methods and quality of studies using decision models or empirical data. Pharmacoeconomics. (2024) 42(9):929–53. doi: 10.1007/s40273-024-01388-6
37
GlantzNMDuncanIAhmedTFanLReedBLKaliraiSet al. Racial and ethnic disparities in the burden and cost of diabetes for US Medicare beneficiaries. Health Equity. (2019) 3(1):211–8. doi: 10.1089/heq.2019.0004
38
BommerCHeesemannESagalovaVManne-GoehlerJAtunRBärnighausenTet al. The global economic burden of diabetes in adults aged 20-79 years: a cost-of-illness study. Lancet Diabetes Endocrinol. (2017) 5(6):423–30. doi: 10.1016/S2213-8587(17)30097-9
39
American Diabetes Association. Economic costs of diabetes in the U.S. in 2017. Diabetes Care. (2018) 41:917–28. doi: 10.2337/dci18-0007
40
AcharyaLDRauNRUdupaNRajanMSVijayanarayanaK. Assessment of cost of illness for diabetic patients in South Indian tertiary care hospital. J Pharm Bioallied Sci. (2016) 8(4):314–20. doi: 10.4103/0975-7406.199336
Summary
Keywords
diabetes mellitus, direct cost, education, empowerment, indirect cost
Citation
Raghav S, Kumar S, Ashraf H, Khanna P and Raghav A (2026) Economic impact of education and empowerment interventions in diabetes care: a two-year prospective longitudinal analysis in a tertiary care population. Front. Clin. Diabetes Healthc. 7:1825788. doi: 10.3389/fcdhc.2026.1825788
Received
08 March 2026
Revised
08 May 2026
Accepted
17 June 2026
Published
01 July 2026
Volume
7 - 2026
Edited by
Anca Pantea Stoian, Carol Davila University of Medicine and Pharmacy, Romania
Reviewed by
Sumit Oberoi, Symbiosis School of Economics, India
Suman Baishnab, Amity Institute of Pharmacy, India
Updates
Copyright
© 2026 Raghav, Kumar, Ashraf, Khanna and Raghav.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Supriya Raghav, 2020phdoddsupriya8972@poornima.edu.in; raghav.supriya@gmail.com
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.