Digital Innovation and Integrated Care: Improving the Effectiveness of Type 2 Diabetes Care in Western Sydney, Australia Abstract: Introduction: The COVID-19 pandemic catalysed the adoption of digital technologies in healthcare. This study assesses a digital-first integrated care model for Type 2 diabetes management in Western Sydney, utilizing Continuous Glucose Monitoring (CGM) and virtual Diabetes Case Conferences (DCC) involving the patient, General Practitioner (GP), Diabetes Specialist and Diabetes Educator at the same time, Methods: In 2020, 833 Type 2 diabetes patients were seen at Western Sydney Diabetes (WSD) clinics. An early cohort of 103 patients was evaluated before and after participation in a virtual DCC, incorporating CGM (Preprint)

Abstract

Diabetes presents a significant global health challenge, disproportionately affecting regions such as Western Sydney, where prevalence exceeds national averages (1, 2). Recognizing this, Western Sydney Diabetes (WSD) has pioneered a hybrid model of integrated care, designed to address the complex needs of over 100,000 local type 2 diabetes patients (9). This model leverages digital technologies—including CGM, virtual DCC, and a suite of digital educational resources—to enhance the coordination between primary and specialist care and empower General Practitioners (GPs) with the tools and knowledge to manage diabetes more effectively. At the heart of WSD's approach is the strategic use of CGM, recognized globally for its effectiveness in diabetes management (5, 6). By integrating real-time glucose data into patient care, WSD facilitates a comprehensive, collaborative treatment planning process. This process not only involves the patient and their GP but also draws on the expertise of dietitians, diabetes educators, and other clinicians. Through virtual case conferencing and access to over 130 patient education videos, WSD's model promotes a deeper understanding of diabetes self-management, encouraging lifestyle modifications that significantly impact glycemic control.

This study aims to retrospectively analyse the outcomes for patients managed under WSD's innovative clinical model, focusing on the use of CGM and other technological solutions in regular practice for individuals with complex Type 2 Diabetes. By examining the model's efficacy in improving patient care and enhancing GP capacity, the research seeks to provide valuable insights into the potential of integrated, technology-enabled care models to address the diabetes burden in high-prevalence regions.

Study Design and Participants: This retrospective cohort study analyzed patient records from the Western Sydney Diabetes (WSD) complex diabetes clinics for the first quarter of 2022. Eligible patients had participated in at least one initial and one follow-up appointment at a WSD hybrid clinic, utilized CGM, and were subsequently discharged to their General Practitioner (GP) for ongoing diabetes management. Discharge criteria included the patient’s stability and the GP’s readiness to assume care, reflecting WSD’s objective to enhance GP capacity for diabetes management in the community. Data Collection: Data were extracted from the electronic medical records, focusing on clinical and demographic information, medication regimens, CGM data, and dietary interventions. A convenience sampling method was employed, culminating in the collection of 103 patient records out of 400 patients from early 2022. Data were securely stored in a dedicated database for analysis. Diabetes Case Conferences (DCC): Central to the study was the evaluation of DCCs, which are collaborative meetings involving the patient, GP, endocrinologist, dietitian, diabetes educator, and other clinicians as needed. These conferences were convened 2-3 times per patient, as required, until consensus was reached on transitioning care back to the GP.Statistical Analysis: Data analysis was conducted using Stata 15. We employed t-tests and chi-squared tests to compare pre- and post-intervention clinical metrics for continuous and categorical variables, respectively. Key CGM metrics analysed included Time in Range (TIR) and the Glucose Management Indicator (GMI). Changes in medication, especially insulin dosages, were also examined. Qualitative Analysis: In addition to quantitative metrics, the study includes a qualitative review of CGM’s clinical utility in managing Type 2 Diabetes, assessing its impact on patient care and treatment outcomes. This qualitative review was conducted as a series of informal interviews with patients and GPs involved in the clinic. Themes from these discussions were collated and reported back by the WSD team.

Patient Demographics and Baseline Characteristics: The study analysed 103 patient records, with 81 patients having complete HbA1c data both at baseline and follow-up. Table 1 provides a comprehensive overview of patient demographics. Glycaemic Control: A significant reduction in HbA1c levels was observed, decreasing from an average of 9.6% at baseline to 8.2% at follow-up. This change represents a mean reduction of 1.4% (95% CI 1.03-1.82%, p<0.0001). Continuous Glucose Monitoring (CGM) Metrics: CGM data highlighted an improvement in Time in Range (TIR), which increased from 46% at baseline to 73% at the follow-up (95% CI 20-32%, p<0.0001). Additionally, the Glucose Management Indicator (GMI) decreased from 7.97% to 6.94%, a reduction of 1.03% (95% CI 0.55-1.2%, p<0.0001), indicating enhanced glucose control. Time above range (TAR) High and TAR-very high reduced by 10.06% and 16.95% respectively. Time below range (TBR)-Low and TBR-very low also reduced by 0.1% and 0.44% respectively. (Table 2) Medication Adjustments: Substantial changes in medication regimens were noted post-consultation. The use of sulfonylureas decreased significantly, with only 17 % of patients (5 of 28) continuing this medication at follow-up. Conversely, 32 patients had a GLP-1 receptor agonist (GLP1 RA) added to their regimen, while 9 discontinued its use. (Figure 2) The proportion of patients on insulin therapy increased from 60% to 73%, although the average total daily insulin dose did not show a statistically significant change (average decrease of 5 units, p=0.10). Patient and GP Experiences: Both patients and General Practitioners (GPs) expressed highly positive feedback regarding their experiences with the virtual clinic model. The use of Continuous Glucose Monitoring (CGM) was particularly highlighted as beneficial. Patients reported improved understanding and management of their condition, attributing this to the insights gained from real-time glucose monitoring. GPs noted the virtual clinic facilitated more effective communication, collaborative care planning, and enhanced their ability to manage diabetes in a primary care setting. This feedback underscores the value of integrating digital health technologies in chronic disease management, aligning with the study's aim to evaluate the impact of a technologically enabled care model on diabetes management. Table 1 – demographics and diabetes medications at baseline for the sample. Data given as mean (SD) or n (%). All (n=61) (n=42) n=103 Age in years (SD) 65.1 (10.0) 61.1 (13.1) - Diabetes Medications At Baseline Metformin 42 (68.9%) 34 (81.0%) 76 (73.8%) Sulfonylurea 24 (39.3%) 4 (9.5%) 28 (27.2%) DPP4I 27 (44.3%) 11 (26.2%) 38 (36.9%) SGLT2I 33 (54.1%) 12 (28.6%) 45 (43.7%) GLP1RA 13 (21.3%) 13 (31.0%) 26 (25.2%) Insulin 33 (54.1%) 29 (69.0%) 62 (60.2%) Mean Insulin TDD in units (SD) 33.6 (42.5) 39.9 (41.0) Mean Baseline HbA1c in % (SD) 9.5 (1.5) 9.9 (1.7) Table 2 Changes in CGM metrics comparing the initial review and prior to discharge CGM metrics Initial review (mean SD) Prior to discharge (mean SD) Change P value Sensor active time (%) 75.44 78.61 +3.17 0.84 Glucose variability (%) 28.94 27.34 - 1.6 0.16 Glucose Management Indicator (GMI) (%) 7.97 6.94 - 1.03 <0.0001 Average blood glucose (mmol/L) 10.74 8.5 - 2.24 <0.0001 Time in Range (TIR) 3.9-10.0mmol/L (%) 46 73 + 27 <0.0001 Time above range (TAR) – High 10.1-13.9mmol/L (%) 31.32 21.26 - 10.06 <0.0001 Time above range (TAR) – Very High >14.0 mmol/L (%) 20.98 4.03 - 16.95 <0.0001 Time Below range (TBR) – Low 3.1-3.8mmo/L (%) 1.29 1.19 - 0.1 0.98 Time Below range (TBR) – Very Low <3.0 mmo/L (%) 0.6 0.16 - 0.44 0.07

In conclusion, the WSD model exemplifies the transformative potential of digital health technologies in redefining diabetes care, heralding substantial glycemic improvements and fostering a collaborative care ethos. Embracing and refining such innovative care approaches hold the promise of scaling diabetes management capabilities, potentially improving health outcomes for a broader spectrum of patients with Type 2 Diabetes. As we move forward, our focus will also extend to objectively measuring patient health outcomes in relation to engagement levels and social determinants, aiming to further enhance the care model's efficacy and reach.