The Diabetes teleMonitoring of patients in insulin Therapy (DiaMonT) trial: study protocol for a randomized controlled trial

Modeling the fasting blood glucose response to basal insulin adjustment in type 2 diabetes: An explainable machine learning approach on real-world data

INTRODUCTION

Optimal basal insulin titration for people with type 2 diabetes is vital to effectively reducing the risk of complications. However, a sizeable proportion of people (30-50 %) remain in suboptimal glycemic control six months post-initiation of basal insulin. This indicates a clear need for novel titration methods that account for individual patient variability in real-world settings.

OBJECTIVE

This study aims to investigate the use of real-world data and explainable machine learning in modeling fasting glucose responses to basal insulin adjustments, focusing on identifying factors influencing fasting glucose variability.

METHODS

A three-step explanatory approach was used to develop models using multiple linear regression, forward feature selection, and three-fold cross-validation. The models were built progressively, starting with a baseline model incorporating fasting blood glucose and insulin dose adjustments, followed by iterative models that in turn included biometric data, social factors, and biochemistry data, and lastly, a comprehensive model without constraints on the feature pool.

RESULTS

The baseline model yielded an average root mean squared error (RMSE) of 1.52 [95% CI: 1.33-1.71]. The iterative models resulted in an average RMSE of 1.49 [95% CI: 1.35-1.62] (biometric data), 1.47 [95% CI: 1.36-1.58] (social factors), and 1.52 [95% CI: 1.34-1.70] (biochemistry data). The comprehensive model yielded an average RMSE of 1.44 [95% CI: 1.41-1.48].

CONCLUSION

Developing explainable machine learning models using real-world data is possible for basal insulin titration. However, model performance is influenced by data's ability to capture everyday behavior, underscoring the need for incorporating more detailed behavioral and social data to optimize future titration models.

From Stability to Variability: Classification of Healthy Individuals, Prediabetes, and Type 2 Diabetes Using Glycemic Variability Indices from Continuous Glucose Monitoring Data

Objective: This study aims to investigate the continuum of glucose control from normoglycemia to dysglycemia (HbA1c ≥ 5.7%/39 mmol/mol) using metrics derived from continuous glucose monitoring (CGM). In addition, we aim to develop a machine learning-based classification model to classify dysglycemia based on observed patterns. Methods: Data from five distinct studies, each featuring at least two days of CGM, were pooled. Participants included individuals classified as healthy, with prediabetes, or with type 2 diabetes mellitus (T2DM). Various CGM indices were extracted and compared across groups. The data set was split 70/30 for training and testing two classification models (XGBoost/Logistic Regression) to differentiate between prediabetes or dysglycemia and the healthy group. Results: The analysis included 836 participants (healthy: n = 282; prediabetes: n = 133; T2DM: n = 432). Across all CGM indices, a progressive shift was observed from the healthy group to those with diabetes (P < 0.001). Statistically significant differences (P < 0.01) were noted in mean glucose, time below range, time above 140 mg/dl, mobility, multiscale complexity index, and glycemic risk index when transitioning from health to prediabetes. The XGBoost models achieved the highest receiver operating characteristic area under the curve values on the test data set ranging from 0.91 [confidence interval (CI): 0.87-0.95] (prediabetes identification) to 0.97 [CI: 0.95-0.98] (dysglycemia identification). Conclusion: Our findings demonstrate a gradual deterioration of glucose homeostasis and increased glycemic variability across the spectrum from normo- to dysglycemia, as evidenced by CGM metrics. The performance of CGM-based indices in classifying healthy individuals and those with prediabetes and diabetes is promising.

Identifying the Relationship Between CGM Time in Range and Basal Insulin Adherence in People With Type 2 Diabetes

BACKGROUND

The study aimed to determine the relationship between basal insulin adherence and glycemic control evaluated by time in range (TIR) in people with insulin-treated type 2 diabetes (T2D), using data from both continuous glucose monitors (CGM) and connected insulin pens. Furthermore, the study aimed to determine the best basal insulin adherence metric.

METHODS

CGM data and basal insulin data were collected from 106 insulin-treated people (aged ≥18 years) with T2D. Three different adherence metrics were employed (dose deviation, dose deviation ≤20%, and a traditional metric) and a three-step methodology was used to measure insulin adherence level. The coefficient of determination (R2), based on a univariate linear regression analysis, was used to determine the relationship between each adherence metric and TIR.

RESULTS

A statistically significant relationship was observed between TIR and adherence quantified as the dose deviation ≤20% metric (R2 = 0.67, P = .006). Neither the relationship between the dose deviation metric and TIR (R2 = 0.43, P = .08) nor the relationship between the traditional metric and TIR (R2 = 0.35, P =.23) was found to be statistically significant.

CONCLUSIONS

Our study indicates a relationship between basal insulin adherence and TIR in people with insulin-treated T2D. This seems to underscore the role of basal insulin adherence for optimal glycemic outcomes and utilizing TIR as a clinical marker. Furthermore, the results suggest that the magnitude of deviation from the recommended basal insulin dose impacts glycemic control, indicating dose deviation ≤20% as a more accurate metric for quantifying adherence.

The intelligent diabetes telemonitoring using decision support to treat patients on insulin therapy (DiaTRUST) trial: study protocol for a randomized controlled trial

BACKGROUND

Diabetes affects 10.5% of adults globally, with type 2 diabetes accounting for 90-95% of cases. Achieving optimal glycemic control is crucial yet challenging, particularly with insulin therapy, where 30-50% of patients fail to meet treatment targets. Telemedicine can improve diabetes management but generates vast amounts of data, burdening healthcare professionals. Integrating clinical decision support tools into telemonitoring systems may enhance care efficiency and glycemic control.

METHODS

The trial is a multicenter 3-month, three-arm, open-label, randomized controlled trial. The trial aims to enroll 51 participants with type 2 diabetes on insulin therapy. Participants will be divided with a 3:1:1 ratio into telemonitoring with decision support, telemonitoring without decision support, and usual care groups. The study employs connected insulin pens, continuous glucose monitors (CGMs), and activity trackers to enable telemonitoring. Outcomes measured include CGM time in range, HbA1c, hypo- and hyperglycemia incidents, total daily insulin dose, body weight, treatment satisfaction, and adherence.

DISCUSSION

Telemonitoring with decision support has the potential to revolutionize diabetes management by offering personalized treatment suggestions, thereby reducing the burden on healthcare professionals, and improving patient outcomes. This study will provide valuable insights into the effectiveness of such an approach in achieving glycemic control in people with type 2 diabetes on insulin therapy. By evaluating both clinical outcomes and patient and healthcare professionals' satisfaction, the study aims to contribute to the development of efficient, scalable telehealth solutions for diabetes care.

TRIAL REGISTRATION

ClinicalTrials.gov NCT06185296. Registered on December 14, 2023.

A Three-Step Data-Driven Methodology to Assess Adherence to Basal Insulin Therapy in Patients With Insulin-Treated Type 2 Diabetes

BACKGROUND

While health care providers (HCPs) are generally aware of the challenges concerning insulin adherence in adults with insulin-treated type 2 diabetes (T2D), data guiding identification of insulin nonadherence and understanding of injection patterns have been limited. Hence, the aim of this study was to examine detailed injection data and provide methods for assessing different aspects of basal insulin adherence.

METHOD

Basal insulin data recorded by a connected insulin pen and prescribed doses were collected from 103 insulin-treated patients (aged ≥18 years) with T2D from an ongoing clinical trial (NCT04981808). We categorized the data and analyzed distributions of correct doses, increased doses, reduced doses, and missed doses to quantify adherence. We developed a three-step model evaluating three aspects of adherence (overall adherence, adherence distribution, and dose deviation) offering HCPs a comprehensive assessment approach.

RESULTS

We used data from a connected insulin pen to exemplify the use of the three-step model to evaluate overall, adherence, adherence distribution, and dose deviation using patient cases.

CONCLUSION

The methodology provides HCPs with detailed access to previously limited clinical data on insulin administration, making it possible to identify specific nonadherence behavior which will guide patient-HCP discussions and potentially provide valuable insights for tailoring the most appropriate forms of support.