Autonomous Artificial Intelligence Increases Access and Health Equity in Underserved Populations with Diabetes.

Diabetic eye disease (DED) is a leading cause of blindness in the world. Early detection and treatment of DED have been shown to be both sight-saving and cost-effective. As such, annual testing for DED is recommended for adults with diabetes and is a Healthcare Effectiveness Data and Information Set (HEDIS) measure. However, adherence to this guideline has historically been low, and access to this sight-saving intervention has particularly been limited for specific populations, such as Black or African American patients. In 2018, the US Food and Drug Agency (FDA) De Novo cleared autonomous artificial intelligence (AI) for diagnosing DED in a primary care setting. In 2020, Johns Hopkins Medicine (JHM), an integrated healthcare system with over 30 primary care sites, began deploying autonomous AI for DED testing in some of its primary care clinics. In this retrospective study, we aimed to determine whether autonomous AI implementation was associated with increased adherence to annual DED testing, and whether this was different for specific populations. JHM primary care sites were categorized as "non-AI" sites (sites with no autonomous AI deployment over the study period and where patients are referred to eyecare for DED testing) or "AI-switched" sites (sites that did not have autonomous AI testing in 2019 but did by 2021). We conducted a difference-in-difference analysis using a logistic regression model to compare change in adherence rates from 2019 to 2021 between non-AI and AI-switched sites. Our study included all adult patients with diabetes managed within our health system (17,674 patients for the 2019 cohort and 17,590 patients for the 2021 cohort) and has three major findings. First, after controlling for a wide range of potential confounders, our regression analysis demonstrated that the odds ratio of adherence at AI-switched sites was 36% higher than that of non-AI sites, suggesting that there was a higher increase in DED testing between 2019 and 2021 at AI-switched sites than at non-AI sites. Second, our data suggested autonomous AI improved access for historically disadvantaged populations. The adherence rate for Black/African Americans increased by 11.9% within AI-switched sites whereas it decreased by 1.2% within non-AI sites over the same time frame. Third, the data suggest that autonomous AI improved health equity by closing care gaps. For example, in 2019, a large adherence rate gap existed between Asian Americans and Black/African Americans (61.1% vs. 45.5%). This 15.6% gap shrank to 3.5% by 2021. In summary, our real-world deployment results in a large integrated healthcare system suggest that autonomous AI improves adherence to a HEDIS measure, patient access, and health equity for patients with diabetes - particularly in historically disadvantaged patient groups. While our findings are encouraging, they will need to be replicated and validated in a prospective manner across more diverse settings.

Use of Diabetes Technologies and Retinopathy in Adults With Type 1 Diabetes.

Importance: Diabetic retinopathy (DR) is a complication of diabetes that can lead to vision loss. Outcomes of continuous glucose monitoring (CGM) and insulin pump use in DR are not well understood. Objective: To assess the use of CGM, insulin pump, or both, and DR and proliferative diabetic retinopathy (PDR) in adults with type 1 diabetes (T1D). Design, Setting, and Participants: A retrospective cohort study of adults with T1D in a tertiary diabetes center and ophthalmology center was conducted from 2013 to 2021, with data analysis performed from June 2022 to April 2023. Exposure: Use of diabetes technologies, including insulin pump, CGM, and both CGM and insulin pump. Main Outcomes and Measures: The primary outcome was development of DR or PDR. A secondary outcome was the progression of DR for patients in the longitudinal cohort. Multivariable logistic regression models assessed for development of DR and PDR and association with CGM and insulin pump use. Results: A total of 550 adults with T1D were included (median age, 40 [IQR, 28-54] years; 54.4% female; 24.5% Black or African American; and 68.4% White), with a median duration of diabetes of 20 (IQR, 10-30) years, and median hemoglobin A1c (HbA1c) of 7.8% (IQR, 7.0%-8.9%). Overall, 62.7% patients used CGM, 58.2% used an insulin pump, and 47.5% used both; 44% (244 of 550) of the participants had DR at any point during the study. On univariate analysis, CGM use was associated with lower odds of DR and PDR, and CGM with pump was associated with lower odds of PDR (all P < .05), compared with no CGM use. Multivariable logistic regression adjusting for age, sex, race and ethnicity, diabetes duration, microvascular and macrovascular complications, insurance type, and mean HbA1c, showed that CGM was associated with lower odds of DR (odds ratio [OR], 0.52; 95% CI, 0.32-0.84; P = .008) and PDR (OR, 0.42; 95% CI, 0.23-0.75; P = .004), compared with no CGM use. In the longitudinal analysis of participants without baseline PDR, 79 of 363 patients (21.8%) had progression of DR during the study. Conclusions and Relevance: In this cohort study of adults with T1D, CGM use was associated with lower odds of developing DR and PDR, even after adjusting for HbA1c. These findings suggest that CGM may be useful for diabetes management to mitigate risk for DR and PDR.

Real-world use of glucagon-like peptide-1 receptor agonists in youth with type 2 diabetes is associated with short-term improvements in HbA1c.

AIM: To assess the short-term, real-world use and effectiveness of glucagon-like peptide-1 receptor agonist (GLP-1RA) medications in the management of type 2 diabetes (T2D) in a diverse cohort of youth. METHODS: This multicentre retrospective study analysed youth prescribed a GLP-1RA for the management of T2D at two academic paediatric diabetes centres prior to June 2022. Change in HbA1c and insulin use from baseline to first (median 91 days) and second (median 190 days) follow-up were evaluated for those taking a GLP-1RA. Multivariable linear mixed effects models adjusting for baseline sex, age, race/ethnicity, insurance, insulin regimen, metformin regimen, GLP-1RA dosing frequency and the body mass index Z-score (BMI-Z) examined the change in HbA1c for participants for up to 6 months after baseline. RESULTS: A total of 136 patients with T2D (median age 16.1 [interquartile range 13.9-18.0] years, 54% female, 56% non-Hispanic Black, 24% Hispanic, 77% with public insurance) were prescribed GLP-1RAs and taking them at first or second follow-up. Median HbA1c decreased from 7.9% to 7.6% (P < .001) at a median follow-up of 91 days (n = 109) and, among those with HbA1c available at baseline and second follow-up (n = 83), from 8.4% to 7.4%. The proportion of patients prescribed insulin decreased from baseline to the first follow-up visit (basal 69% to 60% [P = .008], prandial 46% to 38% [P = .03]). In multivariable analysis, there was a mean decrease in HbA1c by 0.09 percentage points per month (P = .005, 95% confidence interval -0.15, -0.03). CONCLUSIONS: Real-world use of GLP-1RAs in youth with T2D is associated with decreased HbA1c levels, despite challenges with access and adherence. GLP-1RA treatment may reduce insulin doses for youth with T2D.

Machine Learning Models for Prediction of Diabetic Microvascular Complications.

IMPORTANCE AND AIMS: Diabetic microvascular complications significantly impact morbidity and mortality. This review focuses on machine learning/artificial intelligence (ML/AI) in predicting diabetic retinopathy (DR), diabetic kidney disease (DKD), and diabetic neuropathy (DN). METHODS: A comprehensive PubMed search from 1990 to 2023 identified studies on ML/AI models for diabetic microvascular complications. The review analyzed study design, cohorts, predictors, ML techniques, prediction horizon, and performance metrics. RESULTS: Among the 74 identified studies, 256 featured internally validated ML models and 124 had externally validated models, with about half being retrospective. Since 2010, there has been a rise in the use of ML for predicting microvascular complications, mainly driven by DKD research across 27 countries. A more modest increase in ML research on DR and DN was observed, with publications from fewer countries. For all microvascular complications, predictive models achieved a mean (standard deviation) c-statistic of 0.79 (0.09) on internal validation and 0.72 (0.12) on external validation. Diabetic kidney disease models had the highest discrimination, with c-statistics of 0.81 (0.09) on internal validation and 0.74 (0.13) on external validation, respectively. Few studies externally validated prediction of DN. The prediction horizon, outcome definitions, number and type of predictors, and ML technique significantly influenced model performance. CONCLUSIONS AND RELEVANCE: There is growing global interest in using ML for predicting diabetic microvascular complications. Research on DKD is the most advanced in terms of publication volume and overall prediction performance. Both DR and DN require more research. External validation and adherence to recommended guidelines are crucial.

Stakeholder Engagement in Type 1 Diabetes Research, Quality Improvement, and Clinical Care.

The integration of stakeholder engagement (SE) in research, quality improvement (QI), and clinical care has gained significant traction. Type 1 diabetes is a chronic disease that requires complex daily management and care from a multidisciplinary team across the lifespan. Inclusion of key stakeholder voices, including patients, caregivers, health care providers and community advocates, in the research process and implementation of clinical care is critical to ensure representation of perspectives that match the values and goals of the patient population. This review describes the current framework for SE and its application to research, QI, and clinical care across the lifespan.

A New Approach to Staging Diabetic Eye Disease: Staging of Diabetic Retinal Neurodegeneration and Diabetic Macular Edema.

Topic: The goal of this review was to summarize the current level of evidence on biomarkers to quantify diabetic retinal neurodegeneration (DRN) and diabetic macular edema (DME). Clinical relevance: With advances in retinal diagnostics, we have more data on patients with diabetes than ever before. However, the staging system for diabetic retinal disease is still based only on color fundus photographs and we do not have clear guidelines on how to incorporate data from the relatively newer modalities into clinical practice. Methods: In this review, we use a Delphi process with experts to identify the most promising modalities to identify DRN and DME. These included microperimetry, full-field flash electroretinogram, spectral-domain OCT, adaptive optics, and OCT angiography. We then used a previously published method of determining the evidence level to complete detailed evidence grids for each modality. Results: Our results showed that among the modalities evaluated, the level of evidence to quantify DRN and DME was highest for OCT (level 1) and lowest for adaptive optics (level 4). Conclusion: For most of the modalities evaluated, prospective studies are needed to elucidate their role in the management and outcomes of diabetic retinal diseases. Financial Disclosures: Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Autonomous artificial intelligence increases screening and follow-up for diabetic retinopathy in youth: the ACCESS randomized control trial.

Diabetic retinopathy can be prevented with screening and early detection. We hypothesized that autonomous artificial intelligence (AI) diabetic eye exams at the point-of-care would increase diabetic eye exam completion rates in a racially and ethnically diverse youth population. AI for Children's diabetiC Eye ExamS (NCT05131451) is a parallel randomized controlled trial that randomized youth (ages 8-21 years) with type 1 and type 2 diabetes to intervention (autonomous artificial intelligence diabetic eye exam at the point of care), or control (scripted eye care provider referral and education) in an academic pediatric diabetes center. The primary outcome was diabetic eye exam completion rate within 6 months. The secondary outcome was the proportion of participants who completed follow-through with an eye care provider if deemed appropriate. Diabetic eye exam completion rate was significantly higher (100%, 95%CI: 95.5%, 100%) in the intervention group (n = 81) than the control group (n = 83) (22%, 95%CI: 14.2%, 32.4%)(p < 0.001). In the intervention arm, 25/81 participants had an abnormal result, of whom 64% (16/25) completed follow-through with an eye care provider, compared to 22% in the control arm (p < 0.001). Autonomous AI increases diabetic eye exam completion rates in youth with diabetes.

Risk Factors for Nondiagnostic Imaging in a Real-World Deployment of Artificial Intelligence Diabetic Retinal Examinations in an Integrated Healthcare System: Maximizing Workflow Efficiency Through Predictive Dilation.

OBJECTIVE: In the pivotal clinical trial that led to Food and Drug Administration De Novo "approval" of the first fully autonomous artificial intelligence (AI) diabetic retinal disease diagnostic system, a reflexive dilation protocol was used. Using real-world deployment data before implementation of reflexive dilation, we identified factors associated with nondiagnostic results. These factors allow a novel predictive dilation workflow, where patients most likely to benefit from pharmacologic dilation are dilated a priori to maximize efficiency and patient satisfaction. METHODS: Retrospective review of patients who were assessed with autonomous AI at Johns Hopkins Medicine (8/2020 to 5/2021). We constructed a multivariable logistic regression model for nondiagnostic results to compare characteristics of patients with and without diagnostic results, using adjusted odds ratio (aOR). P < .05 was considered statistically significant. RESULTS: Of 241 patients (59% female; median age = 59), 123 (51%) had nondiagnostic results. In multivariable analysis, type 1 diabetes (T1D, aOR = 5.82, 95% confidence interval [CI]: 1.45-23.40, P = .01), smoking (aOR = 2.86, 95% CI: 1.36-5.99, P = .005), and age (every 10-year increase, aOR = 2.12, 95% CI: 1.62-2.77, P < .001) were associated with nondiagnostic results. Following feature elimination, a predictive model was created using T1D, smoking, age, race, sex, and hypertension as inputs. The model showed an area under the receiver-operator characteristics curve of 0.76 in five-fold cross-validation. CONCLUSIONS: We used factors associated with nondiagnostic results to design a novel, predictive dilation workflow, where patients most likely to benefit from pharmacologic dilation are dilated a priori. This new workflow has the potential to be more efficient than reflexive dilation, thus maximizing the number of at-risk patients receiving their diabetic retinal examinations.

The Design of the Electronic Health Record in Type 1 Diabetes Centers: Implications for Metrics and Data Availability for a Quality Collaborative.

BACKGROUND: Systematic and comprehensive data acquisition from the electronic health record (EHR) is critical to the quality of data used to improve patient care. We described EHR tools, workflows, and data elements that contribute to core quality metrics in the Type 1 Diabetes Exchange Quality Improvement Collaborative (T1DX-QI). METHOD: We conducted interviews with quality improvement (QI) representatives at 13 T1DX-QI centers about their EHR tools, clinic workflows, and data elements. RESULTS: All centers had access to structured data tools, nine had access to patient questionnaires and two had integration with a device platform. There was significant variability in EHR tools, workflows, and data elements, thus the number of available metrics per center ranged from four to 17 at each site. Thirteen centers had information about glycemic outcomes and diabetes technology use. Seven centers had measurements of additional self-management behaviors. Centers captured patient-reported outcomes including social determinants of health (n = 9), depression (n = 11), transition to adult care (n = 7), and diabetes distress (n = 3). Various stakeholders captured data including health care professionals, educators, medical assistants, and QI coordinators. Centers that had a paired staffing model in clinic encounters distributed the burden of data capture across the health care team and was associated with a higher number of available data elements. CONCLUSIONS: The lack of standardization in EHR tools, workflows, and data elements captured resulted in variability in available metrics across centers. Further work is needed to support measurement and subsequent improvement in quality of care for individuals with type 1 diabetes.

Blocking AMPKαS496 phosphorylation improves mitochondrial dynamics and hyperglycemia in aging and obesity.

Impaired mitochondrial dynamics causes aging-related or metabolic diseases. Yet, the molecular mechanism responsible for the impairment of mitochondrial dynamics is still not well understood. Here, we report that elevated blood insulin and/or glucagon levels downregulate mitochondrial fission through directly phosphorylating AMPKα at S496 by AKT or PKA, resulting in the impairment of AMPK-MFF-DRP1 signaling and mitochondrial dynamics and activity. Since there are significantly increased AMPKα1 phosphorylation at S496 in the liver of elderly mice, obese mice, and obese patients, we, therefore, designed AMPK-specific targeting peptides (Pa496m and Pa496h) to block AMPKα1S496 phosphorylation and found that these targeting peptides can increase AMPK kinase activity, augment mitochondrial fission and oxidation, and reduce ROS, leading to the rejuvenation of mitochondria. Furthermore, these AMPK targeting peptides robustly suppress liver glucose production in obese mice. Our data suggest these targeting peptides are promising therapeutic agents for improving mitochondrial dynamics and activity and alleviating hyperglycemia in elderly and obese patients.

Development and implementation of a workshop for young adults with diabetes entering college and the workforce.

The process of transitioning from pediatric to adult diabetes care for adolescents and young adults is challenging. This transition period may include many life changes, and can be fraught with worsening glycemic control leading to increased risk for diabetes-related hospitalizations and complications. Research has demonstrated that increased support during this period can help maintain engagement in diabetes care. Transition guidelines highlight the importance of preparation and readiness for transition. In this article, we discuss the development, implementation and content of a workshop for patients and parents/caregivers preparing for the transition to college, the workforce and adult diabetes care.

Real-World Glycemic Outcomes with Early Omnipod 5 Use in Youth with Type 1 Diabetes.

Background: Pivotal trials of diabetes technologies have demonstrated glycemic improvements; however, these trials include patients of limited diversity and ranges of glycemic control. We assessed changes in glycemic control during the first 90 days of Omnipod 5 use in a real-world cohort of youth with type 1 diabetes (T1D). Methods: Youth 2-21 years with T1D initiating Omnipod 5 at two pediatric academic centers were included. Fourteen days of baseline (BL) continuous glucose monitoring (CGM) data were compared against data from the first 90 days of Omnipod 5 use. Outcome measures included changes in time in range (TIR), hemoglobin A1c (HbA1c), and CGM and insulin pump metrics based on the duration of Omnipod 5 use. Results: Among 195 youth (78.9% non-Hispanic White, 15.4% publicly insured, age 11.7 years, T1D duration 3.3 years) TIR increased 11%-points, from 49% to 61% (P < 0.001), and HbA1c decreased 0.5%-points, from 7.5% to 6.9% (P < 0.001). TIR improved within the first 9 days of Omnipod 5 use (p < 0.001) and did not change significantly thereafter (P = 0.1) despite decreases in user-initiated boluses (5.1 vs. 5.0, P = 0.01) and carbohydrate entries (4.2 vs. 4.1, P = 0.005) from days 1-9 to days 1-90. TIR improved 15%-points among youth with BL TIR <60% compared to a 5%-point increase for youth with BL TIR ≥60% (P < 0.001). Conclusions: Glycemic control improved within 9 days of Omnipod 5 initiation in this real-world cohort, and improvements were sustained over the first 90 days of use despite concomitant decreases in user-initiated boluses. These improvements were comparable to those observed in the pivotal trial.

Perioperative maintenance of hybrid closed loop insulin pump systems in youth with type 1 diabetes mellitus: A case series.

Current guidelines support the use of continuous glucose monitoring devices and insulin pumps in minor surgical procedures for pediatric patients with type 1 diabetes mellitus. However, there are few reported cases of using hybrid closed loop technology in the perioperative period. This retrospective case series presents seven pediatric patients with type 1 diabetes who underwent eight surgical procedures with maintenance of hybrid closed loop systems. This paper also provides considerations for future use of hybrid closed loop systems perioperatively.

Short-term use of CGM in youth onset type 2 diabetes is associated with behavioral modifications.

Background: Continuous glucose monitoring (CGM) is beneficial to glycemic control in youth with type 1 diabetes (T1D) and adults with type 2 diabetes (T2D); however, studies in youth with T2D are limited. Objective: Determine if 10-day trial CGM use in youth with T2D improves glycemic control and behavioral modifications. Methods: Youth with T2D > 3 months, on insulin, with no prior CGM use were enrolled. Staff placed CGM and provided education. Participants received 5-day and 10-day follow-up phone calls to review CGM data, behavioral modifications, and adjust insulin doses as needed. We compared 5-day to 10-day TIR, and baseline to 3-6 month HbA1c via paired t-test. Results: Participants (n=41) had median age of 16.2 y, were 61% female, 81% NH Black, median diabetes duration of 0.8 y, and baseline HbA1c of 10.3%. A majority had household income<$50,000 (81%) and parental education level of HS or less (73%). Average 5-day TIR 49% was similar to 10-day TIR 51% (p=0.62). There was no change in HbA1c after 3-6 months (10.2% v 10.3%, p=0.89). Nineteen participants completed full 10-day CGM use; of those, 84% wanted a CGM long-term. Adolescents reported behavioral changes including increased blood sugar checks, increased insulin administration and overall improved diabetes management. Conclusion: Although 10-day CGM use did not impact short-term or long-term glycemic control in youth with T2D, most participants reported behavioral changes and wanted to continue using CGM. Future studies with longer use of CGM may clarify the potential impact of CGM in youth with T2D.

Brief Pictorial Quizzes to Assess Carbohydrate Counting and Nutrition Knowledge in Youth With Type 1 Diabetes.

Available assessments of patient nutrition knowledge and carbohydrate counting ability are lengthy. This article reports on a study to implement and validate a series of brief nutrition quizzes of varying difficulty for use in pediatric type 1 diabetes. Among 129 youth with type 1 diabetes, participants completed an average of 2.4 ± 1 of the six quizzes, with a median score of 4.7 of 5. Higher quiz scores were associated with lower A1C (P <0.001), higher parental education (P = 0.02), and higher income (P = 0.01). Such quizzes can help to identify knowledge gaps and provide opportunities for education, which may improve glycemic outcomes in youth with type 1 diabetes.

A Brief Nutrition Questionnaire for Children With Newly Diagnosed Type 1 Diabetes.

Carbohydrate counting is an important component of type 1 diabetes management that is taught at the time of diagnosis. We implemented and validated a nutrition quiz to assess fundamental carbohydrate counting and nutrition knowledge in newly diagnosed patients. An interactive standard assessment for newly diagnosed type 1 diabetes patients was feasible and reliable to implement for patients and caregivers, but participants with public insurance scored lower overall. This assessment may help to identify nutrition knowledge gaps and provide opportunities for timely education, and providers should place additional focus on nutrition education for patients with public insurance.

Effectiveness of artificial intelligence screening in preventing vision loss from diabetes: a policy model.

The effectiveness of using artificial intelligence (AI) systems to perform diabetic retinal exams ('screening') on preventing vision loss is not known. We designed the Care Process for Preventing Vision Loss from Diabetes (CAREVL), as a Markov model to compare the effectiveness of point-of-care autonomous AI-based screening with in-office clinical exam by an eye care provider (ECP), on preventing vision loss among patients with diabetes. The estimated incidence of vision loss at 5 years was 1535 per 100,000 in the AI-screened group compared to 1625 per 100,000 in the ECP group, leading to a modelled risk difference of 90 per 100,000. The base-case CAREVL model estimated that an autonomous AI-based screening strategy would result in 27,000 fewer Americans with vision loss at 5 years compared with ECP. Vision loss at 5 years remained lower in the AI-screened group compared to the ECP group, in a wide range of parameters including optimistic estimates biased toward ECP. Real-world modifiable factors associated with processes of care could further increase its effectiveness. Of these factors, increased adherence with treatment was estimated to have the greatest impact.