Effectiveness of artificial intelligence vs. human coaching in diabetes prevention: a study protocol for a randomized controlled trial.

BACKGROUND: Prediabetes is a highly prevalent condition that heralds an increased risk of progression to type 2 diabetes, along with associated microvascular and macrovascular complications. The Diabetes Prevention Program (DPP) is an established effective intervention for diabetes prevention. However, participation in this 12-month lifestyle change program has historically been low. Digital DPPs have emerged as a scalable alternative, accessible asynchronously and recognized by the Centers for Disease Control and Prevention (CDC). Yet, most digital programs still incorporate human coaching, potentially limiting scalability. Furthermore, existing effectiveness results of digital DPPs are primarily derived from per protocol, longitudinal non-randomized studies, or comparisons to control groups that do not represent the standard of care DPP. The potential of an AI-powered DPP as an alternative to the DPP is yet to be investigated. We propose a randomized controlled trial (RCT) to directly compare these two approaches. METHODS: This open-label, multicenter, non-inferiority RCT will compare the effectiveness of a fully automated AI-powered digital DPP (ai-DPP) with a standard of care human coach-based DPP (h-DPP). A total of 368 participants with elevated body mass index (BMI) and prediabetes will be randomized equally to the ai-DPP (smartphone app and Bluetooth-enabled body weight scale) or h-DPP (referral to a CDC recognized DPP). The primary endpoint, assessed at 12 months, is the achievement of the CDC's benchmark for type 2 diabetes risk reduction, defined as any of the following: at least 5% weight loss, at least 4% weight loss and at least 150 min per week on average of physical activity, or at least a 0.2-point reduction in hemoglobin A1C. Physical activity will be objectively measured using serial actigraphy at baseline and at 1-month intervals throughout the trial. Secondary endpoints, evaluated at 6 and 12 months, will include changes in A1C, weight, physical activity measures, program engagement, and cost-effectiveness. Participants include adults aged 18-75 years with laboratory confirmed prediabetes, a BMI of ≥ 25 kg/m2 (≥ 23 kg/m2 for Asians), English proficiency, and smartphone users. This U.S. study is conducted at Johns Hopkins Medicine in Baltimore, MD, and Reading Hospital (Tower Health) in Reading, PA. DISCUSSION: Prediabetes is a significant public health issue, necessitating scalable interventions for the millions affected. Our pragmatic clinical trial is unique in directly comparing a fully automated AI-powered approach without direct human coach interaction. If proven effective, it could be a scalable, cost-effective strategy. This trial will offer vital insights into both AI and human coach-based behavioral change strategies in real-world clinical settings. TRIAL REGISTRATION: ClinicalTrials.gov NCT05056376. Registered on September 24, 2021, https://clinicaltrials.gov/study/NCT05056376.

A National Survey of Physicians' Views on the Importance and Implementation of Deintensifying Diabetes Medications.

BACKGROUND: Guidelines recommend deintensifying hypoglycemia-causing medications for older adults with diabetes whose hemoglobin A1c is below their individualized target, but this rarely occurs in practice. OBJECTIVE: To understand physicians' decision-making around deintensifying diabetes treatment. DESIGN: National physician survey. PARTICIPANTS: US physicians in general medicine, geriatrics, or endocrinology providing outpatient diabetes care. MAIN MEASURES: Physicians rated the importance of deintensifying diabetes medications for older adults with type 2 diabetes, and of switching medication classes, on 5-point Likert scales. They reported the frequency of these actions for their patients, and listed important barriers and facilitators. We evaluated the independent association between physicians' professional and practice characteristics and the importance of deintensifying and switching diabetes medications using multivariable ordered logistic regression models. KEY RESULTS: There were 445 eligible respondents (response rate 37.5%). The majority of physicians viewed deintensifying (80%) and switching (92%) diabetes medications as important or very important to the care of older adults. Despite this, one-third of physicians reported deintensifying diabetes medications rarely or never. While most physicians recognized multiple reasons to deintensify, two-thirds of physicians reported barriers of short-term hyperglycemia and patient reluctance to change medications or allow higher glucose levels. In multivariable models, geriatricians rated deintensification as more important compared to other specialties (p=0.027), and endocrinologists rated switching as more important compared to other specialties (p<0.006). Physicians with fewer years in practice rated higher importance of deintensification (p<0.001) and switching (p=0.003). CONCLUSIONS: While most US physicians viewed deintensifying and switching diabetes medications as important for the care of older adults, they deintensified infrequently. Physicians had ambivalence about the relative benefits and harms of deintensification and viewed it as a potential source of conflict with their patients. These factors likely contribute to clinical inertia, and studies focused on improving shared decision-making around deintensifying diabetes medications are needed.

Precision Prognostics for Cardiovascular Disease in Type 2 Diabetes: A Systematic Review and Meta-analysis.

Background Precision medicine has the potential to improve cardiovascular disease (CVD) risk prediction in individuals with type 2 diabetes (T2D). Methods We conducted a systematic review and meta-analysis of longitudinal studies to identify potentially novel prognostic factors that may improve CVD risk prediction in T2D. Out of 9380 studies identified, 416 studies met inclusion criteria. Outcomes were reported for 321 biomarker studies, 48 genetic marker studies, and 47 risk score/model studies. Results Out of all evaluated biomarkers, only 13 showed improvement in prediction performance. Results of pooled meta-analyses, non-pooled analyses, and assessments of improvement in prediction performance and risk of bias, yielded the highest predictive utility for N-terminal pro b-type natriuretic peptide (NT-proBNP) (high-evidence), troponin-T (TnT) (moderate-evidence), triglyceride-glucose (TyG) index (moderate-evidence), Genetic Risk Score for Coronary Heart Disease (GRS-CHD) (moderate-evidence); moderate predictive utility for coronary computed tomography angiography (low-evidence), single-photon emission computed tomography (low-evidence), pulse wave velocity (moderate-evidence); and low predictive utility for C-reactive protein (moderate-evidence), coronary artery calcium score (low-evidence), galectin-3 (low-evidence), troponin-I (low-evidence), carotid plaque (low-evidence), and growth differentiation factor-15 (low-evidence). Risk scores showed modest discrimination, with lower performance in populations different from the original development cohort. Conclusions Despite high interest in this topic, very few studies conducted rigorous analyses to demonstrate incremental predictive utility beyond established CVD risk factors for T2D. The most promising markers identified were NT-proBNP, TnT, TyG and GRS-CHD, with the highest strength of evidence for NT-proBNP. Further research is needed to determine their clinical utility in risk stratification and management of CVD in T2D.

A National Physician Survey of Deintensifying Diabetes Medications for Older Adults With Type 2 Diabetes.

OBJECTIVE: To determine physicians' approach to deintensifying (reducing/stopping) or switching hypoglycemia-causing medications for older adults with type 2 diabetes. RESEARCH DESIGN AND METHODS: In this national survey, U.S. physicians in general medicine, geriatrics, or endocrinology reported changes they would make to hypoglycemia-causing medications for older adults in three scenarios: good health, HbA1c of 6.3%; complex health, HbA1c of 7.3%; and poor health, HbA1c of 7.7%. RESULTS: There were 445 eligible respondents (response rate 37.5%). In patient scenarios, 48%, 4%, and 20% of physicians deintensified hypoglycemia-causing medications for patients with good, complex, and poor health, respectively. Overall, 17% of physicians switched medications without significant differences by patient health. One-half of physicians selected HbA1c targets below guideline recommendations for older adults with complex or poor health. CONCLUSIONS: Most U.S. physicians would not deintensify or switch hypoglycemia-causing medications within guideline-recommended HbA1c targets. Physician preference for lower HbA1c targets than guidelines needs to be addressed to optimize deintensification decisions.

Retrospective Quality Improvement Study of Insulin-Induced Hypoglycemia and Implementation of Hospital-Wide Initiatives.

BACKGROUND: Hospitalized patients who are receiving antihyperglycemic agents are at increased risk for hypoglycemia. Inpatient hypoglycemia may lead to increased risk for morbidity, mortality, prolonged hospitalization, and readmission within 30 days of discharge, which in turn may lead to increased costs. Hospital-wide initiatives targeting hypoglycemia are known to be beneficial; however, their impact on patient care and economic measures in community nonteaching hospitals are unknown. METHODS: This retrospective quality improvement study examined the effects of hospital-wide hypoglycemia initiatives on the rates of insulin-induced hypoglycemia in a community hospital setting from January 1, 2016, until September 30, 2019. The potential cost of care savings has been calculated. RESULTS: Among 49 315 total patient days, 2682 days had an instance of hypoglycemia (5.4%). Mean ± SD hypoglycemic patient days/month was 59.6 ± 16.0. The frequency of hypoglycemia significantly decreased from 7.5% in January 2016 to 3.9% in September 2019 (P = .001). Patients with type 2 diabetes demonstrated a significant decrease in the frequency of hypoglycemia (7.4%-3.8%; P < .0001), while among patients with type 1 diabetes the frequency trended downwards but did not reach statistical significance (18.5%-18.0%; P = 0.08). Based on the reduction of hypoglycemia rates, the hospital had an estimated cost of care savings of $98 635 during the study period. CONCLUSIONS: In a community hospital setting, implementation of hospital-wide initiatives targeting hypoglycemia resulted in a significant and sustainable decrease in the rate of insulin-induced hypoglycemia. These high-leverage risk reduction strategies may be translated into considerable cost savings and could be implemented at other community hospitals.

Development and Validation of a Machine Learning Model to Predict Near-Term Risk of Iatrogenic Hypoglycemia in Hospitalized Patients.

Importance: Accurate clinical decision support tools are needed to identify patients at risk for iatrogenic hypoglycemia, a potentially serious adverse event, throughout hospitalization. Objective: To predict the risk of iatrogenic hypoglycemia within 24 hours after each blood glucose (BG) measurement during hospitalization using a machine learning model. Design, Setting, and Participants: This retrospective cohort study, conducted at 5 hospitals within the Johns Hopkins Health System, included 54 978 admissions of 35 147 inpatients who had at least 4 BG measurements and received at least 1 U of insulin during hospitalization between December 1, 2014, and July 31, 2018. Data from the largest hospital were split into a 70% training set and 30% test set. A stochastic gradient boosting machine learning model was developed using the training set and validated on internal and external validation. Exposures: A total of 43 clinical predictors of iatrogenic hypoglycemia were extracted from the electronic medical record, including demographic characteristics, diagnoses, procedures, laboratory data, medications, orders, anthropomorphometric data, and vital signs. Main Outcomes and Measures: Iatrogenic hypoglycemia was defined as a BG measurement less than or equal to 70 mg/dL occurring within the pharmacologic duration of action of administered insulin, sulfonylurea, or meglitinide. Results: This cohort study included 54 978 admissions (35 147 inpatients; median [interquartile range] age, 66.0 [56.0-75.0] years; 27 781 [50.5%] male; 30 429 [55.3%] White) from 5 hospitals. Of 1 612 425 index BG measurements, 50 354 (3.1%) were followed by iatrogenic hypoglycemia in the subsequent 24 hours. On internal validation, the model achieved a C statistic of 0.90 (95% CI, 0.89-0.90), a positive predictive value of 0.09 (95% CI, 0.08-0.09), a positive likelihood ratio of 4.67 (95% CI, 4.59-4.74), a negative predictive value of 1.00 (95% CI, 1.00-1.00), and a negative likelihood ratio of 0.22 (95% CI, 0.21-0.23). On external validation, the model achieved C statistics ranging from 0.86 to 0.88, positive predictive values ranging from 0.12 to 0.13, negative predictive values of 0.99, positive likelihood ratios ranging from 3.09 to 3.89, and negative likelihood ratios ranging from 0.23 to 0.25. Basal insulin dose, coefficient of variation of BG, and previous hypoglycemic episodes were the strongest predictors. Conclusions and Relevance: These findings suggest that iatrogenic hypoglycemia can be predicted in a short-term prediction horizon after each BG measurement during hospitalization. Further studies are needed to translate this model into a real-time informatics alert and evaluate its effectiveness in reducing the incidence of inpatient iatrogenic hypoglycemia.

Hypoglycemia Communication in Primary Care Visits for Patients with Diabetes.

BACKGROUND: Hypoglycemia is a common and serious adverse effect of diabetes treatment, especially for patients using insulin or insulin secretagogues. Guidelines recommend that these patients be assessed for interval hypoglycemic events at each clinical encounter and be provided anticipatory guidance for hypoglycemia prevention. OBJECTIVE: To determine the frequency and content of hypoglycemia communication in primary care visits. DESIGN: Qualitative study PARTICIPANTS: We examined 83 primary care visits from one urban health practice representing 8 clinicians and 33 patients using insulin or insulin secretagogues. APPROACH: Using a directed content analysis approach, we analyzed audio-recorded primary care visits collected as part of the Achieving Blood Pressure Control Together study, a randomized trial of behavioral interventions for hypertension. The coding framework included communication about interval hypoglycemia, defined as discussion of hypoglycemic events or symptoms; the components of hypoglycemia anticipatory guidance in diabetes guidelines; and hypoglycemia unawareness. Hypoglycemia documentation in visit notes was compared to visit transcripts. KEY RESULTS: Communication about interval hypoglycemia occurred in 24% of visits, and hypoglycemic events were reported in 16%. Despite patients voicing fear of hypoglycemia, clinicians rarely assessed hypoglycemia frequency, severity, or its impact on quality of life. Hypoglycemia anticipatory guidance was provided in 21% of visits which focused on diet and behavior change; clinicians rarely counseled on hypoglycemia treatment or avoidance of driving. Limited discussions of hypoglycemia unawareness occurred in 8% of visits. Documentation in visit notes had low sensitivity but high specificity for ascertaining interval hypoglycemia communication or hypoglycemic events, compared to visit transcripts. CONCLUSIONS: In this high hypoglycemia risk population, communication about interval hypoglycemia and counseling for hypoglycemia prevention occurred in a minority of visits. There is a need to support clinicians to more regularly assess their patients' hypoglycemia burden and enhance counseling practices in order to optimize hypoglycemia prevention in primary care.

Provider Response to Critical Action Values for Hypoglycemia in the Ambulatory Setting: a Retrospective Cohort Study.

BACKGROUND: The blood glucose level triggering a critical action value (CAV) for hypoglycemia is not standardized, and associated outcomes are unknown. OBJECTIVE: To evaluate the clinical consequences of, and provider responses to, CAVs for hypoglycemia. DESIGN: Retrospective cohort study at Johns Hopkins Hospital and Johns Hopkins Bayview Medical Center between April 1, 2013, and January 31, 2017. PARTICIPANTS: Patients with an ambulatory serum glucose < 50 mg/dL. Point-of-care capillary glucose and whole blood glucose samples were excluded. MAIN MEASURES: Electronic medical record (EMR) review for providers' documented response to CAV, associated patient symptoms, and serious adverse events. KEY RESULTS: We analyzed 209 CAVs for hypoglycemia from 154 patients. The median age (IQR) was 59 years (46, 69), 89 (57.8%) were male, and 96 (62.3%) were black. Provider-to-patient contact occurred in 128 of 209 (61.2%) episodes, among which no documented etiology was observed for 81 of 128 (63.3%), no recommendations were provided in 32 of 128 (25.0%), and no patient-reported hypoglycemic symptoms were documented in 103 of 128 (80.5%). Serious adverse events were documented in 4 of 128 episodes (3.1%), two required glucagon administration, and three required an ED visit. Provider-to-patient contact was associated with the patient having malignant neoplasm (adjusted OR 3.63, p = 0.045) or a hypoglycemic disorder (adjusted OR 7.70, p = 0.018) and inversely associated with a longer time from specimen collection to EMR result (adjusted OR 0.90 per hour, p = 0.016). CONCLUSIONS: There is inconsistent provider-to-patient contact following CAVs for hypoglycemia, and the etiology and symptoms of hypoglycemia were infrequently documented. There were few serious documented adverse events associated with hypoglycemia, although undocumented events may have occurred, and the incidence of serious adverse events in non-contacted patients remains unknown. These findings demonstrate a need to standardize provider response to CAVs for hypoglycemia. Decreasing the lag time between sample collection and laboratory result reporting may increase provider-to-patient contact.

Glycemic control and diabetic foot ulcer outcomes: A systematic review and meta-analysis of observational studies.

OBJECTIVE: To evaluate the association between glycemic control (hemoglobin A1C, fasting glucose, and random glucose) and the outcomes of wound healing and lower extremity amputation (LEA) among patients with diabetic foot ulcers (DFUs). RESEARCH DESIGN AND METHODS: Medline, EMBASE, Cochrane Library, and Scopus were searched for observational studies published up to March 2019. Five independent reviewers assessed in duplicate the eligibility of each study based on predefined eligibility criteria and two independent reviewers assessed risk of bias. Ameta-analysis was performed to calculate a pooled odds ratio (OR) or hazard ratio (HR) using random effects for glycemic measures in relation to the outcomes of wound healing and LEA. Subgroup analyses were conducted to explore potential source of heterogeneity between studies. The study protocol is registered with PROSPERO (CRD42018096842). RESULTS: Of 4572 study records screened, 60 observational studies met the study eligibility criteria of which 47 studies had appropriate data for inclusion in one or more meta-analyses(n = 12,604 DFUs). For cohort studies comparing A1C >7.0 to 7.5% vs. lower A1C levels, the pooled OR for LEA was 2.04 (95% CI, 0.91, 4.57) and for studies comparing A1C ≥ 8% vs. <8%, the pooled OR for LEA was 4.80 (95% CI 2.83, 8.13). For cohort studies comparing fasting glucose ≥126 vs. <126 mg/dl, the pooled OR for LEA was 1.46 (95% CI, 1.02, 2.09). There was no association with A1C category and wound healing (OR or HR). There was high risk of bias with respect to comparability of cohorts as many studies did not adjust for potential confounders in the association between glycemic control and DFU outcomes. CONCLUSIONS: Our findings suggest that A1C levels ≥8% and fasting glucose levels ≥126 mg/dl are associated with increased likelihood of LEA in patients with DFUs. A purposively designed prospective study is needed to better understand the mechanisms underlying the association between hyperglycemia and LEA.

Patterns and predictors of antihyperglycemic intensification at hospital discharge for type 2 diabetic patients not on home insulin.

BACKGROUND: Diabetes mellitus is a prevalent condition among hospitalized patients and the inpatient setting presents an opportunity for providers to review and adjust antihyperglycemic medications. We sought to describe practice patterns and predictors of antihyperglycemic intensification (AHI) at hospital discharge for type 2 diabetes mellitus (T2DM) patients not on home insulin. METHODS: We conducted a retrospective study of adult patients with T2DM receiving either non-insulin antihyperglycemic (NIA) or no antihyperglycemic medications prior to admission who were hospitalized within two hospitals in the Johns Hopkins Health System from December 2015 to September 2016. Mean hospital glucose values and observed vs. individualized target hemoglobin A1C values (based on risk of mortality score) were used to define an indication for AHI. Multivariable logistic regression was used to identify predictors of AHI. RESULTS: A total of 554 discharges of 475 unique patients were included. An indication for AHI was present in 104 (18.8%) of discharges, and AHI occurred in 30 (28.8%) of these discharges. Higher mean admission BG values and A1C, fewer pre-admission antihyperglycemic agents, involvement of the diabetes service, and admitting service were associated with AHI, while no association was observed with age, sex, race, risk of mortality and severity of illness scores, or length of stay. AHI was not associated with 30-day readmission. CONCLUSION: An indication for AHI occurs relatively infrequently among hospitalized patients, but when present, AHI occurs in approximately 1 in 3 discharges. AHI appears to be related largely to the degree of hyperglycemia, and diabetes service involvement. Further studies are needed to understand the implications of AHI at hospital discharge on short and long-term outcomes in this population.

Hypoglycemic risk exposures in relation to low serum glucose values in ambulatory patients.

This study aimed to correlate hypoglycemic risk exposures (HREs) with low blood glucose value (BGV) in ambulatory patients to inform selection of a glucose critical action value (CAV).This was a retrospective study of ambulatory patients with at least 1 serum glucose ≤70 mg/dL obtained at 2 laboratories within the Johns Hopkins Health System over 3.8 years. Multivariable logistic regression was used to evaluate association of BGV cut-offs of <60, <54, <50, and <45 mg/dL with HREs. HREs were classified as "high hypoglycemic risk" (HHR), "moderate hypoglycemic risk" (MHR), "low hypoglycemic risk" (LHR), and "no hypoglycemic risk" (NHR).A total of 5404 patient samples of BG ≤70 mg/dL were analyzed, of which 30.3%, 23.2%, 28.5%, 18.0% occurred in NHR, LHR, MHR, and HHR groups, respectively. An inverse relationship was noted between BGV cut-offs and HHR, but no association was observed for LHR or MHR. After adjusting for age, sex, and race, there was an inverse association between BG thresholds and the odds of HHR. For classification of HHR, BGV cut-offs of <60, <54, <50, and <45 mg/dL correctly classified 71.2%, 69.8%, 68.8%, and 67.2% of BG samples, achieved false-positive rates of 13.6%, 4.7%, 1.7%, and 0.5% and positive likelihood ratios of 3.3, 6.0, 11.2, and 23.4, respectively.Nearly 70% of low BGVs occurred in patients with at least 1 HRE, but only ∼20% occurred in HHR patients. Given their high positive likelihood ratios, BGVs <54 or <50 mg/dL are reasonable candidates for CAVs that would allow sufficient clinician response time while minimizing false-positive alerts.

Immune-Modulating Therapy for Rheumatologic Disease: Implications for Patients with Diabetes.

Immune modulators used to treat rheumatologic disease have diverse endocrine effects in patients with diabetes. Providers should be aware of these effects given that diabetes and rheumatologic disease overlap in prevalence and cardiovascular morbidity. In patients with type 1 diabetes, clinical trials have demonstrated that immune modulators used early in the disease can improve pancreatic function, though their efficacy in adults with longstanding autoimmune diabetes is unknown. In patients with type 2 diabetes, hydroxychloroquine is an effective antihyperglycemic and may be preferred for rheumatologic use in patients with difficult glycemic control. In patients without diabetes, hydroxychloroquine and tumor necrosis factor (TNF) inhibitors have been found to decrease diabetes incidence in observational studies. Additionally, dapsone and sulfasalazine alter erythrocyte survival resulting in inaccurate HbA1c values. These multifaceted effects of immune modulators create a need for coordinated care between providers treating patients with diabetes to individualize medication selection and prevent hypoglycemic events. More research is needed to determine the long-term outcomes of immune modulators in patients with diabetes.

Racial differences in acute kidney injury of hospitalized adults with diabetes.

OBJECTIVE: To determine whether there is a racial difference in the risk of acute kidney injury between hospitalized black and white adults with diabetes mellitus in the United States RESEARCH DESIGN AND METHODS: We analyzed cross-sectional data from the 2000-2010 National Hospital Discharge Survey (NHDS) to compare the odds of AKI among hospitalized black and white adults with diabetes. After excluding records in which race status was missing, race was other than white or black, discharge status was not provided, or end-stage renal disease was a diagnosis, we identified 276,138 eligible records for analysis. Multivariable logistic regression was used to analyze the association between race, AKI, and in-hospital mortality. Multivariable linear regression was used to analyze the association between length of stay and race among discharge records with a diagnosis of AKI. RESULTS: In this nationally representative sample of hospitalized U.S. adults with diabetes, blacks had a 50% higher age- and sex-adjusted odds of AKI compared to whites (odds ratio: 1.51; 95% CI 1.37-1.66). The association between black race and increased risk of AKI persisted after additional adjustment for multiple AKI-related risk factors, including chronic kidney disease, sepsis, hypertension, hypotension, length of stay, myocardial infarction, congestive heart failure, angiography, computed tomography scan, cirrhosis, admission source, payor source, hospital region, and hospital bed size (OR 1.71; 95% CI, 1.31-2.25). Among cases of AKI, there was no racial difference in length of stay or in-hospital mortality. CONCLUSIONS: Among hospitalized adults in the U.S. with diabetes, black race is associated with a higher risk of AKI compared to white race.