JMIR Diabetes最新文献

Background: Diabetic ketoacidosis (DKA) is the leading cause of morbidity and mortality in pediatric type 1 diabetes (T1D), occurring in approximately 20% of patients, with an economic cost of $5.1 billion/year in the United States. Despite multiple risk factors for postdiagnosis DKA, there is still a need for explainable, clinic-ready models that accurately predict DKA hospitalization in established patients with pediatric T1D.

Objective: We aimed to develop an interpretable machine learning model to predict the risk of postdiagnosis DKA hospitalization in children with T1D using routinely collected time-series of electronic health record (EHR) data.

Methods: We conducted a retrospective case-control study using EHR data from 1787 patients from among 3794 patients with T1D treated at a large tertiary care US pediatric health system from January 2010 to June 2018. We trained a state-of-the-art; explainable, gradient-boosted ensemble (XGBoost) of decision trees with 44 regularly collected EHR features to predict postdiagnosis DKA. We measured the model's predictive performance using the area under the receiver operating characteristic curve-weighted F₁-score, weighted precision, and recall, in a 5-fold cross-validation setting. We analyzed Shapley values to interpret the learned model and gain insight into its predictions.

Results: Our model distinguished the cohort that develops DKA postdiagnosis from the one that does not (P<.001). It predicted postdiagnosis DKA risk with an area under the receiver operating characteristic curve of 0.80 (SD 0.04), a weighted F₁-score of 0.78 (SD 0.04), and a weighted precision and recall of 0.83 (SD 0.03) and 0.76 (SD 0.05) respectively, using a relatively short history of data from routine clinic follow-ups post diagnosis. On analyzing Shapley values of the model output, we identified key risk factors predicting postdiagnosis DKA both at the cohort and individual levels. We observed sharp changes in postdiagnosis DKA risk with respect to 2 key features (diabetes age and glycated hemoglobin at 12 months), yielding time intervals and glycated hemoglobin cutoffs for potential intervention. By clustering model-generated Shapley values, we automatically stratified the cohort into 3 groups with 5%, 20%, and 48% risk of postdiagnosis DKA.

Conclusions: We have built an explainable, predictive, machine learning model with potential for integration into clinical workflow. The model risk-stratifies patients with pediatric T1D and identifies patients with the highest postdiagnosis DKA risk using limited follow-up data starting from the time of diagnosis. The model identifies key time points and risk factors to direct clinical interventions at both the individual and cohort levels. Further research with data from multiple hospital systems can help us assess how well our model generalizes to o

Background: Blood glucose management around exercise is challenging for youth with type 1 diabetes (T1D). Previous research has indicated interventions including decision-support aids to better support youth to effectively contextualize blood glucose results and take appropriate action to optimize glucose levels during and after exercise. Mobile health (mHealth) apps help deliver health behavior interventions to youth with T1D, given the use of technology for glucose monitoring, insulin dosing, and carbohydrate counting.

Objective: We aimed to develop a novel prototype mHealth app to support exercise management among youth with T1D, detail the application of a co-design process and design thinking principles to inform app design and development, and identify app content and functionality that youth with T1D need to meet their physical activity goals.

Methods: A co-design approach with a user-centered design thinking framework was used to develop a prototype mHealth app "acT1ve" during the 18-month design process (March 2018 to September 2019). To better understand and respond to the challenges among youth with diabetes when physically active, 10 focus groups were conducted with youth aged 13-25 years with T1D and parents of youth with T1D. Thereafter, we conducted participatory design workshops with youth to identify key app features that would support individual needs when physically active. These features were incorporated into a wireframe, which was critically reviewed by participants. A beta version of "acT1ve" was built in iOS and android operating systems, which underwent critical review by end users, clinicians, researchers, experts in exercise and T1D, and app designers.

Results: Sixty youth with T1D, 14 parents, 6 researchers, and 10 clinicians were engaged in the development of "acT1ve." acT1ve included key features identified by youth, which would support their individual needs when physically active. It provided advice on carbohydrates and insulin during exercise, information on hypoglycemia treatment, pre- and postexercise advice, and an educational food guide regarding exercise management. "acT1ve" contained an exercise advisor algorithm comprising 240 pathways developed by experts in diabetes and exercise research. Based on participant input during exercise, acT1ve provided personalized insulin and carbohydrate advice for exercise lasting up to 60 minutes. It also contains other features including an activity log, which displays a complete record of the end users' activities and associated exercise advice provided by the app's algorithm for later reference, and regular reminder notifications for end users to check or monitor their glucose levels.

Conclusions: The co-design approach and the practical application of the user-centered design thinking framework were successfully applied in developing "acT1ve." The design thin

Background: Insurance benefit design influences whether individuals with diabetes who require a continuous glucose monitor (CGM) to provide real-time feedback on their blood glucose levels can obtain the CGM device from either a pharmacy or a durable medical equipment supplier. The impact of the acquisition channel on device adherence and health care costs has not been systematically evaluated.

Objective: This study aims to compare the adherence rates for patients new to CGM therapy and the costs of care for individuals who obtained CGM devices from a pharmacy versus acquisition through a durable medical equipment supplier using retrospective claims analysis.

Methods: Using the Mariner commercial claims database, individuals aged >18 years with documented diabetes and an initial CGM claim during the first quarter of 2021 (2021 Q1, index date) were identified. Patients had to maintain uninterrupted enrollment for a duration of 15 months but file no CGM claim during the 6 months preceding the index date. We used direct matching to establish comparable pharmacy and durable medical equipment cohorts. Outcomes included quarterly adherence, reinitiation, and costs for the period from 2021 Q1 to the third quarter of 2022 (2022 Q3). Between-cohort differences in adherence rates and reinitiation rates were analyzed using z tests, and cost differences were analyzed using 2-tailed t tests.

Results: Direct matching was used to establish comparable pharmacy and durable medical equipment cohorts. A total of 2356 patients were identified, with 1178 in the pharmacy cohort and 1178 in the durable medical equipment cohorts. Although adherence declined over time in both cohorts, the durable medical equipment cohort exhibited significantly superior adherence compared to the pharmacy cohort at 6 months (pharmacy n=615, 52% and durable medical equipment n=761, 65%; P<.001), 9 months (pharmacy n=579, 49% and durable medical equipment cohorts n=714, 61%; P<.001), and 12 months (pharmacy 48% and durable medical equipment n=714, 59%; P<.001). Mean annual total medical costs for adherent patients in the pharmacy cohort were 53% higher than the durable medical equipment cohort (pharmacy US $10,635 and durable medical equipment US $6967; P<.001). In nonadherent patients, the durable medical equipment cohort exhibited a significantly higher rate of therapy reinitiation during the period compared to the pharmacy cohort (pharmacy 61/613, 10% and durable medical equipment 108/485, 22%; P<.001).

Conclusions: The results from this real-world claims analysis demonstrate that, in a matched set, individuals who received their CGM through a durable medical equipment supplier were more adherent to their device. For individuals who experienced a lapse in therapy, those whose supplies were provided through the durable medical equipment channel were more likely to resume use aft

Background: Individuals with chronic diseases often search for health information online. The Diabetes Online Community (DOC) is an active community with members who exchange health information; however, few studies have examined health information brokering in the DOC.

Objective: The aim of this study was to develop and validate the Attitudes Toward Seeking Health Information Online (ATSHIO) scale in a sample of adults with type 1 diabetes (T1D).

Methods: People with T1D were recruited through the DOC, specifically Facebook and Twitter. They were provided with a Qualtrics link to complete the survey. This was a mixed methods study that used thematic analysis along with existing theory and formative research to design the quantitative ATSHIO scale.

Results: A total of 166 people with T1D participated in this study. Confirmatory factor analyses determined a 2-factor scale (Trusting and Evaluating Online Health Information in the DOC and Engaging With Online Health Information in the DOC) with good convergent validity and discriminant validity. Correlations were found between social support, online health information-seeking, diabetes distress, and disease management.

Conclusions: The ATSHIO scale can be used to investigate how people with diabetes are using the internet for obtaining health information, which is especially relevant in the age of telehealth and Health 2.0.

Background: The COVID-19 pandemic created unprecedented challenges for people with type 2 diabetes (T2D) and prediabetes to access in-person health care support. Primary care teams accelerated plans to implement digital health technologies (DHTs), such as remote consultations and digital self-management. There is limited evidence about whether there were inequalities in how people with T2D and prediabetes adjusted to these changes.

Objective: This study aimed to explore how people with T2D and prediabetes adapted to the reduction in in-person health support and the increased provision of support through DHTs during the COVID-19 pandemic and beyond.

Methods: A purposive sample of people with T2D and prediabetes was recruited by text message from primary care practices that served low-income areas. Semistructured interviews were conducted by phone or video call, and data were analyzed thematically using a hybrid inductive and deductive approach.

Results: A diverse sample of 30 participants was interviewed. There was a feeling that primary care had become harder to access. Participants responded to the challenge of accessing support by rationing or delaying seeking support or by proactively requesting appointments. Barriers to accessing health care support were associated with issues with using the total triage system, a passive interaction style with health care services, or being diagnosed with prediabetes at the beginning of the pandemic. Some participants were able to adapt to the increased delivery of support through DHTs. Others had lower capacity to use DHTs, which was caused by lower digital skills, fewer financial resources, and a lack of support to use the tools.

Conclusions: Inequalities in motivation, opportunity, and capacity to engage in health services and DHTs lead to unequal possibilities for people with T2D and prediabetes to self-care and receive care during the COVID-19 pandemic. These issues can be addressed by proactive arrangement of regular checkups by primary care services and improving capacity for people with lower digital skills to engage with DHTs.

Background: Black adolescents with type 1 diabetes (T1D) are at increased risk for suboptimal diabetes health outcomes; however, evidence-based interventions for this population are lacking. Depression affects a high percentage of youth with T1D and increases the likelihood of health problems associated with diabetes.

Objective: Our aim was to test whether baseline levels of depression moderate the effects of a brief eHealth parenting intervention delivered to caregivers of young Black adolescents with T1D on youths' glycemic control.

Methods: We conducted a multicenter randomized controlled trial at 7 pediatric diabetes clinics located in 2 large US cities. Participants (N=149) were allocated to either the intervention group or a standard medical care control group. Up to 3 intervention sessions were delivered on a tablet computer during diabetes clinic visits over a 12-month period.

Results: In a linear mixed effects regression model, planned contrasts did not show significant reductions in hemoglobin A_1c (HbA_1c) for intervention adolescents compared to controls. However, adolescents with higher baseline levels of depressive symptoms who received the intervention had significantly greater improvements in HbA_1c levels at 6-month follow-up (0.94%; P=.01) and 18-month follow-up (1.42%; P=.002) than those with lower levels of depression. Within the intervention group, adolescents had a statistically significant reduction in HbA_1c levels from baseline at 6-month and 18-month follow-up.

Conclusions: A brief, culturally tailored eHealth parenting intervention improved health outcomes among Black adolescents with T1D and depressive symptoms.

Trial registration: ClinicalTrials.gov NCT03168867; https://clinicaltrials.gov/study/NCT03168867.

Background: Digital health programs provide individualized support to patients with chronic diseases and their effectiveness is measured by the extent to which patients achieve target individual clinical outcomes and the program's ability to sustain patient engagement. However, patient dropout and inequitable intervention delivery strategies, which may unintentionally penalize certain patient subgroups, represent challenges to maximizing effectiveness. Therefore, methodologies that optimize the balance between success factors (achievement of target clinical outcomes and sustained engagement) equitably would be desirable, particularly when there are resource constraints.

Objective: Our objectives were to propose a model for digital health program resource management that accounts jointly for the interaction between individual clinical outcomes and patient engagement, ensures equitable allocation as well as allows for capacity planning, and conducts extensive simulations using publicly available data on type 2 diabetes, a chronic disease.

Methods: We propose a restless multiarmed bandit (RMAB) model to plan interventions that jointly optimize long-term engagement and individual clinical outcomes (in this case measured as the achievement of target healthy glucose levels). To mitigate the tendency of RMAB to achieve good aggregate performance by exacerbating disparities between groups, we propose new equitable objectives for RMAB and apply bilevel optimization algorithms to solve them. We formulated a model for the joint evolution of patient engagement and individual clinical outcome trajectory to capture the key dynamics of interest in digital chronic disease management programs.

Results: In simulation exercises, our optimized intervention policies lead to up to 10% more patients reaching healthy glucose levels after 12 months, with a 10% reduction in dropout compared to standard-of-care baselines. Further, our new equitable policies reduce the mean absolute difference of engagement and health outcomes across 6 demographic groups by up to 85% compared to the state-of-the-art.

Conclusions: Planning digital health interventions with individual clinical outcome objectives and long-term engagement dynamics as considerations can be both feasible and effective. We propose using an RMAB sequential decision-making framework, which may offer additional capabilities in capacity planning as well. The integration of an equitable RMAB algorithm further enhances the potential for reaching equitable solutions. This approach provides program designers with the flexibility to switch between different priorities and balance trade-offs across various objectives according to their preferences.

Background: Diabetes and hypertension are some of the most prevalent and costly chronic conditions in the United States. However, outcomes continue to lag behind targets, creating further risk of long-term complications, morbidity, and mortality for people living with these conditions. Furthermore, racial and ethnic disparities in glycemic and hypertension control persist. Flexible telehealth programs leveraging asynchronous care allow for increased provider access and more convenient follow-up, ultimately improving critical health outcomes across demographic groups.

Objective: We aim to evaluate the 12-month clinical outcomes of participants in the 9amHealth web-based clinic for diabetes and hypertension. We hypothesized that participation in the 9amHealth program would be associated with significant improvements in glycemic and blood pressure (BP) control across a diverse group of individuals.

Methods: We enrolled 95 patients in a completely web-based care clinic for diabetes and hypertension who received nutrition counseling, health coaching, and asynchronous physician consultations for medication prescribing. Patients received standard or cellular-connected glucose meters and BP cuffs in order to share data. Laboratory tests were completed either with at-home phlebotomy draws or a self-administered test kit. Patients' first and last hemoglobin A_1c (HbA_1c) and BP results over the 12-month period were compared, and analyses were repeated across race and ethnicity groups.

Results: Among all 95 patients, the average HbA_1c decreased by -1.0 (from 8.2% to 7.2%; P<.001) over 12 months of program participation. In those with a baseline HbA_1c >8%, the average HbA_1c decreased by -2.1 (from 10.2% to 8.1%; P<.001), and in those with a baseline HbA_1c >9%, the average HbA_1c decreased by -2.8 (from 11% to 8.2%; P<.001). Among participants who identified as a race or ethnicity other than White, the HbA_1c decreased by -1.2 (from 8.6% to 7.4%, P=.001). Further examination of subgroups confirmed HbA_1c lowering within each race or ethnicity group. In the overall population, the average systolic BP decreased by 17.7 mm Hg (P=.006) and the average diastolic BP decreased by 14.3 mm Hg (P=.002). Among participants self-identifying as a race or ethnicity other than White, the results similarly showed a decrease in BP (average reduction in systolic BP of 10 mm Hg and in diastolic BP of 9 mm Hg).

Conclusions: A fully web-based model leveraging all-asynchronous physician review and prescribing, combined with synchronous and asynchronous coaching and nutrition support, was associated with clinically meaningful improvement in HbA_1c and BP control over a 12-month period among a diverse group of individuals. Further studies should prospectively evaluate th

Background: Patients with type 2 diabetes mellitus (T2DM) experience multiple barriers to improving self-management. Evidence suggests that motivational interviewing (MI), a patient-centered communication method, can address patient barriers and promote healthy behavior. Despite the value of MI, existing MI studies predominantly used face-to-face or phone-based interventions. With the growing adoption of smartphones, automated MI techniques powered by artificial intelligence on mobile devices may offer effective motivational support to patients with T2DM.

Objective: This study aimed to explore the perspectives of patients with T2DM on the acceptability of app-based MI in routine health care and collect their feedback on specific MI module features to inform our future intervention.

Methods: We conducted semistructured interviews with patients with T2DM, recruited from public primary care clinics. All interviews were audio recorded and transcribed verbatim. Thematic analysis was conducted using NVivo.

Results: In total, 33 patients with T2DM participated in the study. Participants saw MI as a mental reminder to increase motivation and a complementary care model conducive to self-reflection and behavior change. Yet, there was a sense of reluctance, mainly stemming from potential compromise of autonomy in self-care by the introduction of MI. Some participants felt confident in their ability to manage conditions independently, while others reported already making changes and preferred self-management at their own pace. Compared with in-person MI, app-based MI was viewed as offering a more relaxed atmosphere for open sharing without being judged by health care providers. However, participants questioned the lack of human touch, which could potentially undermine a patient-provider therapeutic relationship. To sustain motivation, participants suggested more features of an ongoing supportive nature such as the visualization of milestones, gamified challenges and incremental rewards according to achievements, tailored multimedia resources based on goals, and conversational tools that are interactive and empathic.

Conclusions: Our findings suggest the need for a hybrid model of intervention involving both app-based automated MI and human coaching. Patient feedback on specific app features will be incorporated into the module development and tested in a randomized controlled trial.