Proactive Identification of Patients with Diabetes at Risk of Uncontrolled Outcomes during a Diabetes Management Program: Conceptualization and Development Study Using Machine Learning

BACKGROUND

The growth in the capabilities of telehealth have made it possible to identify individuals with a higher risk of uncontrolled diabetes and provide them with targeted support and resources to help them manage their condition. Thus, predictive modeling has emerged as a valuable tool for the advancement of diabetes management.

OBJECTIVE

This study aimed to conceptualize and develop a novel machine learning (ML) approach to proactively identify participants enrolled in a remote diabetes monitoring program (RDMP) who were at risk of uncontrolled diabetes at 12 months in the program.

METHODS

Registry data from the Livongo for Diabetes RDMP were used to design separate dynamic predictive ML models to predict participant outcomes at each monthly checkpoint of the participants' program journey (month-n models) from the first day of onboarding (month-0 model) up to the 11th month (month-11 model). A participant's program journey began upon onboarding into the RDMP and monitoring their own blood glucose (BG) levels through the RDMP-provided BG meter. Each participant passed through 12 predicative models through their first year enrolled in the RDMP. Four categories of participant attributes (ie, survey data, BG data, medication fills, and health signals) were used for feature construction. The models were trained using the light gradient boosting machine and underwent hyperparameter tuning. The performance of the models was evaluated using standard metrics, including precision, recall, specificity, the area under the curve, the F1-score, and accuracy.

RESULTS

The ML models exhibited strong performance, accurately identifying observable at-risk participants, with recall ranging from 70% to 94% and precision from 40% to 88% across the 12-month program journey. Unobservable at-risk participants also showed promising performance, with recall ranging from 61% to 82% and precision from 42% to 61%. Overall, model performance improved as participants progressed through their program journey, demonstrating the importance of engagement data in predicting long-term clinical outcomes.

CONCLUSIONS

This study explored the Livongo for Diabetes RDMP participants' temporal and static attributes, identification of diabetes management patterns and characteristics, and their relationship to predict diabetes management outcomes. Proactive targeting ML models accurately identified participants at risk of uncontrolled diabetes with a high level of precision that was generalizable through future years within the RDMP. The ability to identify participants who are at risk at various time points throughout the program journey allows for personalized interventions to improve outcomes. This approach offers significant advancements in the feasibility of large-scale implementation in remote monitoring programs and can help prevent uncontrolled glycemic levels and diabetes-related complications. Future research should include the impact of significant changes that can affect a participant's diabetes management.

The Effect of a Digital Mental Health Program on Anxiety and Depression Symptoms: Retrospective Analysis of Clinical Severity

BACKGROUND

Evidence-based digital health programs have shown efficacy in being primary tools to improve emotional and mental health, as well as offering supplementary support to individuals undergoing psychotherapy for anxiety, depression, and other mental health disorders. However, information is lacking about the dose response to digital mental health interventions.

OBJECTIVE

The objective of the study was to examine the effect of time in program and program usage on symptom change among individuals enrolled in a real-world comprehensive digital mental health program (myStrength) who are experiencing severe anxiety or depression.

METHODS

Eligible participants (N=18,626) were adults aged 18 years and older who were enrolled in myStrength for at least four weeks as part of their employee wellness benefit program, who completed baseline, the 2-week, 2-month, and 6-month surveys querying symptoms of anxiety (Generalized Anxiety Disorder-7 [GAD-7]) and depression (Patient Health Questionnaire-9 [PHQ-9]). Linear growth curve models were used to analyze the effect of average weekly program usage on subsequent GAD-7 and PHQ-9 scores for participants with scores indicating severe anxiety (GAD-7≥15) or depression (PHQ-9≥15). All models were adjusted for baseline score and demographics.

RESULTS

Participants in the study (N=1519) were 77.4% female (1176/1519), had a mean age of 45 years (SD 14 years), and had an average enrollment time of 3 months. At baseline, participants reported an average of 9.39 (SD 6.04) on the GAD-7 and 11.0 (SD 6.6) on the PHQ-9. Those who reported 6-month results had an average of 8.18 (SD 6.15) on the GAD-7 and 9.18 (SD 6.79) on the PHQ-9. Participants with severe scores (n=506) experienced a significant improvement of 2.97 (SE 0.35) and 3.97 (SE 0.46) at each time point for anxiety and depression, respectively (t=-8.53 and t=-8.69, respectively; Ps<.001). Those with severe baseline scores also saw a reduction of 0.27 (SE 0.08) and 0.25 (SE 0.09) points in anxiety and depression, respectively, for each additional program activity per week (t=-3.47 and t=-2.66, respectively; Ps<.05).

CONCLUSIONS

For participants with severe baseline scores, the study found a clinically significant reduction of approximately 9 points for anxiety and 12 points for depression after 6 months of enrollment, suggesting that interventions targeting mental health must maintain active, ongoing engagement when symptoms are present and be available as a continuous resource to maximize clinical impact, specifically in those experiencing severe anxiety or depression. Moreover, a dosing effect was shown, indicating improvement in outcomes among participants who engaged with the program every other day for both anxiety and depression. This suggests that digital mental health programs that provide both interesting and evidence-based activities could be more successful in further improving mental health outcomes.

Direct current cardioversion in pregnancy: a multicentre study

OBJECTIVE

Direct current cardioversion (DCCV) in pregnancy is rarely required and typically only documented in single case reports or case series. A recent UK confidential enquiry reported on several maternal deaths where appropriate DCCV appeared to have been withheld.

DESIGN

Retrospective cohort study.

SETTING

Seventeen UK and Ireland specialist maternity centres.

SAMPLE

Twenty-seven pregnant women requiring DCCV in pregnancy.

MAIN OUTCOME MEASURES

Maternal and fetal outcomes following DCCV.

RESULTS

Twenty-seven women had a total of 29 DCCVs in pregnancy. Of these, 19 (70%) initial presentations were to Emergency Departments and eight (30%) to maternity settings. There were no maternal deaths. Seventeen of the women (63%) had a prior history of heart disease. Median gestation at DCCV was 28 weeks, median gestation at delivery was 35 weeks, with a live birth in all cases. The abnormal heart rhythms documented at the first cardioversion were atrial fibrillation in 12/27 (44%) cases, atrial flutter in 8/27 (30%), supraventricular tachycardia in 5/27 (19%) and atrial tachycardia in 2/27 (7%). Fetal monitoring was undertaken following DCCV on 14/29 (48%) occasions (10 of 19 (53%) at ≥26 weeks) and on 2/29 (7%) occasions, urgent delivery was required post DCCV.

CONCLUSIONS

Direct current cardioversion in pregnancy is rarely required but should be undertaken when clinically indicated according to standard algorithms to optimise maternal wellbeing. Once the woman is stable post DCCV, gestation-relevant fetal monitoring should be undertaken. Maternity units should develop multidisciplinary processes to ensure pregnant women receive the same standard of care as their non-pregnant counterparts.

Association of physical activity on blood glucose in individuals with type 2 diabetes

Regular physical activity (PA) has been shown to improve glycemic control in persons with type 2 diabetes. This study aimed to investigate the impact of PA on blood glucose after controlling for medication use, demographics, and week of activation using a real-world population of individuals with type 2 diabetes. A longitudinal, retrospective study was performed evaluating weekly PA of Livongo members (N = 9,509), which analyzed fasting blood glucose (FBG), step counts, and daily active minutes. Linear mixed-effect modeling technique was used to investigate within member and between member effects of input variables on average weekly FBG. Of members enrolled, 6,336 (32%) had self-reported body mass index, qualified week with diabetes medications, and FBG measures. Members' baseline average age was 49.4 (SD 10.1) years old, 43% female, and 45,496 member weeks with an average of 7.2 qualified weeks (PA observable in ≥4 days) per member. Average weekly FBG was 140.5 mg/dL (SD 39.8), and average daily step counts were 4,833 (SD 3,266). Moving from sedentary (<5,000 steps per day) to active (≥5,000 steps per day) resulted in mean weekly FBG reduction of 13 mg/dL (95% CI: -22.6 to -3.14). One additional day of ≥8,000 steps reduced mean weekly FBG by 0.47 mg/dL (95% CI: -0.77 to -0.16). Members who completed 30 min of moderate to vigorous PA above the population average reduced mean weekly FBG by 7.7 mg/dL (95% CI: -13.4 to -2.0). PA is associated with a mean weekly FBG reduction of 13 mg/dL when changing from a sedentary to active lifestyle while participating in a remote diabetes monitoring program.

Effects of a multiple chronic condition (CC) remote monitoring program on clinical outcomes among older adults

Older adults are faced with an increased risk of comorbid chronic disease such as diabetes. While multiple health behavior change interventions (MHCIs) are known to improve clinical outcomes more than targeted interventions, less is known whether such effects persist in older populations. The objective of the study was to examine the effects of multiple chronic condition (CC) remote monitoring program enrollment and mental health program enrollment on glucose and blood pressure reduction, adjusting for self-monitoring behaviors. In a sample of 594 older adults (age 55+, 14% 65+ years, 46.8% female) evaluated over a 12-month period, statistical models showed that older adults with uncontrolled diabetes (A1c >= 7.0%) had a 7.9 pt. reduction in blood glucose for each additional program enrolled and a 22.7 pt. reduction in blood glucose when enrolled in mental health compared to those not enrolled. Similarly, older adults with uncontrolled hypertension (BP >= 130/80) had a 4.8 pt. reduction in systolic blood pressure for each additional program enrolled and a 7.2 pt. reduction in systolic blood pressure when enrolled in mental health compared to those not enrolled. The findings indicate the potential for multiprogram digital health interventions that incorporate mental health to further improve clinical outcomes in older adults suffering from multiple chronic diseases, namely diabetes and hypertension.

The role of the Radiographer in osteoporosis and fracture prevention services - a narrative review

OBJECTIVE

To explore relevant literature and policy around the role of the radiographer working within osteoporosis services. Discussion will examine the value of radiographers in these services, as well as current limitations and future opportunities for advancing practice in these domains.

KEY FINDINGS

Osteoporosis and fracture prevention are a public health issue that must be addressed to improve patient outcomes following fractures. DXA radiographers currently fulfill an important role in the diagnosis of osteoporosis and collaborative working between radiology and osteoporosis services is to be encouraged. Radiographers are able to extend their role into advanced practice within osteoporosis services such as fracture liaison and rheumatology, they have expert knowledge and experience to bring to these roles and post graduate education can further increase radiographer's expertise in this field. The inability of diagnostic radiographers to become independent prescribers is a current limitation for radiographers working within osteoporosis services.

CONCLUSION

The role of the radiographer working within DXA and osteoporosis services is evolving and is an exciting area of advanced practice. Promoting this specialty within radiography may help to improve job satisfaction as well as recruitment and retention rates. As radiographers scope of practice in osteoporosis changes and evolves, it is hoped that current legislation may change to allow independent prescribing for diagnostic radiographers, which can in turn streamline patient pathways and reduce the burden on primary and secondary care.

Baseline brain function in the preadolescents of the ABCD Study

The Adolescent Brain Cognitive Development (ABCD) Study® is a 10-year longitudinal study of children recruited at ages 9 and 10. A battery of neuroimaging tasks are administered biennially to track neurodevelopment and identify individual differences in brain function. This study reports activation patterns from functional MRI (fMRI) tasks completed at baseline, which were designed to measure cognitive impulse control with a stop signal task (SST; N = 5,547), reward anticipation and receipt with a monetary incentive delay (MID) task (N = 6,657) and working memory and emotion reactivity with an emotional N-back (EN-back) task (N = 6,009). Further, we report the spatial reproducibility of activation patterns by assessing between-group vertex/voxelwise correlations of blood oxygen level-dependent (BOLD) activation. Analyses reveal robust brain activations that are consistent with the published literature, vary across fMRI tasks/contrasts and slightly correlate with individual behavioral performance on the tasks. These results establish the preadolescent brain function baseline, guide interpretation of cross-sectional analyses and will enable the investigation of longitudinal changes during adolescent development.

Kidney transplant surgical director training: Urologists represent a functional alternative to general surgeons

INTRODUCTION

Kidney transplant (KT) directors are general surgeons or urologists. All KT centers must meet established performance standards. However, it has not been established if general surgery and urology led programs have disparate outcomes.

METHODS

Transplant outcomes and donor-recipient characteristics by director training were investigated. Organ Procurement and Transplantation Network (OPTN) directory, program websites were analyzed for surgical director demographics. Scientific Registry of Transplant Recipients (SRTR) 1-year kidney survival and deceased donor (DD) wait-time rankings were evaluated. A retrospective analysis of 142 157 KT recipients from 2010 to 2019 was performed using the United Network for Organ Sharing (UNOS) database.

RESULTS

One hunderd and seventy three (90.6%) KT programs were led by general surgeons. There were no significant differences in gender, ethnicity, region, credentials, or fellowship completion. Recipients undergoing KT with urology led programs were older (P = .002) and had longer wait-times (P < .001). These centers used higher KDPI (.47 vs. .45, P < .001) and higher HLA mismatch (3.92 vs. 3.89, P = .02) kidneys. Urology led centers utilized living donors less frequently (32.1% vs. 35.8%, P < .001) and had longer CIT (15.44 vs. 12.21, P < .001). Both had similar SRTR ranking of 1-year survival and DD wait-time.

CONCLUSION

Most directors were general surgeon. Patient outcomes did not differ by transplant director training. Urologists represent a viable option for KT leadership and recruitment should be encouraged.

Impact of a diabetes remote monitoring program on medication adherence

BACKGROUND: Medication nonadherence in diabetes is well documented to be associated with inadequate glycemic control. Through remote blood glucose (BG) monitoring, unlimited test strip and lancet supplies, personal coaching, and online access to clinical information and educational resources, diabetes remote monitoring (DRM) programs may provide a solution. OBJECTIVE: To examine the relationship between patient participation in a DRM solution and adherence to oral antidiabetic drugs (OAD). METHODS: A retrospective, propensity score-matched cohort study was conducted using deidentified administrative claims data from a large pharmacy benefit manager. Commercially insured patients aged 18 years or older and having 2 or more 30-day adjusted OAD claims comprised the target sample. Patients enrolled in insurance plans that implemented DRM, who had at least 1 BG check (ever engaged) between April 1, 2015, and March 31, 2018 (exposure) were matched to patients enrolled in insurance plans that did not implement DRM (nonexposure). After a 1:2 matching on baseline demographics, disease burden proxy, total pharmacy out-of-pocket costs, previous adherence and insulin use, nonexposure group participants were assigned the same first BG check date as their matched DRM participants. Medication adherence measured as proportion of days covered (PDC) in the 365 days following first BG check was examined as a continuous and binary outcome measure (PDC > 80% or adherent vs < 80% or nonadherent). Multivariable linear and logistic regression were conducted to examine differential magnitude in adherence and likelihood of being adherent, respectively. RESULTS: The final sample consisted of 6,002 exposure and 12,004 nonexposure group patients. DRM participants who were ever engaged had a 4.5% higher adherence rate (P < 0.001) and 42% higher odds of being adherent (P < 0.001) in the period after engagement compared with non-DRM participants. Sensitivity analyses showed that patients engaged continuously (> 1 BG check per week) for 3, 6, and 12 months had 5.1%, 5.2%, and 6.4% higher adherence rates, respectively (P < 0.001), and 52%, 64%, and 98% higher odds of being adherent, respectively (P < 0.001), compared with non-DRM participants. CONCLUSIONS: The study findings offer evidence that DRM engagement is associated with higher odds of medication adherence. DRM solutions that provide access to glucose test results, personalized coaching, educational resources, and lower testing supply cost can also influence adherence. Our findings have important implications for payers and patients related to improved health outcomes due to higher medication adherence. DISCLOSURES: Funding for this study was provided by Express Scripts. Munshi, Amelung, Carter, and Henderson are employed by Express Scripts. James and Shah are employed by Livongo, which provided the DRM solution.

Engagement in a Multi-Feature Digital Health Lifestyle Change Program as a Predictor of Weight Loss

Background: Modern digital health interventions targeting weight loss employ multiple evidence-based strategies, including nutrition tracking, coaching, and activity monitoring, providing users with choice as they set and achieve their own goals. Still, limited research exists on the partial effects of each component of such interventions, and whether participants choose to use all of the features of a program. The objective of this study was to test the individual components of a fully-featured digital health lifestyle intervention as predictors of weight loss in a single statistical model.

Methods: Participants in the study (N=25,273) were enrolled in the Livongo for Weight Loss program as part of their employee wellness benefit across 57 states/territories of the US from April, 2019 to January, 2021. Participants received a cellular-connected scale to use daily in the program; they were asked to track their eating via the app and physical activity via smartphones or wearables. Additionally, participants could engage with coaches voluntarily or by receiving feedback from coaches on their recorded food logs. A mixed-effects generalized linear model was used to test the effects of scale usage, physical activity, human telephonic coaching, and food logs without and without coaching feedback on the percent weight loss the following month. Predictors were disaggregated into between- and within-subject components to understand the impact of each component relative to one’s own mean. The month in the program and whether or not that month occurred during the COVID-19 pandemic were entered as time-varying covariates. Baseline age, gender, and BMI were entered as time-invariant covariates.

Results: Participants were 45% male and had average age of 54.3 years old (SD =11.4), with an average BMI of 33.10 kg/m² (SD: 6.2). On average, participants were enrolled in the program for 10.4 months (SD: 5.1). Each additional use of the scale above one’s own average was associated with an overall 7.4% weight loss (z=21.06,p<0.001). Similarly, each additional minute of moderate-vigorous physical activity (MVPA) above one’s own average was associated with an overall 2.4% weight loss (z=3.14, p<0.01). Lastly, coaching and food logging with coaching feedback at a frequency above one’s own average were associated with approximately a 6% weight loss throughout the program (z=3.08 and 2.35, respectively; p<0.05 for both).

Conclusion: We found that frequency of use of a scale in a weight loss intervention was most predictive of weight loss, followed by human coaching interaction and physical activity. However, food logging without feedback did not significantly impact weight loss among participants. Additional work is needed to understand drivers associated with increased utilization of beneficial program features, including optimizing the use of coaching, which offers great benefit, but may be costly to scale.

The development and cognitive testing of the positive outcomes HIV PROM: a brief novel patient-reported outcome measure for adults living with HIV

BACKGROUND

People living with HIV experience burdensome multidimensional symptoms and concerns requiring person-centred care. Routine use of patient reported outcome measures can improve outcomes. There is no brief patient reported outcome measure (PROM) that currently reflects the breadth of concerns for people living with HIV. This study aimed to develop and cognitively test a brief novel patient reported outcome measure for use within routine adult HIV care- the "Positive Outcomes" HIV PROM.

METHODS

Development followed the COSMIN taxonomy and guidance for relevance and comprehensiveness, and Rothrock guidance on development of valid patient reported outcome measures. The Positive Outcomes HIV PROM was developed by a steering group (people living with HIV, HIV professionals and health services researchers) using findings from a previously reported qualitative study of priority outcomes for people living with HIV. The prototype measure was cognitively tested with a purposive sample of people living with HIV.

RESULTS

The Positive Outcomes HIV PROM consists of 23 questions (22 structured, and one open question) informed by the priorities of key stakeholders (n = 28 people living with HIV, n = 21 HIV professionals and n = 8 HIV commissioners) to ensure face and content validity, and refined through cognitive testing (n = 6 people living with HIV). Cognitive testing demonstrated high levels of acceptability and accessibility.

CONCLUSIONS

The Positive Outcomes HIV PROM is the first brief patient reported outcome measure reflecting the diverse needs of people living with HIV designed specifically for use in the clinical setting to support patient assessment and care, and drive service quality improvement. It is derived from primary data on the priority outcomes for people living with HIV and is comprehensive and acceptable. Further psychometric testing is required to ensure reliability and responsiveness.

Drivers of weight loss in a CDC-recognized digital diabetes prevention program

INTRODUCTION

To investigate the impact of the digital Livongo Diabetes Prevention Program (DPP) on weight at 12 months, understand participants' self-monitoring behaviors associated with greater weight loss, and evaluate the impact of coaching interactions on more frequent self-monitoring behaviors.

RESEARCH DESIGN AND METHODS

A retrospective analysis was performed using data from 2037 participants enrolled in the Livongo DPP who completed lesson 1 and recorded a starting weight during 2016-2017. Self-monitoring behaviors, including weigh-ins, food logging, activity, and coach-participant interactions, were analyzed at 6 and 12 months. Subgroup analysis was conducted based on those who were highly engaged versus those minimally engaged. Multiple regression analysis was performed using demographic, self-monitoring, and lesson attendance data to determine predictors of weight loss at 12 months and coaching impact on self-monitoring.

RESULTS

Participants had a mean age of 50 years (SD ±12), with a starting weight of 94 kg (SD ±21), were college-educated (78%), and were female (74%). Overall, participants lost on average 5.1% of their starting weight. Highly engaged participants lost 6.6% of starting body weight, with 25% losing ≥10% at 12 months. Logistic regression analysis showed each submitted food log was associated with 0.23 kg (p<0.05) weight loss, each lesson completed was associated with 0.14 kg (p<0.05) weight loss, and a week of 150 active minutes was associated with 0.1 kg (p<0.01) weight loss. One additional coach-participant message each week was associated with 1.4 more food logs per week, 1.6% increase in weeks with four or more weigh-ins, and a 2.7% increase in weeks with 150 min of activity.

CONCLUSIONS

Food logging had the largest impact on weight loss, followed by lesson engagement and physical activity. Future studies should examine further opportunities to deliver nutrition-based content to increase and sustain weight loss for DPP.

Consensus recommendations on calf- and herd-level passive immunity in dairy calves in the United States

Passive immunity in calves is evaluated or quantified by measuring serum or plasma IgG or serum total protein within the first 7 d of age. While these measurements inform about circulating concentrations of this important protein, they are also a proxy for evaluating all of the additional benefits of colostral ingestion. The current individual calf standard for categorizing dairy calves with successful passive transfer or failure of passive transfer of immunity are based on serum IgG concentrations of ≥10 and <10 g/L, respectively. This cutoff was based on higher mortality rates in calves with serum IgG <10 g/L. Mortality rates have decreased since 1991, but the percentage of calves with morbidity events has not changed over the same time period. Almost 90% of calves sampled in the USDA National Animal Health Monitoring System's Dairy 2014 study had successful passive immunity based on the dichotomous standard. Based on these observations, a group of calf experts were assembled to evaluate current data and determine if changes to the passive immunity standards were necessary to reduce morbidity and possibly mortality. In addition to the USDA National Animal Health Monitoring System's Dairy 2014 study, other peer-reviewed publications and personal experience were used to identify and evaluate potential standards. Four options were evaluated based on the observed statistical differences between categories. The proposed standard includes 4 serum IgG categories: excellent, good, fair, and poor with serum IgG levels of ≥25.0, 18.0-24.9, 10.0-17.9, and <10 g/L, respectively. At the herd level, we propose an achievable standard of >40, 30, 20, and <10% of calves in the excellent, good, fair, and poor categories, respectively. Because serum IgG concentrations are not practical for on-farm implementation, we provide corresponding serum total protein and %Brix values for use on farm. With one-third of heifer calves in 2014 already meeting the goal of ≥25 g/L serum IgG at 24 h of life, this achievable standard will require more refinement of colostrum management programs on many dairy farms. Implementation of the proposed standard should further reduce the risk of both mortality and morbidity in preweaned dairy calves, improving overall calf health and welfare.