Identification of undiagnosed atrial fibrillation using a machine learning risk prediction algorithm and diagnostic testing (PULsE-AI) in primary care: cost-effectiveness of a screening strategy evaluated in a randomized controlled trial in England.

OBJECTIVE: The PULsE-AI trial sought to determine the effectiveness of a screening strategy that included a machine learning risk prediction algorithm in conjunction with diagnostic testing for identification of undiagnosed atrial fibrillation (AF) in primary care. This study aimed to evaluate the cost-effectiveness of implementing the screening strategy in a real-world setting. METHODS: Data from the PULsE-AI trial - a prospective, randomized, controlled trial conducted across six general practices in England from June 2019 to February 2021 - were used to inform a cost-effectiveness analysis that included a hybrid screening decision tree and Markov AF disease progression model. Model outcomes were reported at both individual- and population-level (estimated UK population ≥30 years of age at high-risk of undiagnosed AF) and included number of patients screened, number of AF cases identified, mean total and incremental costs (screening, events, treatment), quality-adjusted-life-years (QALYs), and incremental cost-effectiveness ratio (ICER). RESULTS: The screening strategy was estimated to result in 45,493 new diagnoses of AF across the high-risk population in the UK (3.3 million), and an estimated additional 14,004 lifetime diagnoses compared with routine care only. Per-patient costs for high-risk individuals who underwent the screening strategy were estimated at £1,985 (vs £1,888 for individuals receiving routine care only). At a population-level, the screening strategy was associated with a cost increase of approximately £322 million and an increase of 81,000 QALYs. The screening strategy demonstrated cost-effectiveness versus routine care only at an accepted ICER threshold of £20,000 per QALY-gained, with an ICER of £3,994/QALY. CONCLUSIONS: Compared with routine care only, it is cost-effective to target individuals at high risk of undiagnosed AF, through an AF risk prediction algorithm, who should then undergo diagnostic testing. This AF risk prediction algorithm can reduce the number of patients needed to be screened to identify undiagnosed AF, thus alleviating primary care burden.

Budget impact analysis of a machine learning algorithm to predict high risk of atrial fibrillation among primary care patients.

AIMS: We investigated whether the use of an atrial fibrillation (AF) risk prediction algorithm could improve AF detection compared with opportunistic screening in primary care and assessed the associated budget impact. METHODS AND RESULTS: Eligible patients were registered with a general practice in UK, aged 65 years or older in 2018/19, and had complete data for weight, height, body mass index, and systolic and diastolic blood pressure recorded within 1 year. Three screening scenarios were assessed: (i) opportunistic screening and diagnosis (standard care); (ii) standard care replaced by the use of the algorithm; and (iii) combined use of standard care and the algorithm. The analysis considered a 3-year time horizon, and the budget impact for the National Health Service (NHS) costs alone or with personal social services (PSS) costs. Scenario 1 would identify 79 410 new AF cases (detection gap reduced by 22%). Scenario 2 would identify 70 916 (gap reduced by 19%) and Scenario 3 would identify 99 267 new cases (gap reduction 27%). These rates translate into 2639 strokes being prevented in Scenario 1, 2357 in Scenario 2, and 3299 in Scenario 3. The 3-year NHS budget impact of Scenario 1 would be £45.3 million, £3.6 million (difference ‒92.0%) with Scenario 2, and £46.3 million (difference 2.2%) in Scenario 3, but for NHS plus PSS would be ‒£48.8 million, ‒£80.4 million (64.8%), and ‒£71.3 million (46.1%), respectively. CONCLUSION: Implementation of an AF risk prediction algorithm alongside standard opportunistic screening could close the AF detection gap and prevent strokes while substantially reducing NHS and PSS combined care costs.

Updated Software for Automated Assessment of Glucose Variability and Quality of Glycemic Control in Diabetes.

Background: Glycemic variability is an important factor to consider in diabetes management. It can be assessed with multiple glycemic variability metrics and quality of control indices based on continuous glucose monitoring (CGM) recordings. For this, a robust repeatable calculation is important. A widely used tool for automated assessment is the EasyGV software. The aim of this work is to implement new methods of glycemic variability assessment in EasyGV and to validate implementation of each glucose metric in EasyGV against a reference implementation of the calculations. Methods: Validation data used came from the JDRF CGM study. Validation of the implementation of metrics that are available in EasyGV software v9 was carried out and the following new methods were added and validated: personal glycemic state, index of glycemic control, times in ranges, and glycemic variability percentage. Reference values considered gold standard calculations were derived from MATLAB implementation of each metric. Results: The Pearson correlation coefficient was above 0.98 for all metrics, except for mean amplitude of glycemic excursion (r = 0.87) as EasyGV implements a fuzzy logic approach to assessment of variability. Bland-Altman plots demonstrated validation of the new software. Conclusions: The new freely available EasyGV software v10 (www.phc.ox.ac.uk/research/technology-outputs/easygv) is a validated robust tool for analyzing different glycemic variabilities and control metrics.

Assessment of Glucose Control Metrics by Discriminant Ratio.

Objective: Increasing use of continuous glucose monitoring (CGM) data has created an array of glucose metrics for glucose variability, temporal patterns, and times in ranges. However, a gold standard metric has not been defined. We assess the performance of multiple glucose metrics to determine their ability to detect intra- and interperson variability to determine a set of recommended metrics. Methods: The Juvenile Diabetes Research Foundation data set, a randomized controlled study of CGM and self-monitored blood glucose conducted in children and adults with type 1 diabetes (T1D), was used. To determine the ability of the evaluated glycemic metrics to discriminate between different subjects and attenuate the effect of within-subject variation, the discriminant ratio was calculated and compared for each metric. Then, the findings were confirmed using data from two other recent randomized clinical trials. Results: Mean absolute glucose (MAG) has the highest discriminant ratio value (2.98 [95% confidence interval {CI} 1.64-3.67]). In addition, low blood glucose index and index of glycemic control performed well (1.93 [95% CI 1.15-3.44] and 1.92 [95% CI 1.27-2.93], respectively). For percentage times in glucose target ranges, the optimal discriminator was percentage time in glucose target 70-180 mg/dL. Conclusions: MAG is the optimal index to differentiate glucose variability in people with T1D, and may be a complementary therapeutic monitoring tool in addition to glycated hemoglobin and a measure of hypoglycemia. Percentage time in glucose target 70-180 mg/dL is the optimal percentage time in range to report.

Cost-effectiveness of targeted screening for the identification of patients with atrial fibrillation: evaluation of a machine learning risk prediction algorithm.

Aims: As many cases of atrial fibrillation (AF) are asymptomatic, patients often remain undiagnosed until complications (e.g. stroke) manifest. Risk-prediction algorithms may help to efficiently identify people with undiagnosed AF. However, the cost-effectiveness of targeted screening remains uncertain. This study aimed to assess the cost-effectiveness of targeted screening, informed by a machine learning (ML) risk prediction algorithm, to identify patients with AF.Methods: Cost-effectiveness analyses were undertaken utilizing a hybrid screening decision tree and Markov disease progression model. Costs and outcomes associated with the detection of AF compared traditional systematic and opportunistic AF screening strategies to targeted screening informed by a ML risk prediction algorithm. Model analyses were based on adults ≥50 years and adopted the UK NHS perspective.Results: Targeted screening using the ML risk prediction algorithm required fewer patients to be screened (61 per 1,000 patients, compared to 534 and 687 patients in the systematic and opportunistic strategies) and detected more AF cases (11 per 1,000 patients, compared to 6 and 8 AF cases in the systematic and opportunistic screening strategies). The targeted approach demonstrated cost-effectiveness under base case settings (cost per QALY gained of £4,847 and £5,544 against systematic and opportunistic screening respectively). The targeted screening strategy was predicted to provide an additional 3.40 and 2.05 QALYs per 1,000 patients screened versus systematic and opportunistic strategies. The targeted screening strategy remained cost-effective in all scenarios evaluated.Limitations: The analysis relied on assumptions that include the extended period of patient life span and the lack of consideration for treatment discontinuations/switching, as well as the assumption that the ML risk-prediction algorithm will identify asymptomatic AF.Conclusions: Targeted screening using a ML risk prediction algorithm has the potential to enhance the clinical and cost-effectiveness of AF screening, improving health outcomes through efficient use of limited healthcare resources.

Expansion of the homeostasis model assessment of ß-cell function and insulin resistance to enable clinical trial outcome modeling through the interactive adjustment of physiology and treatment effects: iHOMA2.

OBJECTIVE: To describe and make available an interactive, 24-variable homeostasis model assessment (iHOMA2) that extends the HOMA2 model, enabling the modeling of physiology and treatment effects, to present equations of the HOMA2 and iHOMA2 models, and to exemplify iHOMA2 in two widely differing scenarios: changes in insulin sensitivity with thiazolidinediones and changes in renal threshold with sodium glucose transporter 2 (SGLT2) inhibition. RESEARCH DESIGN AND METHODS: iHOMA2 enables a user of the available software to examine and modify the mathematical functions describing the organs and tissues involved in the glucose and hormonal compartments. We exemplify this with SGLT2 inhibition modeling (by changing the renal threshold parameters) using published data of renal effect, showing that the modeled effect is concordant with the effects on fasting glucose from independent data. RESULTS: iHOMA2 modeling of thiazolidinediones effect suggested that changes in insulin sensitivity in the fasting state are predominantly hepatic. SGLT2 inhibition modeled by iHOMA2 resulted in a decrease in mean glucose of 1.1 mmol/L. Observed data showed a decrease in glucose of 0.9 mmol/L. There was no significant difference between the model and the independent data. Manipulation of iHOMA2's renal excretion threshold variable suggested that a decrease of 17% was required to obtain a 0.9 mmol/L decrease in mean glucose. CONCLUSIONS: iHOMA2 is an extended mathematical model for the assessment of insulin resistance and ß-cell function. The model can be used to evaluate therapeutic agents and predict effects on fasting glucose and insulin and on ß-cell function and insulin sensitivity.