Examining the impact of structural uncertainty across ten type 2 diabetes models: Results from the 2022 Mount Hood Challenge.

OBJECTIVE: The Mount Hood Diabetes Challenge Network aimed to examine the impact of model structural uncertainty on the estimated cost-effectiveness of interventions for type 2 diabetes. METHODS: Ten independent modelling groups completed a blinded simulation exercise to estimate the cost-effectiveness of three interventions in two type 2 diabetes populations. Modelling groups were provided with a common baseline population, cost and utility values associated with different model health states, and instructions regarding time horizon and discounting. We collated the results to identify variation in predictions of net monetary benefit (NMB), and the drivers of those differences. RESULTS: Overall, modelling groups agreed which interventions had a positive NMB (i.e. were cost-effective), though estimates of NMB varied substantially- by up to £23,696 for one intervention. Variation was mainly driven through differences in risk equations for complications of diabetes and their implementation between models. The number of modelled health states was also a significant predictor of NMB. CONCLUSIONS: This exercise demonstrates that structural uncertainty between different health economic models impacts cost-effectiveness estimates. Whilst it is reassuring that a decision maker would likely reach similar conclusions on which interventions were cost-effective using most models, the range in numerical estimates generated across different models would nevertheless be important for price-setting negotiations with intervention developers. Minimising the impact of structural uncertainty on healthcare decision making therefore remains an important priority. Model registries, which record and compare the impact of structural assumptions, offer one potential avenue to improve confidence in the robustness of health economic modelling.

Effects of a multicomponent communication training to involve older people in decisions to DEPRESCRIBE cardiometabolic medication in primary care (CO-DEPRESCRIBE): protocol for a cluster randomized controlled trial with embedded process and economic evaluation.

BACKGROUND: Deprescribing of medication for cardiovascular risk factors and diabetes has been incorporated in clinical guidelines but proves to be difficult to implement in primary care. Training of healthcare providers is needed to enhance deprescribing in eligible patients. This study will examine the effects of a blended training program aimed at initiating and conducting constructive deprescribing consultations with patients. METHODS: A cluster-randomized trial will be conducted in which local pharmacy-general practice teams in the Netherlands will be randomized to conducting clinical medication reviews with patients as usual (control) or after receiving the CO-DEPRESCRIBE training program (intervention). People of 75 years and older using specific cardiometabolic medication (diabetes drugs, antihypertensives, statins) and eligible for a medication review will be included. The CO-DEPRESCRIBE intervention is based on previous work and applies models for patient-centered communication and shared decision making. It consists of 5 training modules with supportive tools. The primary outcome is the percentage of patients with at least 1 cardiometabolic medication deintensified. Secondary outcomes include patient involvement in decision making, healthcare provider communication skills, health/medication-related outcomes, attitudes towards deprescribing, medication regimen complexity and health-related quality of life. Additional safety and cost parameters will be collected. It is estimated that 167 patients per study arm are needed in the final intention-to-treat analysis using a mixed effects model. Taking loss to follow-up into account, 40 teams are asked to recruit 10 patients each. A baseline and 6-months follow-up assessment, a process evaluation, and a cost-effectiveness analysis will be conducted. DISCUSSION: The hypothesis is that the training program will lead to more proactive and patient-centered deprescribing of cardiometabolic medication. By a comprehensive evaluation, an increase in knowledge needed for sustainable implementation of deprescribing in primary care is expected. TRIAL REGISTRATION: The study is registered at ClinicalTrials.gov (identifier: NCT05507177).

Evaluating the Cost-Utility of Continuous Glucose Monitoring in Individuals with Type 1 Diabetes: A Systematic Review of the Methods and Quality of Studies Using Decision Models or Empirical Data.

INTRODUCTION: This review presents a critical appraisal of differences in the methodologies and quality of model-based and empirical data-based cost-utility studies on continuous glucose monitoring (CGM) in type 1 diabetes (T1D) populations. It identifies key limitations and challenges in health economic evaluations on CGM and opportunities for their improvement. METHODS: The review and its documentation adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for systematic reviews. Searches for articles published between January 2000 and January 2023 were conducted using the MEDLINE, Embase, Web of Science, Cochrane Library, and Econlit databases. Published studies using models and empirical data to evaluate the cost utility of all CGM devices used by T1D patients were included in the search. Two authors independently extracted data on interventions, populations, model settings (e.g., perspectives and time horizons), model types and structures, clinical outcomes used to populate the model, validation, and uncertainty analyses. They subsequently met to confirm consensus. Quality was assessed using the Philips checklist for model-based studies and the Consensus Health Economic Criteria (CHEC) checklist for empirical studies. Model validation was assessed using the Assessment of the Validation Status of Health-Economic decision models (AdViSHE) checklist. The extracted data were used to generate summary tables and figures. The study protocol is registered with PROSPERO (CRD42023391284). RESULTS: In total, 34 studies satisfied the selection criteria, two of which only used empirical data. The remaining 32 studies applied 10 different models, with a substantial majority adopting the CORE Diabetes Model. Model-based studies often lacked transparency, as their assumptions regarding the extrapolation of treatment effects beyond available evidence from clinical studies and the selection and processing of the input data were not explicitly stated. Initial scores for disagreements concerning checklists were relatively high, especially for the Philips checklist. Following their resolution, overall quality scores were moderate at 56%, whereas model validation scores were mixed. Strikingly, costing approaches differed widely across studies, resulting in little consistency in the elements included in intervention costs. DISCUSSION AND CONCLUSION: The overall quality of studies evaluating CGM was moderate. Potential areas of improvement include developing systematic approaches for data selection, improving uncertainty analyses, clearer reporting, and explaining choices for particular modeling approaches. Few studies provided the assurance that all relevant and feasible options had been compared, which is required by decision makers, especially for rapidly evolving technologies such as CGM and insulin administration. High scores for disagreements indicated that several checklists contained questions that were difficult to interpret consistently for quality assessment. Therefore, simpler but comprehensive quality checklists may be needed for model-based health economic evaluation studies.

A Blueprint for Multi-use Disease Modeling in Health Economics: Results from Two Expert-Panel Consultations.

BACKGROUND: The current use of health economic decision models in HTA is mostly confined to single use cases, which may be inefficient and result in little consistency over different treatment comparisons, and consequently inconsistent health policy decisions, for the same disorder. Multi-use disease models (MUDMs) (other terms: generic models, whole disease models, disease models) may offer a solution. However, much is uncertain about their definition and application. The current research aimed to develop a blueprint for the application of MUDMs. METHODS: We elicited expert opinion using a two-round modified Delphi process. The panel consisted of experts and stakeholders in health economic modelling from various professional backgrounds. The first questionnaire concerned definition, terminology, potential applications, issues and recommendations for MUDMs and was based on an exploratory scoping review. In the second round, the panel members were asked to reconsider their input, based on feedback regarding first-round results, and to score issues and recommendations for priority. Finally, adding input from external advisors and policy makers in a structured way, an overview of issues and challenges was developed during two team consensus meetings. RESULTS: In total, 54 respondents contributed to the panel results. The term 'multi-use disease models' was proposed and agreed upon, and a definition was provided. The panel prioritized 10 potential applications (with comparing alternative policies and supporting resource allocation decisions as the top 2), while 20 issues (with model transparency and stakeholders' roles as the top 2) were identified as challenges. Opinions on potential features concerning operationalization of multi-use models were given, with 11 of these subsequently receiving high priority scores (regular updates and revalidation after updates were the top 2). CONCLUSIONS: MUDMs would improve on current decision support regarding cost-effectiveness information. Given feasibility challenges, this would be most relevant for diseases with multiple treatments, large burden of disease and requiring more complex models. The current overview offers policy makers a starting point to organize the development, use, and maintenance of MUDMs and to support choices concerning which diseases and policy decisions they will be helpful for.

Trajectories of clinical characteristics, complications and treatment choices in data-driven subgroups of type 2 diabetes.

AIMS/HYPOTHESIS: This study aimed to explore the added value of subgroups that categorise individuals with type 2 diabetes by k-means clustering for two primary care registries (the Netherlands and Scotland), inspired by Ahlqvist's novel diabetes subgroups and previously analysed by Slieker et al. METHODS: We used two Dutch and Scottish diabetes cohorts (N=3054 and 6145; median follow-up=11.2 and 12.3 years, respectively) and defined five subgroups by k-means clustering with age at baseline, BMI, HbA1c, HDL-cholesterol and C-peptide. We investigated differences between subgroups by trajectories of risk factor values (random intercept models), time to diabetes-related complications (logrank tests and Cox models) and medication patterns (multinomial logistic models). We also compared directly using the clustering indicators as predictors of progression vs the k-means discrete subgroups. Cluster consistency over follow-up was assessed. RESULTS: Subgroups' risk factors were significantly different, and these differences remained generally consistent over follow-up. Among all subgroups, individuals with severe insulin resistance faced a significantly higher risk of myocardial infarction both before (HR 1.65; 95% CI 1.40, 1.94) and after adjusting for age effect (HR 1.72; 95% CI 1.46, 2.02) compared with mild diabetes with high HDL-cholesterol. Individuals with severe insulin-deficient diabetes were most intensively treated, with more than 25% prescribed insulin at 10 years of diagnosis. For severe insulin-deficient diabetes relative to mild diabetes, the relative risks for using insulin relative to no common treatment would be expected to increase by a factor of 3.07 (95% CI 2.73, 3.44), holding other factors constant. Clustering indicators were better predictors of progression variation relative to subgroups, but prediction accuracy may improve after combining both. Clusters were consistent over 8 years with an accuracy ranging from 59% to 72%. CONCLUSIONS/INTERPRETATION: Data-driven subgroup allocations were generally consistent over follow-up and captured significant differences in risk factor trajectories, medication patterns and complication risks. Subgroups serve better as a complement rather than as a basis for compressing clustering indicators.

Tailoring Type II Diabetes Treatment: Investigating the Effect of 5-HTT Polymorphisms on HbA1c Levels after Metformin Initiation.

Aims: To investigate the effect of serotonin transporter (5-HTT) polymorphisms on change in HbA1c levels six months after metformin initiation in type 2 diabetes patients. Materials and Methods: Participants of PROVALID (PROspective cohort study in patients with type 2 diabetes mellitus for VALidation of biomarkers) within the GIANTT (Groningen Initiative to ANalyse Type 2 Diabetes Treatment) cohort who initiated metformin were genotyped for combined 5-HTTLPR/rs25531 (L∗L∗, L∗S∗, and S∗S∗) and 5-HTT VNTR (STin 2.12, 12/-, and 10/-) polymorphisms, respectively. Multiple linear regression was applied to determine the change in HbA1c level from baseline date to six months across 5-HTTLPR/VNTR genotype groups, adjusted for baseline HbA1c, age, gender, triglyceride level, low-density lipoprotein level, and serum creatinine. Results: 157 participants were included, of which 56.2% were male. The average age was 59.3 ± 9.23 years, and the mean baseline HbA1c was 7.49% ± 1.21%. 5-HTTLPR was characterized in 46 patients as L∗L∗, 70 patients as L∗S∗, and 41 patients as S∗S∗ genotypes. No significant association was found between 5-HTTLPR and 5-HTT VNTR genotypes and change in HbA1c after adjustments. Conclusions: 5-HTT polymorphisms did not affect HbA1c levels six months after the start of metformin. Further long-term studies in large samples would be relevant to determine which polymorphisms can explain the variation in response to metformin treatment.

The Association between Deductibles and Cardiovascular Medication Adherence: A Retrospective Inception Cohort Study.

OBJECTIVE: Drug non-adherence in primary preventive cardiovascular therapy is one of the most important modifiable drivers of cardiovascular events. The effect of deductibles in healthcare cost-sharing plans (the amount that has to be paid for healthcare services before the insurance company starts to pay) on such non-adherence in a European setting is unknown. Therefore, we estimated the association between deductibles and the adherence to primary preventive antihypertensive and antihyperlipidemic medication. METHODS: Using the claims database of Menzis Health Insurer in the Netherlands, we applied ordered beta regression mixed modelling to estimate the association between deductibles and adherence taking several demographic and social-economic factors, repeated measurements and within-patient variation into account. RESULTS: All in all, 106,316 patients starting primary preventive antihypertensive or antihyperlipidemic monotherapy were eligible for analysis. At index date, mean age of the study population was 58 years and 52% were male. Reaching the deductible limit and no need to pay for medication anymore increased the adherence [relative adherence ratio (RAR) 1.03, 95% confidence interval (95% CI): 1.00-1.05] for antihyperlipidemic therapy and 1.02 (95% CI: 1.00-1.04) for antihypertensive therapy. A larger deductible amount decreases the adherence of antihyperlipidemic and antihypertensive therapy (RAR 0.83; 95% CI: 0.69-1.00 and RAR 0.85, 95% CI: 0.74-0.98, respectively). CONCLUSION: Independent of other risk factors for non-adherence, presence of deductibles in health insurance is associated with a small negative effect on the adherence to both primary preventive antihypertensive as well as antihyperlipidemic therapy. Further study is needed on the potential health-economic consequences.

Comparative analysis of algorithm-guided treatment and predefined duration treatment programmes for depression: exploring cost-effectiveness using routine care data.

BACKGROUND: More knowledge on the cost-effectiveness of various depression treatment programmes can promote efficient treatment allocation and improve the quality of depression care. OBJECTIVE: This study aims to compare the real-world cost-effectiveness of an algorithm-guided programme focused on remission to a predefined duration, patient preference-centred treatment programme focused on response using routine care data. METHODS: A naturalistic study (n=6295 in the raw dataset) was used to compare the costs and outcomes of two programmes in terms of quality-adjusted life years (QALY) and depression-free days (DFD). Analyses were performed from a healthcare system perspective over a 2-year time horizon. Incremental cost-effectiveness ratios were calculated, and the uncertainty of results was assessed using bootstrapping and sensitivity analysis. FINDINGS: The algorithm-guided treatment programme per client yielded more DFDs (12) and more QALYs (0.013) at a higher cost (€3070) than the predefined duration treatment programme. The incremental cost-effectiveness ratios (ICERs) were around €256/DFD and €236 154/QALY for the algorithm guided compared with the predefined duration treatment programme. At a threshold value of €50 000/QALY gained, the programme had a probability of <10% of being considered cost-effective. Sensitivity analyses confirmed the robustness of these findings. CONCLUSIONS: The algorithm-guided programme led to larger health gains than the predefined duration treatment programme, but it was considerably more expensive, and hence not cost-effective at current Dutch thresholds. Depending on the preferences and budgets available, each programme has its own benefits. CLINICAL IMPLICATION: This study provides valuable information to decision-makers for optimising treatment allocation and enhancing quality of care cost-effectively.

Life years lost for users of specialized mental healthcare.

BACKGROUND: Mental disorders are burdensome and are associated with increased mortality. Mortality has been researched for various mental disorders, especially in countries with national registries, including the Nordic countries. Yet, knowledge gaps exist around national differences, while also relatively less studies compare mortality of those seeking help for mental disorders in specialized mental healthcare (SMH) by diagnosis. Additional insight into such mortality distributions for SMH users would be beneficial for both policy and research purposes. We aim to describe and compare the mortality in a population of SMH users with the mortality of the general population. Additionally, we aim to investigate mortality differences between sexes and major diagnosis categories: anxiety, depression, schizophrenia spectrum and other psychotic disorders, and bipolar disorder. METHODS: Mortality and basic demographics were available for a population of N = 10,914 SMH users in the north of The Netherlands from 2010 until 2017. To estimate mortality over the adult lifespan, parametric Gompertz distributions were fitted on observed mortality using interval regression. Life years lost were computed by calculating the difference between integrals of the survival functions for the general population and the study sample, thus correcting for age. Survival for the general population was obtained from Statistics Netherlands (CBS). RESULTS: SMH users were estimated to lose 9.5 life years (95% CI: 9.4-9.6). Every major diagnosis category was associated with a significant loss of life years, ranging from 7.2 (95% CI: 6.4-7.9) years for anxiety patients to 11.7 (95% CI: 11.0-12.5) years for bipolar disorder patients. Significant differences in mortality were observed between male SMH users and female SMH users, with men losing relatively more life years: 11.0 (95% CI: 10.9-11.2) versus 8.3 (95% CI: 8.2-8.4) respectively. This difference was also observed between sexes within every diagnosis, although the difference was insignificant for bipolar disorder. CONCLUSION: There were significant differences in mortality between SMH users and the general population. Substantial differences were observed between sexes and between diagnoses. Additional attention is required, and possibly specific interventions are needed to reduce the amount of life years lost by SMH users.

Cost-Effectiveness of Four Tobacco Control Interventions in Mongolia.

INTRODUCTION: The aim of this study is to quantify the cost-effectiveness of four tobacco control interventions: Tobacco taxation, mass media campaigns, school programs, and cessation support, and to illustrate how available evaluation tools can be adapted to the local setting. AIMS AND METHODS: We used the dynamic population health modeling-health impact assessment tool to project the future smoking prevalence associated with the interventions and to simulate the resulting smoking-related disease burden over time. Applying the most recent available national Mongolian data as input, the costs and effects of four interventions were compared to a business-as-usual scenario, resulting in costs per life year gained and per disability-adjusted life years (DALYs) averted. RESULTS: Three years after implementation, all interventions reduce the prevalence of current smoking, with the strongest reduction observed with the increase in tobacco tax (5.1% points), followed by mass media campaigns (1.6% points), school programs (1.3% points), and cessation support interventions (0.6% points). School programs were a cost-saving tobacco control intervention compared to current practice in Mongolia, while the other programs resulted in additional costs compared to business as usual. Compared to the World Health Organization (WHO) thresholds, all interventions would be considered "very cost-effective" in terms of cost per DALY averted (below US$ 4295 per DALY averted) in Mongolia. CONCLUSIONS: Large-scale interventions such as taxation and mass media campaigns result in both cost-effectiveness and important health benefits in relation to intervention costs. Reducing the prevalence of smoking among the male population would be particularly worthwhile in Mongolia. IMPLICATIONS: This study shows that in Mongolia school programs were a cost-saving intervention, while the cost-effectiveness ratios were US$ 25 per disability-adjusted life year (DALY) averted for mass media campaigns, US$ 74 for taxation, and US$ 1961 for cessation support interventions. Compared to the WHO thresholds, all interventions would be considered "very cost-effective" in terms of expenses per DALY averted (

Cost-Effectiveness of SGLT2 Inhibitors in a Real-World Population: A MICADO Model-Based Analysis Using Routine Data from a GP Registry.

OBJECTIVE: Sodium-glucose cotransporter 2 inhibitors (SGLT2i) have been shown to reduce the risk of cardiovascular complications, which largely drive diabetes' health and economic burdens. Trial results indicated that SGLT2i are cost effective. However, these findings may not be generalizable to the real-world target population. This study aims to evaluate the cost effectiveness of SGLT2i in a routine care type 2 diabetes population that meets Dutch reimbursement criteria using the MICADO model. METHODS: Individuals from the Hoorn Diabetes Care System cohort (N = 15,392) were filtered to satisfy trial inclusion criteria (including EMPA-REG, CANVAS, and DECLARE-TIMI58) or satisfy the current Dutch reimbursement criteria for SGLT2i. We validated a health economic model (MICADO) by comparing simulated and observed outcomes regarding the relative risks of events in the intervention and comparator arm from three trials, and used the validated model to evaluate the long-term health outcomes using the filtered cohorts' baseline characteristics and treatment effects from trials and a review of observational studies. The incremental cost-effectiveness ratio (ICER) of SGLT2i, compared with care-as-usual, was assessed from a third-party payer perspective, measured in euros (2021 price level), using a discount rate of 4% for costs and 1.5% for effects. RESULTS: From Dutch individuals with diabetes in routine care, 15.8% qualify for the current Dutch reimbursement criteria for SGLT2i. Their characteristics were significantly different (lower HbA1c, higher age, and generally more preexisting complications) than trial populations. After validating the MICADO model, we found that lifetime ICERs of SGLT2i, when compared with usual care, were favorable (< €20,000/QALY) for all filtered cohorts, resulting in an ICER of €5440/QALY using trial-based treatment effect estimates in reimbursed population. Several pragmatic scenarios were tested, the ICERs remained favorable. CONCLUSIONS: Although the Dutch reimbursement indications led to a target group that deviates from trial populations, SGLT2i are likely to be cost effective when compared with usual care.

Potential Value of Identifying Type 2 Diabetes Subgroups for Guiding Intensive Treatment: A Comparison of Novel Data-Driven Clustering With Risk-Driven Subgroups.

OBJECTIVE: To estimate the impact on lifetime health and economic outcomes of different methods of stratifying individuals with type 2 diabetes, followed by guideline-based treatment intensification targeting BMI and LDL in addition to HbA1c. RESEARCH DESIGN AND METHODS: We divided 2,935 newly diagnosed individuals from the Hoorn Diabetes Care System (DCS) cohort into five Risk Assessment and Progression of Diabetes (RHAPSODY) data-driven clustering subgroups (based on age, BMI, HbA1c, C-peptide, and HDL) and four risk-driven subgroups by using fixed cutoffs for HbA1c and risk of cardiovascular disease based on guidelines. The UK Prospective Diabetes Study Outcomes Model 2 estimated discounted expected lifetime complication costs and quality-adjusted life-years (QALYs) for each subgroup and across all individuals. Gains from treatment intensification were compared with care as usual as observed in DCS. A sensitivity analysis was conducted based on Ahlqvist subgroups. RESULTS: Under care as usual, prognosis in the RHAPSODY data-driven subgroups ranged from 7.9 to 12.6 QALYs. Prognosis in the risk-driven subgroups ranged from 6.8 to 12.0 QALYs. Compared with homogenous type 2 diabetes, treatment for individuals in the high-risk subgroups could cost 22.0% and 25.3% more and still be cost effective for data-driven and risk-driven subgroups, respectively. Targeting BMI and LDL in addition to HbA1c might deliver up to 10-fold increases in QALYs gained. CONCLUSIONS: Risk-driven subgroups better discriminated prognosis. Both stratification methods supported stratified treatment intensification, with the risk-driven subgroups being somewhat better in identifying individuals with the most potential to benefit from intensive treatment. Irrespective of stratification approach, better cholesterol and weight control showed substantial potential for health gains.

Early symptom change contributes to the outcome prediction of cognitive behavioral therapy for depression patients: A machine learning approach.

BACKGROUND: Limited evidence exists regarding the association between early symptom change and later outcomes of cognitive behavioral therapy (CBT). This study aimed to apply machine learning algorithms to predict continuous treatment outcomes based on pre-treatment predictors and early symptom changes and to uncover whether additional variance could be explained compared to regression methods. Additionally, the study examined early subscale symptom changes to determine the most significant predictors of treatment outcome. METHODS: We investigated CBT outcomes in a large naturalistic dataset (N = 1975 depression patients). The sociodemographic profile, pre-treatment predictors, and early symptom change, including total and subscale scores were used to predict the Symptom Questionnaire (SQ)48 score at the 10th session as a continuous outcome. Different machine learners were compared to linear regression. RESULTS: Early symptom change and baseline symptom score were the only significant predictors. Models with early symptom change explained 22.0 % to 23.3 % more variance than those without early symptom change. Specifically, the baseline total symptom score, and the early symptom score changes of the subscales pertaining to depression and anxiety were the top three predictors of treatment outcome. LIMITATION: Excluded patients with missing treatment outcomes had slightly higher symptom scores at baseline, indicating possible selection bias. CONCLUSION: Early symptom change improved the prediction of treatment outcomes. The prediction performance achieved is far from clinical relevance: the best learner could only explain 51.2 % of the variance in outcomes. Compared to linear regression, more sophisticated preprocessing and learning methods did not substantially improve performance.

Prediction of complications in health economic models of type 2 diabetes: a review of methods used.

AIM: Diabetes health economic (HE) models play important roles in decision making. For most HE models of diabetes 2 diabetes (T2D), the core model concerns the prediction of complications. However, reviews of HE models pay little attention to the incorporation of prediction models. The objective of the current review is to investigate how prediction models have been incorporated into HE models of T2D and to identify challenges and possible solutions. METHODS: PubMed, Web of Science, Embase, and Cochrane were searched from January 1, 1997, to November 15, 2022, to identify published HE models for T2D. All models that participated in The Mount Hood Diabetes Simulation Modeling Database or previous challenges were manually searched. Data extraction was performed by two independent authors. Characteristics of HE models, their underlying prediction models, and methods of incorporating prediction models were investigated. RESULTS: The scoping review identified 34 HE models, including a continuous-time object-oriented model (n = 1), discrete-time state transition models (n = 18), and discrete-time discrete event simulation models (n = 15). Published prediction models were often applied to simulate complication risks, such as the UKPDS (n = 20), Framingham (n = 7), BRAVO (n = 2), NDR (n = 2), and RECODe (n = 2). Four methods were identified to combine interdependent prediction models for different complications, including random order evaluation (n = 12), simultaneous evaluation (n = 4), the 'sunflower method' (n = 3), and pre-defined order (n = 1). The remaining studies did not consider interdependency or reported unclearly. CONCLUSIONS: The methodology of integrating prediction models in HE models requires further attention, especially regarding how prediction models are selected, adjusted, and ordered.

A Promising Approach to Optimizing Sequential Treatment Decisions for Depression: Markov Decision Process.

The most appropriate next step in depression treatment after the initial treatment fails is unclear. This study explores the suitability of the Markov decision process for optimizing sequential treatment decisions for depression. We conducted a formal comparison of a Markov decision process approach and mainstream state-transition models as used in health economic decision analysis to clarify differences in the model structure. We performed two reviews: the first to identify existing applications of the Markov decision process in the field of healthcare and the second to identify existing health economic models for depression. We then illustrated the application of a Markov decision process by reformulating an existing health economic model. This provided input for discussing the suitability of a Markov decision process for solving sequential treatment decisions in depression. The Markov decision process and state-transition models differed in terms of flexibility in modeling actions and rewards. In all, 23 applications of a Markov decision process within the context of somatic disease were included, 16 of which concerned sequential treatment decisions. Most existing health economic models relating to depression have a state-transition structure. The example application replicated the health economic model and enabled additional capacity to make dynamic comparisons of more interventions over time than was possible with traditional state-transition models. Markov decision processes have been successfully applied to address sequential treatment-decision problems, although the results have been published mostly in economics journals that are not related to healthcare. One advantage of a Markov decision process compared with state-transition models is that it allows extended action space: the possibility of making dynamic comparisons of different treatments over time. Within the context of depression, although existing state-transition models are too basic to evaluate sequential treatment decisions, the assumptions of a Markov decision process could be satisfied. The Markov decision process could therefore serve as a powerful model for optimizing sequential treatment in depression. This would require a sufficiently elaborate state-transition model at the cohort or patient level.

Personalizing the Use of a Intermittently Scanned Continuous Glucose Monitoring (isCGM) Device in Individuals With Type 1 Diabetes: A Cost-Effectiveness Perspective in the Netherlands (FLARE-NL 9).

AIMS: Intermittently scanned continuous glucose monitoring (isCGM) is a method to monitor glucose concentrations without using a finger prick. Among persons with type 1 diabetes (T1D), isCGM results in improved glycemic control, less disease burden and improved health-related quality of life (HRQoL). However, it is not clear for which subgroups of patients isCGM is cost-effective. We aimed to provide a real-world cost-effectiveness perspective. METHODS: We used clinical data from a 1-year nationwide Dutch prospective observational study (N = 381) and linked these to insurance records. Health-related quality of life was assessed with the EQ-5D-3L questionnaire. Individuals were categorized into 4 subgroups: (1) frequent hypoglycemic events (58%), (2) HbA1c > 70 mmol/mol (8.5%) (19%), (3) occupation that requires avoiding finger pricks and/or hypoglycemia (5%), and (4) multiple indications (18%). Comparing costs and outcomes 12 months before and after isCGM initiation, incremental cost-effectiveness ratios (ICERs) were calculated for the total cohort and each subgroup from a societal perspective (including healthcare and productivity loss costs) at the willingness to pay of €50,000 per quality-adjusted life year (QALY) gained. RESULTS: From a societal perspective, isCGM was dominant in all subgroups (ie higher HRQoL gain with lower costs) except for subgroup 1. From a healthcare payer perspective, the probabilities of isCGM being cost-effective were 16%, 9%, 30%, 98%, and 65% for the total cohort and subgroup 1, 2, 3, and 4, respectively. Most sensitivity analyses confirmed these findings. CONCLUSIONS: Comparing subgroups of isCGM users allows to prioritize them based on cost-effectiveness. The most cost-effective subgroup was occupation-related indications, followed by multiple indications, high HbA1c and the frequent hypoglycemic events subgroups. However, controlled studies with larger sample size are needed to draw definitive conclusions.

Cost Effectiveness and Budget Impact of the Boston University Approach to Psychiatric Rehabilitation for Increasing the Social Participation of Individuals With Severe Mental Illnesses.

Background: The purpose of this study was to investigate the cost-effectiveness and budget impact of the Boston University Approach to Psychiatric Rehabilitation (BPR) compared to an active control condition (ACC) to increase the social participation (in competitive employment, unpaid work, education, and meaningful daily activities) of individuals with severe mental illnesses (SMIs). ACC can be described as treatment as usual but with an active component, namely the explicit assignment of providing support with rehabilitation goals in the area of social participation. Method: In a randomized clinical trial with 188 individuals with SMIs, BPR (n = 98) was compared to ACC (n = 90). Costs were assessed with the Treatment Inventory of Costs in Patients with psychiatric disorders (TIC-P). Outcome measures for the cost-effectiveness analysis were incremental cost per Quality Adjusted Life Year (QALY) and incremental cost per proportional change in social participation. Budget Impact was investigated using four implementation scenarios and two costing variants. Results: Total costs per participant at 12-month follow-up were € 12,886 in BPR and € 12,012 in ACC, a non-significant difference. There were no differences with regard to social participation or QALYs. Therefore, BPR was not cost-effective compared to ACC. Types of expenditure with the highest costs were in order of magnitude: supported and sheltered housing, inpatient care, outpatient care, and organized activities. Estimated budget impact of wide BPR implementation ranged from cost savings to €190 million, depending on assumptions regarding uptake. There were no differences between the two costing variants meaning that from a health insurer perspective, there would be no additional costs if BPR was implemented on a wider scale in mental health care institutions. Conclusions: This was the first study to investigate BPR cost-effectiveness and budget impact. The results showed that BPR was not cost-effective compared to ACC. When interpreting the results, one must keep in mind that the cost-effectiveness of BPR was investigated in the area of social participation, while BPR was designed to offer support in all rehabilitation areas. Therefore, more studies are needed before definite conclusions can be drawn on the cost-effectiveness of the method as a whole.

The transferability and validity of a population-level simulation model for the economic evaluation of interventions in diabetes: the MICADO model.

AIMS: Valid health economic models are essential to inform the adoption and reimbursement of therapies for diabetes mellitus. Often existing health economic models are applied in other countries and settings than those where they were developed. This practice requires assessing the transferability of a model developed from one setting to another. We evaluate the transferability of the MICADO model, developed for the Dutch 2007 setting, in two different settings using a range of adjustment steps. MICADO predicts micro- and macrovascular events at the population level. METHODS: MICADO simulation results were compared to observed events in an Italian 2000-2015 cohort (Casale Monferrato Survey [CMS]) and in a Dutch 2008-2019 (Hoorn Diabetes Care Center [DCS]) cohort after adjusting the demographic characteristics. Additional adjustments were performed to: (1) risk factors prevalence at baseline, (2) prevalence of complications, and (3) all-cause mortality risks by age and sex. Model validity was assessed by mean average percentage error (MAPE) of cumulative incidences over 10 years of follow-up, where lower values mean better accuracy. RESULTS: For mortality, MAPE was lower for CMS compared to DCS (0.38 vs. 0.70 following demographic adjustment) and adjustment step 3 improved it to 0.20 in CMS, whereas step 2 showed best results in DCS (0.65). MAPE for heart failure and stroke in DCS were 0.11 and 0.22, respectively, while for CMS was 0.42 and 0.41. CONCLUSIONS: The transferability of the MICADO model varied by event and per cohort. Additional adjustments improved prediction of events for MICADO. To ensure a valid model in a new setting it is imperative to assess the impact of adjustments in terms of model accuracy, even when this involves the same country, but a new time period.

Health-related quality of life for normal glycaemia, prediabetes and type 2 diabetes mellitus: Cross-sectional analysis of the ADDITION-PRO study.

AIMS: We estimated and compared health-related quality of life for individuals with normal glucose tolerance, prediabetes and diabetes. METHODS: Participants in the ADDITION-PRO study, Denmark, who attended a health assessment between 2009 and 2011, and who completed the 3-level EuroQoL 5-dimensions (EQ-5D-3L) questionnaire were included. For the present study, they were classified as normal glucose tolerance, prediabetes and diabetes (screen-detected and known) using the 2019 American Diabetes Association criteria. Prediabetes was defined as impaired fasting glucose, impaired glucose tolerance or HbA1c between 5.7-6.4% (39-47 mmol/mol). EQ-5D-3L data were converted into utility scores using Danish and UK values, where '1' equals full health and '0' equals death. Regression models estimated the association between utility and the different glucose health states. RESULTS: The mean EQ-5D-3L score in the sample population was 0.86 ± 0.17 (median 0.85, interquartile range 0.76 to 1) using UK values. Almost half of the sample (48%) reported full health with an EQ-5D score of '1'. Individuals with known diabetes reported the lowest EQ-5D-3L utility scores (0.81 ± 0.20), followed by individuals with screen-detected diabetes (0.85 ± 0.19), prediabetes (0.86 ± 0.17) and normal glucose tolerance (0.90 ± 0.15). The differences were statistically significant for normal glucose and known diabetes relative to prediabetes, after adjusting for sex, age, smoking, BMI and physical activity. These findings also held using Danish values albeit the differences were of smaller magnitude. CONCLUSIONS: Having prediabetes and diabetes was significantly associated with lower health-related quality of life relative to normal glucose tolerance. Our estimates will be useful to inform the value of interventions to prevent diabetes or prediabetes.