Estimating the Health and Economic Impact of Improved Management in Prevalent Chronic Obstructive Pulmonary Disease Populations in England, Germany, Canada, and Japan: A Modelling Study

Introduction

Chronic obstructive pulmonary disease (COPD) is a progressive respiratory disorder characterised by irreversible airflow limitation and symptoms including dyspnoea, persistent cough, and sputum production. Exacerbation of symptoms is common and occurs more frequently as the disease progresses.^1,2 Severe exacerbations are the primary cause of hospitalisation, and re-admission following discharge is common.^3,4

In 2019, COPD was the third leading cause of death worldwide, responsible for 3.23 million deaths.⁵ The global cost of illness was estimated to be US$2.1 trillion in 2010 and is expected to more than double by 2030.⁶ Hospitalisations and re-admissions are primary drivers of healthcare system costs, and their frequency is highly influenced by exacerbation frequency and severity.^7,8

According to the Global Initiative for Chronic Obstructive Lung Disease (GOLD) case definition (a fixed ratio value of forced expiratory volume in one second (FEV₁) to forced vital capacity (FVC): FEV1/FVC<0.7).,⁹ the global prevalence of COPD was estimated at 10.3% (392 million cases, ages 30–79) in 2019.¹⁰ Prevalence is higher in men, urban areas and high-income countries.¹¹

Currently, no curative treatments exist for COPD. However, effective symptom management and slower disease progression can be achieved through pharmacological and non-pharmacological interventions.¹² GOLD describes international recommendations for the optimal diagnosis and management of COPD,⁹ and many countries publish national guidelines to support best practices.^13–17 Despite this, challenges remain in delivering optimal care to patients. A recent study in high-income settings reported common barriers to include low consideration of COPD, underutilisation of effective pharmacological and non-pharmacological interventions, and sub-optimal management of exacerbations.¹⁸

Improvements in care for people with COPD are expected to reduce hospitalisations, improve patient outcomes, and increase life expectancy.^17,19,20 Interventions to identify undiagnosed patients, such as active case finding, improve patient outcomes and are cost-effective.^21–23 However, evidence on the health and economic benefit of interventions that provide optimal management in line with best practices is more limited.²⁴

Early follow-up review post-discharge (within one month) is recommended in international guidelines⁹ and quality standards¹⁹ for patients hospitalised due to severe exacerbation and can reduce re-admission risk.²⁵ Despite this, uptake is low;^26,27 not attending early follow-up represents a missed opportunity to review and amend discharge therapy and may contribute to re-admissions and poor survival.⁹

Integrated disease management (IDM) programmes provide structured, multidisciplinary care and address the complex nature of managing COPD.²⁸ Administered by teams in both primary and secondary care settings, these programmes vary in composition. They often include pharmacological and non-pharmacological interventions (eg, treatment review, smoking cessation counselling, telemonitoring, etc.). Pulmonary rehabilitation is a well-known form of IDM; the organisation of two primary care programmes from Canada^29,30 and Germany³¹ is detailed in Table S1. A recent meta-analysis of 52 studies found improvements in disease-specific quality of life, exercise capacity and respiratory-related hospital admissions for patients with IDM.²⁸ Despite this evidence of effectiveness, access and uptake remain limited.³¹

This study evaluates the potential health and economic impact of two policy scenarios to improve the management of prevalent, on-treatment COPD populations in England, Germany, Canada, and Japan, through interventions that align with recommendations in international quality standards on management.¹⁹ The interventions were 1) early follow-up review after severe exacerbation hospitalisation and 2) provision of IDM to optimise COPD care.

Methods

Model Overview

A discrete-time cohort-level Markov state-transition model was developed to simulate the natural disease history for COPD. Consistent with prior models,^21,32,33 health states were based on GOLD airflow obstruction stages (GOLD I–IV).⁹ While the GOLD ABE (ABCD pre-2023) clinical risk classifications are central to individualised therapeutic decision-making, spirometry remains a key prognostic measure for COPD, and the airflow obstruction stages are preferred for population-level modelling. Death was included as an absorbing state.

A lifetime horizon was assumed, and the cycle length was three months; a half-cycle correction was applied. In each cycle, simulated individuals either remain in their current GOLD stage, improve and move to the previous stage, or worsen and progress to the next stage while lung function typically declines over time, some improvements to earlier stages were observed in the source data and included in the disease progression model parameters.^21,34 Simulated individuals are also subject to risks of severe exacerbation hospitalisation, subsequent re-admission, and death. Deaths were determined by background COPD and in-hospital mortality. Severe exacerbation and background COPD mortality event risks were conditional upon GOLD stage.

The model predicts total life years, severe exacerbation hospitalisations (including re-admissions), in-hospital deaths, and lifetime costs for defined cohorts of patients. Modelled costs include inhaled medication, background disease management, hospitalisation, and intervention (ie, early follow-up review and IDM administration) costs.

Figure 1 presents an overview of the model structure. The model was implemented in TreeAgePro 2021 (TreeAgePro Software, Williamstown, Massachusetts, USA).³⁵

Figure 1 Schematic diagram of the model structure. Abbreviation: GOLD, Global Initiative for Chronic Obstructive Lung Disease.

Inputs

The model was parameterised using published evidence on usual care (UC) for prevalent COPD populations in each country (Table 1). Data were obtained from population studies, prior modelling studies, national-level audits, and cost databases. Disease progression and severe exacerbation risk profiles were representative of single long-acting muscarinic antagonist (LAMA) therapy. All patients were assumed to be on-treatment. Intervention effects were based on the findings of retrospective population-based studies^25,36 for early follow-up review and randomised controlled trials (RCTs)^28,29 for IDM. A healthcare payer perspective was assumed. Costs reflect 2021/2022 prices. Future costs and benefits were discounted at country-specific rates: England, 3.5%; Germany, 3%; Canada, 1.5%; and Japan, 2%.

Table 1 Input Parameters Applied in the Model

Further detail on the input sources and employed assumptions is provided in Supplemental S1.

Primary Analyses

Considerable gaps and uncertainty exist in the evidence on UC in each country and the intervention effect sizes. To assess the potential impact of improvements in care, primary analyses employed the most robust intervention effect evidence and assumed realistic uptake targets for the policy scenarios.

Early Follow-Up Review - Policy 1

• Intervention effect: the 90-day odds of re-admission were 66% (odds ratio [OR]: 0.34)²⁵ lower for patients attending early follow-up review within one month of severe exacerbation hospitalisation.
• Policy impact: assumed at 50%, the percentage of patients attending early follow-up review was increased for the policy. UC attendance levels ranged from 31.1–37.8% across settings.^26,53

IDM - Policy 2

• Intervention effect: patients receiving IDM were at reduced odds (36% lower; OR: 0.64²⁸) of severe exacerbation hospitalisation across the modelled time horizon.
• Policy impact: the percentage of patients receiving lifelong management with IDM was assumed to be 30% in the policy, while UC levels were 5–10% across settings (Table 2).

Table 2 Overview of Usual Care and Policy Scenario Intervention Implementations

Early follow-up review intervention costs reflected healthcare provider interactions for local practice (Tables S2–S5), while annual IDM administration costs were assumed to be £306, €346, $500, and ¥48,111. Intervention costs do not include additional costs due to treatment regimen changes, other non-pharmacological interventions, or other resource use.

Further information on the intervention effect data sources and assumptions made is provided in Supplemental S2. The impact of the policies was defined by differences in life years, severe exacerbation hospitalisations, in-hospital deaths, and total costs from UC.

Scenario and Uncertainty Analyses

Given considerable uncertainty in several key inputs and assumptions due to limited data, extensive scenario analyses were conducted to assess the impact for a range of intervention effect sizes and uptake levels and at alternative IDM administration costs (Table 2). One-way sensitivity analysis was conducted for all other key parameters.

Ethics approval was not required as no patient-specific information was used in this study.

Results

Primary Analyses

Table 3 presents discounted, primary analysis results for cohorts of 100,000 patients.

Table 3 Cohort-Level (n=100,000) Discounted Primary Analysis Results by Country and Policy

Policy 1

Increasing attendance of early follow-up review to 50% (from 37.8% in England, Germany and Japan, and 31.1% in Canada) resulted in fewer severe exacerbation hospitalisations (England: 1894; Germany: 3354; Canada: 5528; Japan: 1911) and fewer in-hospital deaths (England: 96; Germany: 127; Canada: 210; Japan: 97). Life year gains were predicted for all countries, ranging from 512 in Japan to 1316 in Canada.

While the costs of inhaled medication, background disease management, and early follow-up review were higher for the policy, hospitalisation costs were notably lower and the policy generated lifetime cost savings of £2.89 million, €6.58 million, $40.08 million, and ¥735.58 million.

Policy 2

Increasing IDM access to 30% (from 10% in England and Germany; from 5% in Canada and Japan) yielded fewer severe exacerbation hospitalisations (England: 8516; Germany: 16,153; Canada: 18,290; Japan: 14,412) and in-hospital deaths (England: 425; Germany: 613; Canada: 695; Japan: 735).

While this reduced hospitalisation-related costs, inhaled medication, background disease management, and IDM administration costs were higher for the policy, and the overall cost impact varied by country. In England, Germany, and Japan, the policy was estimated to generate additional life years (2299; 3619; 3656) at higher total lifetime costs (£38.15 million; €35.58 million; ¥1091.53 million), while in Canada, the policy generated 4299 additional life years at cost savings of $20.52 million.

Scenario and Uncertainty Analyses

Figures 2–5 show incremental differences between UC and the policies for different intervention effect sizes and uptake levels. As expected, for both policies, the number of hospitalisations avoided and life years gained increased with greater effect sizes and higher uptake levels. Estimated benefits were larger for Policy 2 than Policy 1, in all countries. The cost impact varied with the effect size. Policy 1 was cost-saving for implementations where the intervention effect corresponded to a ≥23% reduction in the risk (odds/ probability) of re-admission in England and Japan; in Germany and Canada, the policy was cost-saving for all effect sizes. For Policy 2, cost savings were predicted for implementations where the intervention effect corresponded to a ≥36% reduction in severe exacerbation risk in Canada and a ≥50% reduction in Japan. The policy was associated with higher total costs for all effect sizes in England and Germany.

Figure 2 Cohort-level (n=100,000) incremental policy intervention scenario results by outcome - England. (A) Policy 1 - Early follow-up review: Values represent incremental differences between the policy intervention scenario and UC for the range of considered policy uptake levels and intervention effect sizes. (B) Policy 2 - IDM: Per A. Abbreviations: IDM, integrated disease management; Incr, incremental (Policy - UC); OR, Odds ratio; RR, relative risk.

Figure 3 Cohort-level (n=100,000) incremental policy intervention scenario results by outcome - Germany. (A) Policy 1 - Early follow-up review: Values represent incremental differences between the policy intervention scenario and UC for the range of considered policy uptake levels and intervention effect sizes. (B) Policy 2 - IDM: Per A. Abbreviations: IDM, integrated disease management; Incr, incremental (Policy - UC); OR, Odds ratio; RR, relative risk.

Figure 4 Cohort-level (n=100,000) incremental policy intervention scenario results by outcome - Canada. (A) Policy 1 - Early follow-up review: Values represent incremental differences between the policy intervention scenario and UC for the range of considered policy uptake levels and intervention effect sizes. (B) Policy 2 - IDM: Per A. Abbreviations: IDM, integrated disease management; Incr, incremental (Policy - UC); OR, Odds ratio; RR, relative risk.

Figure 5 Cohort-level (n=100,000) incremental policy intervention scenario results by outcome - Japan. (A) Policy 1 - Early follow-up review: Values represent incremental differences between the policy intervention scenario and UC for the range of considered policy uptake levels and intervention effect sizes. (B) Policy 2 - IDM: Per A. Abbreviations: IDM, integrated disease management; Incr, incremental (Policy - UC); OR, Odds ratio; RR, relative risk.

Table 4 shows the impact for alternative IDM administration cost levels. The policy was cost-neutral (ie, total policy costs equivalent to UC) at annual, per-patient administration costs of £95, €170, $584, and ¥42,681, under primary analysis intervention effect sizes and uptake levels.

Table 4 Cohort-Level (n=100,000) Incremental Cost Estimates for IDM Programme Administration Cost Scenarios

Figures S1–S8 present tornado plots of incremental outcomes for one-way sensitivity analyses. As expected, parameters related to severe exacerbations, including severe exacerbation costs and event probabilities, and in-hospital mortality probabilities, were influential; results were also sensitive to variations in baseline age and cost and benefit discount rates. Notably, in some cases (Japan and Canada, Policy 2), the cost impact of the policy depended on the severe exacerbation cost level.

Discussion

Main Findings

This modelling study suggests a significant opportunity to realise health gains and potentially generate cost savings through better management of prevalent, on-treatment COPD populations, via increased uptake of two evidence-based interventions, in England, Germany, Canada and Japan. These findings are likely generalisable to other high-income settings.

While increased attendance of early follow-up review after severe exacerbation hospitalisation (Policy 1) and increased access to IDM (Policy 2) are both predicted to reduce hospitalisations and improve survival, results suggest Policy 2 has a larger potential for impact. This highlights the importance of sustained, multidisciplinary care to prevent avoidable exacerbations and improve patient outcomes; all patients may achieve the benefits of IDM while the potential of early follow-up review is limited to patients experiencing severe exacerbation.

Policy 1, however, is likely of value to both patients and payers, with improved outcomes predicted at reduced total costs for a wide range of intervention effect sizes and across settings. This contrasts with Policy 2 where benefits come with higher total costs, except where cost savings were predicted in Canada and Japan under certain effect sizes. Sensitivity analyses found the cost impact for Policy 2 depended on the severe exacerbation cost level at primary analysis effect sizes in Canada and Japan. The cost impact at a given effect size depends on the underlying cost of COPD management, including during additional years of life, and the cost of the intervention, and the finding of directional variations across and within countries, coupled with the absence of evidence for key parameters, highlights the need for further study to verify our findings.

Findings in the Context of Existing Literature

Early follow-up review after severe exacerbation hospitalisation has been shown to reduce re-admission rates^25,36 and mortality risks.⁵⁵ However, no prior studies have evaluated the value of improved attendance of early follow-up review in terms of health outcomes and costs.

Literature on the cost and effectiveness of IDM is mixed. Short-term studies in Italy⁵⁶ (2 years) and Poland⁵⁷ (six months) found that IDM reduced hospitalisations and was cost-effective. Assessing outcomes over three years, a large study of the German IDM programme found participants had reduced mortality risk (hazard ratio: 0.89, 95% confidence interval [CI]: 0.84–0.94) and higher total costs (€553 per year).³¹ IDM can improve quality of life;^28–30 a recent cost-effectiveness analysis of IDM in Canadian primary care by Scarffe et al, found the programme dominated UC, with cost savings and higher quality-adjusted life years (QALYs) in the short- (1-year; within-trial) and long-term (30-years; model-based analysis).⁵⁸ Approximating the effect on severe exacerbation from the same RCT,²⁹ corresponding analyses from our study (Canada, Policy 2; RR: 0.40) corroborate this finding of improved health outcomes and cost savings and provide additional evidence on the impact at alternative uptake levels.

Notably, Scarffe et al, reported IDM to generate incremental lifetime QALYs (1.732) greater than incremental life years (0.244).⁵⁸ Our study did not evaluate QALYs. However, considering incremental cost-per-life year estimates from the primary analyses (England: £16,594; Germany: €9832; Japan: ¥298,560) and cost-per-QALY willingness-to-pay thresholds of £20,000,³⁷ €20,000, and ¥5,000,000, the policy is likely also cost-effective in these settings. There are no formal willingness-to-pay threshold exists in Germany or Japan, so the currency equivalent cost-per-QALY threshold for England has been assumed.⁵⁹

Differences in findings across studies are likely due to variations in the composition of IDM and UC, evaluation follow-up time,²⁸ as well as measured outcomes and patient and provider factors.⁶⁰ IDM typically yields health benefits, mainly due to reduced exacerbation frequency, but the cost impact varies across settings. This is consistent with findings across the four countries in our study.

Strengths

The policies were selected to align with the international consensus statement on COPD management and designed to be feasible to implement in the study countries, ie based on improving existing management and reducing the variability of care.¹⁹ Modelled intervention effects were based on the best available evidence, including data from a contemporary meta-analysis of 52 RCTs for IDM.²⁸ In practice, the interventions may be implemented differently according to local health services provisions and their costs and effect sizes vary accordingly. In addition, the achievable uptake will depend on the design of the associated policies, which may vary based on local-level infrastructure and existing care processes as well as budgetary considerations. Broad applicability of the findings was ensured through scenario analyses considering an extensive range of intervention effect sizes, uptake levels, and IDM administration costs.

Limitations

The management of COPD is highly complex and the care provision varies both on an individual basis and by setting. Capturing these intricacies and differences in outcomes between countries presents challenges. Where possible, country-specific data were used for UC profiles, and the impact of evaluated interventions was informed by published evidence. However, relevant data were lacking for many parameters, and differences exist in how studies measured outcomes or interventions. Combining this evidence robustly presented challenges requiring assumptions. The model framework makes these explicit, and sensitivity analyses highlight priority areas where further data on COPD care is needed.

The achievable impact of the policies may differ in practice in several ways. 1) Exacerbation history and symptom severity are core components of the GOLD ABE clinical assessment tool and key for therapeutic decision-making.⁹ Based on these factors, both interventions may result in treatment changes (likely escalation), which were not modelled. Therefore, treatment costs may increase alongside slower disease progression rates and improved health outcomes. 2) Only the effect of IDM on severe exacerbations was modelled. IDM could beneficially impact disease progression, but limited data were available to inform this. Whilst our model captures the indirect effect of reductions in severe exacerbations on mortality, capturing the broader effects could yield more optimistic outcomes. 3) Studies of IDM report short-term (3–36 months) effects. Our model extrapolated these to a lifetime horizon, which may overestimate benefits in the case of waning. 4) The effects of IDM and early follow-up review were assumed to apply equally across patients, GOLD stages, and settings. In practice, not all patients may benefit equally from or be able to access these interventions; the effect may vary by disease stage and setting, and the benefits be greater for high-risk populations. Barriers such as distance from care may also limit the achievable impact.^9,25 However, this is a reasonable simplifying assumption for a population-level model, and a range of uptake and effect size scenarios were evaluated.

Policy Implications and Future Research

While the socio-economic burden of COPD is well established, the impact of policies that improve its management is less so. This study provides policymakers with evidence on the potential health and economic impact of improved COPD care in different healthcare settings by quantifying the potential value of policy scenarios to prevent and better manage severe exacerbations. Additionally, in making explicit the notable evidence gaps, priority areas for further research to support evidence-based policy decisions are highlighted.

Additional country-specific data on current and guideline-recommended care should be collected to support more robust assessment of the chosen interventions and to validate the impact shown in this study. Future studies could evaluate treatment-regimen changes due to early follow-up review and IDM, the impact on mild to moderate exacerbations which may also influence patients’ disease trajectory,⁶¹ and the impact in high-risk populations (eg, smokers). With prevalence estimates ranging from 5.1% to 13.5% for 40- to 64-year-olds high-income country prevalence is based on the GOLD case definition,¹⁰ many COPD patients are of working age and the economic impact on patients and governments may be considerable.^9,62,63 Future studies could also assess the impact of improved outcomes on productivity losses ie, societal costs.

Conclusion

Preventing severe exacerbations is key to improving patient outcomes in COPD, and reductions in related hospitalisations will support healthcare system resilience and sustainability. This study provides evidence that suggests improved adherence to early follow-up review post severe exacerbation hospitalisation (Policy 1) and increased access to IDM (Policy 2) can positively impact patient outcomes, reduce hospitalisations, and may generate cost savings across four high-income countries while highlighting priority areas for further research to support evidence-based policy decisions.