Evaluating Value-Based Healthcare: The Role of Econometric Models in Integrated Care Pathways

Value-Based Healthcare (VBHC) adoption in Europe faces barriers, including conflicting priorities on outcomes and limitations in traditional economic models like Health Technology Assessment. Econometric methods such as Difference-in-Differences and Synthetic Control offer solutions by evaluating interventions like Integrated Care Pathways (ICPs). These models assess clinical, operational and societal value, linking improvements in outcomes and efficiency to sustainable, evidence-based healthcare decisions.

Key Points

• VBHC in Europe struggles with conflicting priorities and inadequate traditional evaluation models. 
• Econometric methods like Difference-in-Differences improve VBHC impact assessment. 
• Integrated Care Pathways streamline care, enhancing outcomes and operational efficiency. 
• Econometric models link interventions to clinical, operational and societal healthcare value. 
• Evidence-based approaches support sustainable, scalable VBHC initiatives across systems. 

The adoption of Value-Based Healthcare (VBHC) in Europe is progressing slowly, and there are valid reasons for this hesitancy. One significant issue is the lack of consensus on which health outcomes should be prioritised and the methods used to measure them. Patients, payers and society often have differing priorities, which complicates the task of aligning these perspectives within a unified framework. Moreover, existing economic modelling approaches, such as those used in Health Technology Assessment (HTA), tend to focus narrowly on individual therapies. This narrow focus makes them poorly suited for evaluating the broader dynamics of care delivery, including the interplay between acute care, community-based treatment and coordinated pathways. Adopting an econometric approach, rather than relying solely on traditional HTA, presents a promising way to bridge these gaps. This article explores how econometric modelling can be utilised to evaluate VBHC initiatives, using Integrated Care Pathways as a practical example.

Integrated Care Pathways (ICPs) are structured, evidence-based frameworks designed to streamline care delivery for specific conditions or clinical processes (van der Feltz-Cornelis et al. 2023). ICPs can be implemented in both acute care settings and community-based environments. By standardising interventions, promoting multidisciplinary collaboration and coordinating care across different settings, they aim to improve clinical outcomes, enhance operational efficiency and reduce healthcare costs. Their relevance to Value-Based Healthcare (VBHC) lies in their systemic approach, encompassing multiple interventions rather than focusing on a single treatment or diagnostic procedure. This makes their evaluation more complex but critical for understanding the broader creation of value in healthcare.

Economic Modelling Approaches

Econometric approaches offer a unique advantage in determining value because they address causality, ie quantify the degree of deviation in outcomes attributable to a particular treatment or pathway (Angrist 2022). By establishing a causal relationship between the intervention and the observed outcomes, these methods provide an evidence-based starting point for assessing value. This approach firstly shifts the focus to outcomes, allowing providers and payers to subsequently assess the financial and operational implications within a business context.

In contrast, Health Technology Assessment (HTA) often begins with a life sciences organisation setting a price for a therapy. Following this, the HTA process evaluates whether this price aligns with the payer's willingness to pay, using cost-effectiveness thresholds as a framework. While HTA excels in assessing whether a specific intervention justifies its cost (Mundy et al. 2024), econometric modelling is more suitable for a broader VBHC framework.  In this context, the priority is to understand how coordinated care interventions create value across systems and stakeholders without being tied to predetermined pricing.

Furthermore, econometric methods are inherently more useful for assessing the value of indirect innovation in healthcare. For example, tools that improve workflow efficiency can indirectly enhance patient outcomes by enabling faster access to treatment. However, it is challenging to link these improvements to measurable outcomes like quality-of-life scores. Similarly, an overly narrow focus on disease-specific outcome measures or patient experience metrics risks undervaluing innovations that, while not directly therapeutic, play a critical role in enhancing the overall healthcare ecosystem or in identifying deteriorations in patient health (Mantovani et al. 2023). Without a more comprehensive approach, significant advancements in areas such as diagnostics, coordination and system efficiency may fail to receive the recognition they deserve.

Two particularly valuable approaches to VBHC initiatives evaluation are Difference-in-Differences (Diff-in-Diff) (Card et al. 1994) and the Synthetic Control Method (SCM) (Abadie et al. 2010). These methods are widely used to isolate the specific impact of an intervention in complex, real-world healthcare settings. Both of these econometric models rely on panel data, which follows multiple individuals over time. In healthcare, this data is often derived from electronic health records, which include remote patient monitoring records. Such data is instrumental in mapping patient journeys across various healthcare settings, providing comprehensive insights into clinical outcomes, treatment and resource utilisation.

This longitudinal information is crucial for evaluating interventions like Integrated Care Pathways (ICPs) and their impact on both the quality of care and overall costs. For instance, a systematic review highlighted that predictive models using EHR data can effectively forecast hospital readmissions, thereby aiding in the allocation of healthcare resources and the development of targeted interventions (Mahmoudi et al. 2020). Additionally, research indicates that EHR-based clinical decision support systems enhance the management of chronic diseases by improving provider adherence to guidelines and facilitating better monitoring of disease indicators (Hazazi 2021).

Difference-in-Differences

Difference-in-Differences (Diff-in-Diff) is a comparative method used to evaluate changes in outcomes over time for two groups: one that has been exposed to the intervention (eg patients enrolled in a diabetes ICP) and one that has not been exposed (eg patients receiving standard care) (Wang 2024). By examining the difference in outcome changes between the two groups, Diff-in-Diff helps isolate the effect of the intervention while also accounting for broader trends affecting both groups.

For example, consider a diabetes ICP designed to improve the management of Type 2 Diabetes through structured interventions like regular monitoring, patient education and enhanced care coordination. A published review of ICPs in diabetes care (Grant et al. 2014) highlighted their positive impact on clinical outcomes, such as improved glycaemic control, and operational outcomes, including reductions in hospital admissions. Importantly, reviewing this paper with an economic lens suggests that a Diff-in-Diff model could be readily applied to evaluate the value of diabetes ICPs. This model would measure changes in key outcomes before and after pathway implementation across comparable patient groups.For instance, hospitalisation rates could be used as the primary outcome of interest, comparing a cohort of patients enrolled in the diabetes ICP to a comparable cohort receiving standard care.

By analysing trends before and after implementation, Diff-in-Diff estimates the extent to which changes in hospitalisation rates in the treated group deviate from those in the control group. If hospitalisation rates for both groups were declining at a similar rate before the ICP, a steeper decline in the ICP group post-implementation can be attributed to the pathway itself.

A key requirement for the validity of Diff-in-Diff is the assumption of parallel trends (Kahn-Lang et al. 2020): in the absence of ICP, both groups must have experienced similar changes in hospitalisation rates over time. This assumption ensures that any divergence in post-implementation outcomes can be attributed to the ICP rather than external factors. Temporal factors like seasonal variations or broader healthcare system changes are also adjusted for, making this a robust tool for evaluating coordinated care models.

Synthetic Control Method

The Synthetic Control Method (SCM) is a novel econometric approach that allows for the evaluation of a causal impact when a natural control group is not available (Abadie 2021) (Krajewski et al. 2024). SCM constructs a “synthetic” control group by creating a weighted combination of untreated units (e.g., patients, hospitals or regions) that closely matches the treated unit’s pre-intervention characteristics and outcomes (Bonander et al. 2021).

SCM is particularly powerful in settings where traditional comparison methods may not be robust due to the absence of an identifiable comparable control group. It also can be linked to the concept of digital twins in healthcare through its use of weighted untreated groups to create highly individualised and dynamic controls, although the two concepts are not identical.

For example, consider the implementation of a stroke ICP in a single centre. This pathway is designed to improve functional independence through faster thrombolysis, standardised rehabilitation protocols and enhanced follow-up care (Sulch et al. 2000). In situations where there is no natural control group of patients not exposed to the pathway but otherwise comparable, SCM can create a synthetic control group by combining data from untreated patients across multiple centres.

Using weights, SCM ensures that the synthetic control group mirrors the treated group in terms of baseline characteristics such as age, gender, stroke severity, comorbidities and pre-intervention functional independence scores. For instance, if the hospital implementing the stroke ICP had a pre-intervention functional independence rate of 50%, the synthetic control is constructed to replicate this baseline while incorporating other relevant predictors of outcomes.

After the intervention, in this case, the ICP, SCM compares the treated group’s observed outcomes to those of the synthetic control group. A significant improvement in functional independence for the treated group, compared to the synthetic control, would indicate the effectiveness of the pathway. By creating a weighted synthetic control group, SCM effectively serves as a practical application of digital twin technology, generating a counterfactual that is tailored to the specific characteristics of the treated group.

In the case of the stroke ICP, SCM could provide insights into the pathway’s impact on functional independence, hospital readmissions and time-to-thrombolysis. However, the challenge lies in identifying outcome measures that are truly meaningful to stakeholders, especially payers. From a payer’s perspective, outcomes must not only reflect clinical improvements but also demonstrate clear links to value. This includes showing reductions in overall healthcare costs, improved resource utilisation or longer-term cost avoidance through enhanced recovery. 

Outcome Measures in VBHC: Accounting for Clinical Outcomes

Emphasising healthcare utilisation as a primary outcome reflects the efficiency of resource use and its broader societal impact. Metrics such as hospital readmissions, emergency department visits and length of stay are valuable for assessing how interventions alleviate strain on healthcare systems. For instance, reductions in hospitalisations following a stroke Integrated Care Pathway (ICP) not only ease system capacity but also generate economic benefits, enabling improved access to care for other patients and reducing associated costs.

However, while healthcare utilisation metrics offer valuable insights, they must be integrated alongside clinical outcomes to provide a comprehensive assessment of value (Damman et al. 2020). To achieve this, predefined targets for patient-reported outcome measures (PROMs) can ensure that clinical benefits remain central to the analysis. Additionally, econometric modelling can be used to quantify how reductions in healthcare utilisation align with improvements in clinical outcomes.

In the second stage of the analysis, it is important to ensure that patient-centred outcomes remain integral to the evaluation. This can be achieved by adjusting the VBHC model, which primarily focuses on cost reductions linked to healthcare utilisation, to account for patient-reported outcome measures. This adjustment can be achieved by incorporating PROMs as additional covariates in a multivariate regression model. This approach allows for the simultaneous evaluation of cost reductions and their relationship to clinical or patient-relevant outcomes, ensuring a comprehensive assessment of value. Ultimately, this structured framework will link efficiency gains to meaningful benefits for both patients and the healthcare system.

Conclusion

Econometric modelling, particularly Integrated Care Pathways (ICPs), has proven useful to evaluate Value-Based Healthcare (VBHC). By leveraging methods such as Difference-in-Differences (Diff-in-Diff) and Synthetic Control Method (SCM), we have demonstrated how these models provide robust frameworks for establishing causal relationships between interventions and outcomes. This approach addresses the limitations of traditional methods like Health Technology Assessment (HTA).

While outcome measurement remains a critical consideration in VBHC, particularly in balancing healthcare utilisation with clinical outcomes, the primary focus here is on illustrating how econometric models can be employed to evaluate the system-wide value created by coordinated care pathways. This methodological approach offers a pathway for assessing not only clinical effectiveness but also broader operational and societal benefits. By doing so, it enables stakeholders to make evidence-based decisions about the scalability and sustainability of VBHC initiatives.

Conflict of Interest

None