An INCEPTION trial-based model projected ECPR costs and quality-adjusted survival over 20 years. The reported ICER was €242,122 per QALY, with a 46% probability of cost-effectiveness at a Dutch €80,000 threshold. Small per-protocol groups, few survivors, and long-term assumptions limit certainty and transferability.
Can long-term survival justify the initial cost of ECPR?
Extracorporeal cardiopulmonary resuscitation (ECPR) uses veno-arterial ECMO to restore circulation when conventional resuscitation fails. Its economic value depends on more than the price of the circuit: it also depends on how many additional patients survive, their subsequent quality of life, and the resources required during and after hospitalization.
van de Koolwijk and colleagues examined whether extending the economic assessment beyond one year would change the conclusions drawn from the Dutch INCEPTION trial. Their analysis used trial-derived costs and outcomes as inputs for a Markov model with a 20-year horizon. The resulting estimates describe a particular delivery system and patient population; they do not establish a universal cost-effectiveness result for every ECPR program.
Trial data and the lifetime model
The main analysis included 80 patients from the prespecified per-protocol population: 33 treated with ECPR and 47 with conventional CPR (CCPR). Patients had refractory out-of-hospital cardiac arrest with an initially shockable rhythm. Mean age was approximately 56 years, and 91% were men. The per-protocol analysis excluded crossovers, patients who regained spontaneous circulation before emergency-department arrival, and ECPR initiation beyond the specified time window.
This distinction matters because the input population was smaller and more selected than the randomized trial population. The economic results should not be presented as though they came from an unrestricted intention-to-treat comparison of all enrolled patients.
The model used two states: survivor and non-survivor. Healthcare and societal costs were collected during the first year, including emergency services, cannulation, ECMO support, hospital care, procedures, follow-up care, and productivity losses. Beyond that observed period, the investigators extrapolated annual costs and quality-adjusted life years (QALYs). A QALY combines duration of survival with health-related quality of life.
Utilities were informed by EQ-5D-5L measurements. The model applied the same post-arrest health utility of 0.74 to survivors in both groups and assumed annual survival of 98.4% beyond the first year. Costs were discounted by 3% annually and utilities by 1.5%, following Dutch economic-evaluation guidance. These assumptions are essential to interpreting the long-term estimates.
Higher first-year costs and uncertain incremental benefit
Mean first-year costs were €26,372 with ECPR and €10,356 with CCPR, a difference of €16,016. Higher ECPR costs reflected cannulation and support as well as longer hospitalization, transfusions, and diagnostic or therapeutic procedures. The economic burden therefore extended beyond the direct ECMO equipment cost.
Reported survival was approximately 15% with ECPR and 9% with CCPR. With only a small number of survivors, the difference was uncertain and did not establish a statistically significant survival advantage in this analyzed population. A modest change in survivor numbers can substantially change the estimated additional QALYs and the cost-effectiveness ratio.
Over the modeled 20 years, the authors reported incremental costs of €160,969 and an incremental gain of 0.66 QALYs, yielding an incremental cost-effectiveness ratio (ICER) of €242,122 per QALY gained. These are model outputs, not directly observed 20-year follow-up outcomes. The paper calls this a lifetime horizon, implemented as a 20-year projection.
What the 46% probability means
Using 5,000 probabilistic simulations, the investigators estimated a 46% probability that ECPR would be cost-effective at a willingness-to-pay threshold of €80,000 per QALY. This probability describes uncertainty in the economic comparison under the model assumptions. It is not a patient’s probability of survival and does not mean that 46% of ECPR cases individually represent good economic value.
The base-case ICER exceeded the selected Dutch threshold, while the probabilistic result showed substantial decision uncertainty. Costs among non-survivors and differences in QALYs were major drivers. The findings support cautious interpretation of program-level value rather than a categorical conclusion that ECPR can never be cost-effective.
An intention-to-treat scenario produced a lower reported ICER of €146,781 per QALY, although this also exceeded €80,000. Modeling greater long-term mortality worsened cost-effectiveness. These scenarios show how the population definition and survival assumptions influence the estimate.
Survival and system performance drive economic value
In an exploratory scenario that held other parameters unchanged, increasing ECPR survival to approximately 27% brought the ICER near €80,000 per QALY. At 40% survival, the modeled ICER was €48,085 per QALY. These hypothetical scenarios are not validated eligibility criteria, performance targets, or guarantees of cost-effectiveness.
Improving outcomes may itself require additional spending on team readiness, training, transport, communication, and infrastructure. Those costs would need to be included in a local assessment. Moreover, higher ECPR survival alone does not fully describe incremental value: outcomes under conventional resuscitation remain the comparator.
For perfusion services, the study emphasizes the importance of evaluating the entire resuscitation pathway. Circuit costs, cannulation logistics, hospital resource use, survival, and neurologic recovery all contribute to the value of a program. Local outcome data should accompany resource estimates when interpreting how this Dutch analysis might apply elsewhere.
Limitations that affect interpretation
The major strength is the use of prospectively collected data from a multicenter randomized trial. However, this report is a model-based economic evaluation using a selected per-protocol subset, with only 33 ECPR patients and 47 controls. Small survivor numbers and large variation in costs create considerable uncertainty.
Only the first year was directly observed for the economic inputs. The remainder depends on assumptions about annual survival, utility, and ongoing care costs. The two-state structure cannot fully represent different trajectories of neurologic disability and functional recovery. Assigning the same survivor utility to both groups also makes the survival difference particularly influential.
Dutch prices, societal costs, discount rates, and willingness-to-pay thresholds should not be transferred directly to other healthcare systems. Results from highly specialized centers may differ from pragmatic multicenter delivery. Potential benefits associated with organ donation were not included in the model.
Implications for perfusion practice
The study provides a useful framework for discussing ECPR resources alongside outcomes. Extending the horizon improved the projected economic picture relative to a shorter assessment, but did not bring the base-case estimate below the chosen threshold. The practical message is to assess incremental survival and long-term recovery together with the full cost of delivering the service.
This economic model does not establish bedside criteria for withholding ECPR or withdrawing support. Its contribution is to program evaluation: it identifies the assumptions, outcomes, and resource use that most strongly influence the projected value of ECPR.
Keywords: ECPR, ECMO, Out-of-Hospital Cardiac Arrest, INCEPTION Trial, Cost-effectiveness, QALY, Health Economics, Markov Model.
Source: van de Koolwijk AF, Gabrio A, Delnoij TSR, et al.; INCEPTION investigators and Dutch ECLS study group. European Heart Journal: Acute Cardiovascular Care. 2026;15:516–523. doi:10.1093/ehjacc/zuag058.
Funding: The original trial was financially supported by Getinge and ZonMW (843001707).
