Revisiting the Interpretation of Cardiopulmonary Exercise Testing “Availability” After Cardiac Resynchronization Therapy: Methodological and Data Considerations

Dear editor,

I read with great interest a recent article by Kuhara et al. titled “Factors influencing the availability of cardiopulmonary exercise testing for patients undergoing cardiac resynchronization therapy in Japan” [1].

This nationwide analysis provides valuable insights into the current state of cardiopulmonary exercise testing (CPET) among patients undergoing cardiac resynchronization therapy (CRT) or CRT defibrillator (CRT‐D) implantation in Japan.

Although I commend the authors for their valuable contribution to this important area, I would like to highlight several methodological and interpretative considerations that could improve the clarity and reliability of the reported findings.

1. Dataset Discrepancy Between the Present and Previous Reports

A significant numerical inconsistency was noticed between this study and the authors' earlier publications using the same Diagnosis Procedure Combination (DPC) dataset [2].

Both studies explicitly referenced the same ethics approval code (R2‐007, University of Occupational and Environmental Health) and covered an identical study period (2014–2018).

In the 2022 study, 6531 CRT and CRT‐D devices were included after excluding patients aged < 15 years and emergency admissions.

Conversely, the current report states that only 3965 CRT/CRT‐D cases were initially identified, and 3859 were retained for analysis following exclusion.

Given an identical data source and timeframe, a 39% reduction in eligible cases raises concerns about potential changes in the extraction codes, diagnostic inclusion criteria, or exclusion logic.

Clarifying the precise differences in data selection between the two analyses is essential for reproducibility and to determine whether the present cohort represents a subset of a previously published population.

2. Facility‐Level Confounding and Supply‐Side Limitation

As noted by the authors, approximately 21% of hospitals in Japan are equipped with CPET systems.

Consequently, many patients in the DPC dataset were hospitalized in facilities where CPET could not be performed.

Including such cases in the same regression model may have introduced bias in the estimated odds ratios, as “no CPET” outcomes in these hospitals reflect institutional capability rather than patient‐related factors.

A stratified or multilevel logistic model limited to CPET‐capable hospitals or a sensitivity analysis including hospitals with at least one recorded CPET would help isolate true patient‐level determinants from structural constraints. This adjustment would also allow for a clear separation between facility‐level accessibility and individual‐level clinical determinants.

3. Mediation Structure Involving the Barthel Index and Cardiac Rehabilitation

In the multivariate model, both the Barthel Index and rehabilitation type (none, other, or cardiac) were included as explanatory variables.

However, in real‐world practice, the Barthel Index (Activities of Daily Living (ADL) score) is a major determinant of whether a patient receives cardiac or other forms of rehabilitation.

When cardiac rehabilitation is initiated, CPET is frequently performed as part of the evaluation or exercise prescription.

Therefore, a mediating pathway of “Barthel Index→cardiac rehabilitation→CPET” is highly plausible.

Including both variables in the same model may represent statistical overadjustment, thus attenuating the true impact of functional status.

Moreover, the authors stated that “patients with higher ADL or those who underwent cardiac rehabilitation were more likely to undergo CPET.”

This interpretation implicitly assumes that cardiac rehabilitation precedes CPET. However, in clinical practice, CPET is often conducted before or as part of cardiac rehabilitation.

Specifying this temporal sequence would greatly strengthen the conceptual accuracy of the analysis.

4. Complementary Real‐World Data Demonstrating Higher CPET Implementation

Our institutional clinical data provide a markedly different perspective on CPET practice in a setting where testing capabilities are reliably available.

In our previously published cohort [3], which included 659 patients with implantable cardioverter defibrillators (ICD) or CRT‐D enrolled in a comprehensive cardiac rehabilitation (CCR) program, CPET was performed at the start of CCR in 420 patients (64%).

Specifically among CRT‐D recipients, the CPET implementation rate in the same cohort was approximately 63%.

This rate sharply contrasts with the 3% CPET rate reported in a DPC‐based nationwide analysis [1].

This discrepancy strongly suggests that the inclusion of hospitals lacking CPET capabilities led to a substantial underestimation of real‐world utilization.

These findings collectively suggest that CPET availability in Japan is largely constrained by facility resources rather than by patient selection.

5. Concluding Remarks

In summary, I respectfully suggest that:

1. The relationship between the current dataset and the authors' prior DPC analysis should be explicitly clarified.

2. Facility‐level capabilities should be accounted for to avoid structural confounding.

3. The mediation structure among the Barthel Index, cardiac rehabilitation, and CPET should be recognized to prevent overadjustment and misinterpretation.

Addressing these points would strengthen the validity and interpretability of this valuable nationwide investigation.

I sincerely appreciate the authors' contribution to understanding exercise testing following CRT in Japan and hope that the comments will aid in further refinement of DPC‐based analyses.

Funding

This study was supported by the Japan Society for the Promotion of Science, KAKENHI (grant number 24K20490).

Conflicts of Interest

The author declares no conflicts of interest.

Yanagi H., “Revisiting the Interpretation of Cardiopulmonary Exercise Testing “Availability” After Cardiac Resynchronization Therapy: Methodological and Data Considerations,” Journal of Arrhythmia 41, no. 6 (2025): e70252, 10.1002/joa3.70252.