Calculated METs systematically overestimated CPET-derived METs and predicted mortality at a higher cutoff (4.6 vs 4.0 METs; HR 0.54, 95% CI 0.32-0.91, p=0.02), limiting their clinical reliability.
Observational (n=567)
No
Do calculated METs accurately reflect CPET-derived METs and provide reliable prognostic value in patients undergoing phase II cardiac rehabilitation?
Calculated METs systematically overestimate functional capacity compared to CPET-derived METs, altering prognostic thresholds and limiting their reliability for individual risk stratification in cardiac rehabilitation.
Hazard Ratio: 0.54 (95% CI 0.32–0.91)
p-value: p=0.02
Abstract Introduction Cardiac rehabilitation (CR) programs frequently rely on estimated metabolic equivalents (METs) derived from exercise testing formulas; however, the accuracy and clinical validity of these calculated METs relative to directly measured values remains uncertain. Purpose To evaluate the agreement between calculated METs and cardiopulmonary exercise test (CPET)-derived METs, and to assess their prognostic implications. Methods We conducted an observational, single-center study with patients enrolled in a phase II CR program between 2015 and 2025. CPETs performed at admission and after completion of the program were analyzed. CPET-derived METs were calculated as peak VO2 divided by 3.5. Estimated METs were obtained using the "Cycle Ergometer Metabolic Calculator". Agreement between both methods was assessed using Pearson correlation, intraclass correlation coefficients (ICC), and Bland–Altman analyses. Prognostic performance was evaluated using ROC analysis to identify optimal MET cutoffs and Cox proportional hazards models for all-cause mortality and hospitalization. Results Of 697 patients, 567 (81%) were included (total 842 CPETs, mean age 61 years, 80% male). Mean estimated and CPET-derived METs were 5.9 and 4.5 pre-CR, and 6.5 and 5.0 post-CR, respectively. Although correlations were strong in both CPETs (r=0.843 and 0.877; p0.001), agreement was only moderate (ICC 0.55 and 0.59). Bland-Altman plots demonstrated substantial positive bias (+1.4 and +1.5 METs), with wide limits of agreement (~0 to +3 METs). In the pre-CR CPET, estimated METs predicted mortality with an ROC-derived cutoff of 4.6 METs (HR 0.54, 95% CI 0.32-0.91, p=0.02). CPET-derived METs showed a similar result (HR 0.55, 95% CI 0.34-0.89, p=0.014) but at a lower cutoff of 4.0 METs, indicating systematic overestimation by estimated values. Conclusion Although calculated METs correlate strongly with CPET-derived METs, their moderate agreement, systematic overestimation, and altered prognostic thresholds may limit their reliability for individual clinical assessment and risk stratification.For image description, please refer to the figure legend and surrounding text.
Cazeiro et al. (Mon,) conducted a observational in Phase II cardiac rehabilitation (n=567). Calculated METs vs. CPET-derived METs was evaluated on All-cause mortality (HR 0.54, 95% CI 0.32-0.91, p=0.02). Calculated METs systematically overestimated CPET-derived METs and predicted mortality at a higher cutoff (4.6 vs 4.0 METs; HR 0.54, 95% CI 0.32-0.91, p=0.02), limiting their clinical reliability.