Quantification of metabolic equivalents (METs) by the MET-REPAIR questionnaire: A validation study in patients with a high cardiovascular burden.

ABSTRACT

STUDY OBJECTIVE: Quantifying functional capacity is a core component of preoperative cardiovascular risk assessment. Lower metabolic equivalents (METs) are associated with higher morbidity/mortality in non-surgical and surgical populations. However, actually measuring METs preoperatively is rare. We sought to determine the correlation of self-reported METs using the questionnaire of the MET REevaluation for Perioperative cArdIac Risk (MET-REPAIR) study and objectively measured METs by gold-standard cardiopulmonary exercise testing (CPET). DESIGN: Single-center prospective validation study. SETTING: University hospital. PATIENTS We enrolled adult patients aged ≥45 undergoing out-patient cardiac rehabilitation. INTERVENTION Patients completed the MET-REPAIR Questionnaire and the Duke Activity Status Index (DASI), had blood samples drawn, and underwent undergoing routine CPET. MEASUREMENTS We compared measured METs by CPET to 1) self-reported METs (the MET-REPAIR Questionnaire), 2) the DASI score, 3) stand-alone questions, and 4) N-terminal pro-brain natriuretic peptide (NT-proBNP) concentrations. MAIN RESULTS: 140 patients were recruited. Measured METs by CPET correlated with 1) self-reported METs by the MET-REPAIR Questionnaire (ρ = 0.489, "fair"), 2) self-reported physical activity by the DASI (ρ = 0.487, "fair"), 3) the self-reported continual stair climbing ability (one of the stand-alone questions; ρ = 0.587, "fair") and 4) NT-proBNP concentrations (ρ = -0.353, "poor"). The area under the receiver operating characteristic curve (AUROC) for the ability to perform more than 4 METs were highest for flights of stairs (0.841 [95%CI 0.735-0.948], p < 0.05 to rest, optimum 3 flights), not significantly different between MET-REPAIR Questionnaire and DASI (0.666 [95%CI 0.551-0.781], optimum 6 METs vs. 0.704 [95%CI 0.578-0.829], optimum 32.2 points or 6.7 METs, p = 0.405), and not significant for NT-proBNP (0.623 [95%CI 0.483-0.763]). CONCLUSIONS: The MET-REPAIR Questionnaire correlates with measured METs; all utilized forms of self-reported physical activity overestimate measured METs. NT-proBNP correlates poorly with measured METs.