Reconsidering the recommendation against self-reported metabolic equivalents of task in pre-operative assessment in light of emerging evidence

Editor, In your updated January 2025 guideline on pre-operative assessment of adults undergoing elective noncardiac surgery, it is stated that self-reported metabolic equivalents of task (MET) should not be used for routine peri-operative risk stratification.1 This recommendation is largely grounded in the findings of the METS study, which concluded that subjective assessment of functional capacity did not predict 30-day death or myocardial infarction and did not outperform structured risk indices.2 However, in one of the most recent issues of Anesthesiology, Valadkhani et al.3 reported the largest prospective cohort study to date evaluating self-reported MET and postoperative outcomes. In 38 293 elective noncardiac surgical patients, categorised MET levels (1, 2 to 3, 4 to 5, 6 to 8, ≥9) demonstrated a clear dose–response association with both 30 and 365-day mortality, as well as with days at home alive (DAH). Adjusted survival time progressively shortened with decreasing MET category, and 1-year mortality trajectories were distinctly separated across strata. These findings raise the question of whether the current guideline position may warrant refinement rather than categorical discouragement. Important methodological differences may explain the apparent discrepancy between the METS study and the recent large-scale cohort. Firstly, the METS study primarily evaluated a binary functional capacity threshold and focused on short-term cardiovascular endpoints with a relatively limited number of events.2 In contrast, the recent cohort treated MET as a graded ordinal exposure and reported 2061 deaths at 365 days, enabling more stable survival modelling and long-term risk discrimination.3 Dichotomisation may obscure clinically relevant gradients that become apparent when functional capacity is analysed categorically. Secondly, the outcome framework differs fundamentally. Although 30-day cardiovascular events are highly relevant, longer-term all-cause mortality and DAH arguably better reflect global physiological reserve and overall health vulnerability.3 Functional capacity may represent a multidimensional construct that is more closely linked to longer-term resilience than to isolated peri-operative myocardial events. Thirdly, statistical association does not automatically translate into incremental clinical utility. The recent study did not report discrimination metrics such as C-statistics, calibration analyses or reclassification indices when MET is added to established risk models.3 Consequently, whether graded MET meaningfully improves prediction beyond ASA Physical Status, age, surgical risk or validated composite indices remains to be determined. Nevertheless, it is notable that the contemporary prospective dataset exceeds the combined size of many previous studies and demonstrates consistent dose–response associations across both short-term and long-term outcomes.3 Given this scale and consistency, complete dismissal of self-reported MET may be disproportionate to the evolving evidence base. We fully acknowledge the limitations inherent to subjective MET assessment, including recall bias and interobserver variability. The METS study provided critical insight into these limitations and appropriately cautioned against simplistic binary thresholds.2 However, the emerging data suggest that refinement – rather than abandonment – of functional capacity assessment may be the more proportionate response. A balanced interpretation may therefore include: Avoidance of binary cut-offs (