Beyond blood pressure: can the Easy D_AO₂ index bring oxygen delivery to the bedside?

The fundamental purpose of hemodynamic management is not simply to maintain arterial pressure but rather to ensure adequate oxygen delivery to metabolically active tissues. Mean arterial pressure (MAP), the most ubiquitously monitored hemodynamic variable in the operating room and intensive care unit, represents pressure rather than flow. A patient may therefore maintain an apparently acceptable MAP despite reduced cardiac output or impaired oxygen-carrying capacity. In this issue of the Korean Journal of Anesthesiology, Eun et al. [1] introduce the Easy D_AO₂ Index (EDI), a simple composite parameter intended to estimate systemic oxygen delivery using routinely available clinical variables.

The physiological rationale for EDI derives from the conventional arterial oxygen delivery index (D_AO₂I), which is determined by cardiac index and arterial oxygen content. Because the dissolved fraction of oxygen contributes minimally to arterial oxygen content under normal conditions, oxygen transport is determined predominantly by hemoglobin concentration, hemoglobin oxygen saturation, and blood flow. EDI simplifies this relationship by substituting the product of pulse pressure and heart rate for cardiac output, and pulse oximetry–derived oxygen saturation (SpO₂) for arterial oxygen saturation, while incorporating hemoglobin concentration and indexing the result to body surface area.

This is an attractive simplification, but its physiological assumptions warrant scrutiny. Pulse pressure reflects stroke volume but is also strongly influenced by arterial compliance, vascular impedance, and wave reflection [2]. Consequently, the product of pulse pressure and heart rate is not synonymous with cardiac output, particularly when vascular tone or arterial compliance changes substantially. Indeed, arterial pressure waveform–based approaches to cardiac output estimation have demonstrated variable accuracy even when considerably more waveform information is incorporated [3]. EDI should therefore be regarded as a physiologically informed surrogate or composite risk index rather than a quantitative substitute for measured D_AO₂I. Its dimensionless value should not be interpreted as actual oxygen delivery.

The concept of estimating oxygen delivery without pulmonary artery catheterization is also not entirely new. In an infant with critical anemia, Steil et al. [4] demonstrated continuous non-invasive oxygen delivery monitoring by combining electrical velocimetry–derived cardiac output with pulse co-oximetry–derived hemoglobin and oxygen saturation. A subsequent review described the feasibility of deriving a non-invasive oxygen delivery index by combining finger-cuff–derived cardiac output with non-invasive hemoglobin measurement [5]. More recently, Ng et al. [6] integrated cardiac output obtained from the ClearSight system with continuous pulse oximetry–derived hemoglobin (SpHb) and SpO₂ measurements to generate real-time non-invasive oxygen delivery trends in cardiac surgical patients. The innovation of EDI therefore lies less in the concept of non-invasive oxygen delivery assessment itself than in its radical simplification: it eliminates the need for a dedicated cardiac output monitor or proprietary flow algorithm. In this respect, terms such as “readily obtainable” or “device-independent” may describe its practical advantage more precisely than “non-invasive,” because hemoglobin measurement still generally requires blood sampling, and blood pressure may be measured invasively in higher-risk patients.

The study nevertheless provides intriguing clinical observations. In two large and distinct populations, i.e., 95 115 non-cardiac surgical cases and 90 420 intensive care unit (ICU) admissions, greater exposure to low EDI was independently associated with 7-day mortality [1]. Notably, among patients with hypotension, concurrent low EDI was associated with higher mortality than was hypotension with preserved EDI. This finding reinforces the physiological distinction between pressure and oxygen transport and suggests that EDI may capture information not reflected by MAP alone.

However, several findings argue against interpreting EDI as a therapeutic target at this stage. First, the observational design establishes a prognostic association but not causality; actively increasing EDI has not been shown to improve outcomes. Second, the thresholds defining low EDI were data-driven, cohort-specific 5th percentiles rather than independently established physiological cutoffs. Their applicability to other populations therefore remains uncertain. Third, EDI alone showed limited discrimination for mortality, with areas under the receiver operating characteristic curves of 0.581 in surgical patients and 0.529 in ICU patients. These values suggest that EDI is better viewed as one component of a broader hemodynamic assessment than as a stand-alone prognostic tool.

Measurement limitations are equally important. Hemoglobin was measured intermittently, and the most recent value was carried forward, potentially rendering EDI insensitive to acute hemorrhage, hemodilution, or transfusion. The relationship between pulse pressure and stroke volume may become unreliable during major changes in arterial compliance or during vasoactive drug administration. Moreover, physiological validation against D_AO₂I was performed in only 86 selected patients who required pulmonary artery catheterization. Correlation indicates that two variables change together but does not establish agreement or interchangeability. Finally, neither EDI nor conventional global oxygen delivery accounts for oxygen consumption, extraction reserve, or the regional and microcirculatory distribution of blood flow.

The immediate value of EDI may therefore lie in serving as a readily available warning signal rather than as another number to normalize. A low EDI could prompt clinicians to undertake a more careful assessment of hemodynamic adequacy. Prospective external validation, determination of clinically meaningful thresholds, assessment of its incremental value beyond established clinical variables, and, ultimately, interventional trials will be required before EDI-guided treatment can be advocated. If these steps prove successful, the greatest contribution of EDI may well be its simplicity. For now, its principal contribution is perhaps conceptual: EDI reminds us that successful hemodynamic management requires looking beyond arterial pressure toward the ultimate purpose of the circulation, delivering oxygen to the tissues.