Crisis in Physiological Reasoning in Intensive Care: Left Ventricular Systolic Function

A shortage of trained staff can compromise the cognitive environment in clinical care. Effective training should prioritise physiological reasoning, supported by robust clinical evidence. While technology can enhance decision-making, it should not replace clinical judgement. Complex concepts, such as left ventricular systolic function, require more than simplified measures like left ventricular ejection fraction. Training must be rooted in strong foundational principles rather than rapid, surface-level approaches.

Introduction

The shortage of trained staff, accelerated training, the redistribution of roles and growing reliance on technology can lead to a more protocol-driven and less physiologically informed approach; for this reason, reviving basic cardiovascular reasoning is also a strategy for patient safety and the sustainability of work in critical care settings.

The development of new technologies must be integrated into the new role of specialised medical training in critical care settings (A&E departments, perioperative areas, general wards and intensive care units). Dynamic bedside assessment and reasoning regarding the physiology and pathophysiology involved in each patient will strengthen interpretation with a probabilistic mindset and the generation of knowledge under uncertainty, with biological plausibility and sound clinical reasoning.

In contexts of limited resources, redistribution of roles and accelerated adoption of technology, there is a risk of undermining physiological reasoning and reducing cardiovascular assessment to isolated metrics. Left ventricular systolic function must be reclaimed as an interpretative process rather than a number taken out of context.

"Speaking of echocardiography in critical areas; numbers are cruel and they don't lie; the one who makes a mistake is the one who interprets them or simply ignores them…"

Technology is No Substitute for Clinical Judgement

The current debate on smart ICUs focuses on automation, interoperability and clinical intelligence, but emphasises that this must be human-centred. Furthermore, a recent roadmap in critical care argues that clinicians must learn to interpret AI outputs in conditions of uncertainty, assessing plausibility, consistency with biological principles and coherence with clinical reasoning.

The real risk is not in using technology, but in using it without a proper grasp of the fundamentals. The digital divide in healthcare cannot be bridged simply by introducing new tools, but rather by developing the skills needed to use them safely, critically and in a context-appropriate manner. Recent European frameworks emphasise training, dedicated time and alignment between skills and real-world practice.

Why Start With Left Ventricular Systolic Function?

It plays a role in most of the haemodynamic syndromes seen in critically ill patients. Beginning with left ventricular (LV) systolic function is not based on an arbitrary hierarchy, but rather on a clinical and physiological necessity. In critically ill patients, most haemodynamic syndromes directly or indirectly compromise the LV's ability to convert contractile energy into effective blood flow. This occurs not only in cardiogenic shock, where impaired pumping is explicit, but also in distributive, obstructive or mixed states, in which the interaction between contractility, afterload, preload and ventriculo-arterial coupling determines whether cardiac output will be sufficient to sustain tissue perfusion. In other words, even when the primary lesion does not originate in the myocardium, the LV often becomes the final organ of haemodynamic integration. Consequently, its assessment provides a strategic entry point for understanding why a patient maintains or loses flow, pressure and cardiovascular reserve (Bastos et al. 2020; Monge García et al. 2020).

Left ventricular systolic function is also often misunderstood. It is frequently misinterpreted because it is often reduced to a seemingly intuitive figure: the left ventricular ejection fraction (LVEF). However, the LVEF does not represent pure contractility; it reflects the end result of an interaction between myocardial shortening, ventricular geometry, remodelling and loading conditions. A ventricle may retain a normal LVEF and yet present with low flow, longitudinal impairment or ventricular-arterial decoupling; similarly, a reduced LVEF may correspond to a relatively compensated chronic condition and not, on its own, explain a current state of shock. The error lies not in measuring LVEF, but in attributing to it a physiological significance that exceeds what it actually provides (Marwick et al. 2018; Stokke et al. 2017).

In intensive care, this confusion is exacerbated because LVEF is measured in situations where the load changes abruptly. Septic vasodilation can 'normalise' ejection despite underlying myocardial depression; conversely, a significant increase in afterload can reduce LVEF even if intrinsic contractility is not the primary issue. Therefore, interpreting the ejection fraction in isolation leads to simplistic conclusions, particularly when the geometry of the left ventricle, the pattern of longitudinal deformation, the stroke volume, and the relationship between the ventricle and the arterial tree are overlooked. In reality, LVEF remains useful, but only when considered within the context of a specific physiological question (Sanfilippo et al. 2018; Thomas et al. 2025; Monge García et al. 2019).

A high rate of clinical errors stems from the assumption that LVEF is equivalent to overall cardiovascular function. A common source of clinical error is the assumption that LVEF alone summarises overall cardiovascular function. This equivalence is false because cardiovascular function does not depend solely on the percentage of volume ejected by the left ventricle, but rather on the amount of forward flow it generates, the filling pressure, the vascular resistance, the degree of synchrony, and the efficiency of coupling. Organ perfusion may deteriorate even with preserved LVEF when there is severe vasoplegia, significant diastolic dysfunction, tachyarrhythmia, filling obstruction, pulmonary hypertension with right-sided failure, or insufficient systolic flow. Conversely, patients with chronically depressed LVEF may maintain acceptable haemodynamic stability if ventricular-arterial coupling and afterloads are favourable (Monge García et al. 2020; Bastos et al. 2020).

‘Ignoring echocardiography in critical areas does not make decisions simpler; it only makes errors less visible and consequences more costly…’

Left Ventricular Systolic Function

Physiological definition

From a physiological perspective, left ventricular systolic function can be defined as the ventricle's ability to generate tension, increase pressure, shorten its fibres and transfer mechanical energy to the arterial system during systole. However, this definition requires a conceptual clarification: contractility is not synonymous with ejection. Contractility refers to the intrinsic property of the myocardium to generate force relatively independent of the load; ejection, on the other hand, expresses the observable result of that force once modulated by preload, afterload, chamber geometry and coupling with the vascular system. Consequently, discussing 'systolic function' in critical care involves distinguishing between muscle mechanics, chamber performance and the performance of the cardiovascular system as a whole (Marwick et al. 2018; Bastos et al. 2020).

This distinction is central to critical care medicine because many patients do not fail due to a complete absence of contraction, but rather due to a decoupling between the energy generated by the ventricle and the conditions under which it must eject blood. The problem may lie in the myocardium, in the vasculature, in both, or in their interaction. Hence, a useful definition of systolic function cannot be reduced to 'how much the ventricle empties', but rather to 'how effectively it converts its contraction into useful flow under real-world load conditions' (Monge García et al. 2020; Monge García et al. 2019).

Left ventricular loop; single-beat method

The left ventricular pressure-volume loop remains the most robust integrated model for understanding systole. It simultaneously depicts filling, isovolumetric contraction, ejection and isovolumetric relaxation, allowing us to observe how pressure and volume change throughout the cardiac cycle. Within this framework, telesystolic elastance (Ees) has been considered a measure more closely related to contractility because it describes the relationship between pressure and volume at the end of systole and is relatively less sensitive to load variations than LVEF. Furthermore, effective arterial elastance (Ea) summarises the total load imposed by the arterial system. The relationship between the two provides a means of understanding ventricular-arterial coupling and, consequently, the efficiency with which ventricular energy is converted into external work (Chen et al. 2001; Bastos et al. 2020).

The single-beat method was developed precisely to bring these concepts into clinical practice without the need for repeated invasive procedures. Although its use is not part of routine echocardiography in most ICUs, its educational value is enormous: it serves as a reminder that systolic function is not merely a percentage figure, but the expression of a dynamic system between the heart chambers and the arteries. From an educational perspective, reviewing the LV via a single-beat loop forces one to consider pressure, volume, energy and efficiency, rather than merely the image or ejection fraction (Chen et al. 2001; Gamarra et al. 2024).

The relationship between contractility and performance

Contractility and performance are not the same thing, although in practice they are often confused. Contractility refers to the intrinsic capacity of the myocardium to generate force; performance describes the macroscopic outcome of that capacity, expressed as stroke volume, cardiac output, blood pressure or LVEF. The same ventricle may show different performance with similar contractility if preload or afterload conditions change. Therefore, the finding of a 'preserved' LVEF does not rule out myocardial depression, and a reduced LVEF does not in itself prove that the myocardium is primarily responsible for haemodynamic collapse (Marwick et al. 2018; Stokke et al. 2017).

This distinction has immediate clinical implications. When poor performance is mistaken for poor contractility, it is easy to over-diagnose cardiogenic shock and administer inotropes when the real problem lies in afterload, rhythm, biventricular interaction or vascular resistance. Conversely, when a preserved LVEF is interpreted as a guarantee of systolic integrity, a more subtle mechanical abnormality, such as longitudinal impairment or ventricular-arterial decoupling, may be overlooked. Therefore, the physiological analysis must ask not only how much the LV ejects, but why it does so in this way and under what conditions it achieves this (Sanfilippo et al. 2018; Monge García et al. 2020).

The difference between ejection, deformation and coupling

Ejection, strain and coupling address different physiological questions and should not be treated as synonyms. Ejection, classically represented by LVEF, describes the relative change in volume between the end of diastole and the end of systole; in essence, it indicates the extent to which the chamber empties. Strain, on the other hand, assesses the relative change in the length or thickness of the myocardium, and is therefore closer to fibre mechanics; global longitudinal strain (GLS) has proved particularly useful in identifying subclinical systolic dysfunction, even when LVEF remains preserved (Stokke et al. 2017; Thomas et al. 2025).

Ventricular-arterial coupling adds a third dimension: it does not ask how much the ventricle empties or how much its fibres deform, but rather how well the left ventricle's contractile capacity matches the load imposed by the arterial system. This relationship is particularly valuable in critically ill patients because it reflects cardiovascular efficiency. A ventricle may eject with a reasonable LVEF and yet do so at the cost of significant energy decoupling. Therefore, distinguishing between ejection, deformation and coupling is not an academic refinement, but a prerequisite for avoiding confusion between different metrics and distinct physiological phenomena (Monge García et al. 2020; Gamarra et al. 2024).

The Crisis in Physiological Reasoning in the ICU

Excessive formalism

The use of protocols has helped to standardise processes, reduce oversights and speed up responses in high-pressure clinical settings. However, when a protocol ceases to be a guide and becomes a substitute for clinical judgement, a form of cognitive impoverishment arises that is particularly dangerous in the ICU. Real-world haemodynamic syndromes are rarely pure: they are usually mixed, evolving and context dependent. Therefore, the application of rigid algorithms without a pathophysiological understanding of the case can lead to decisions that appear orderly but are biologically inaccurate. Standardisation is useful for organising care; it is not sufficient for understanding it (Cecconi et al. 2025; Bienefeld et al. 2024). In this scenario, specialised training faces an additional risk: learning to perform procedures without understanding the underlying assumptions.

Automation bias

Automation bias describes the tendency to uncritically accept the recommendation of an automated system and to pay less attention to information that might contradict it. In healthcare, this phenomenon is not merely theoretical: it has been documented in clinical decision support systems and is regaining relevance with the expansion of artificial intelligence-based tools. The risk increases when the user perceives the system as highly useful, when time is short, or when independent verification requires greater cognitive effort. In the ICU, where decisions must be made quickly and workloads are high, the ground for this bias is particularly fertile (Goddard et al. 2012; Lyell et al. 2017; Kücking et al. 2024).

Over-interpretation of structured reports

The structured report represents a significant step forward in standardisation, traceability and communication. However, it can also create a false sense of certainty if interpreted as a direct equivalent of the patient's physiological reality. A well-organised report does not eliminate window limitations, load dependency, the influence of mechanical ventilation, rhythm disturbances or observer bias. Therefore, even the best-structured documents should be read as contextualised interpretative summaries rather than as self-contained truths. The report marks the conclusion of a diagnostic process, not its replacement (Kossaify & Grollier 2014; Sanfilippo et al. 2021).

Overinterpretation occurs when the reader treats an echocardiographic label as an established clinical cause. For this reason, findings must always be linked back to the patient's original problem: what clinical question prompted the examination, what were its technical limitations, and how does the finding relate to the patient's current haemodynamic status. Only in this way does the structured report retain its value without becoming a sophisticated form of oversimplification (Kossaify & Grollier 2014; American College of Cardiology 2022).

The Skill to be Regained is not 'Parroting' but Thinking Physiologically

The idea is not to oppose technology, but to argue that without an understanding of physiology, it simply amplifies errors. This ties in very well with recent discussions on smart ICUs, digital skills and training in how to interpret tools in situations of uncertainty.

Discussion

The echocardiographic assessment of left ventricular systolic function in critical care settings should be understood as a multi-parameter process. The left ventricular ejection fraction (LVEF) measured using the biplane Simpson method remains the most standardised benchmark for quantifying global systolic function and classifying its severity, with normal ranges widely accepted by the ASE/EACVI. However, the same literature emphasises that LVEF reflects ejection performance under specific loading conditions rather than pure myocardial contractility; therefore, its isolated interpretation may lead to incomplete or even erroneous conclusions in critically ill patients.

On this basis, it is essential to recognise a first tier of parameters focused on overall ventricular performance. This group includes 2D LVEF, 3D LVEF and integrated visual assessment. 3D LVEF offers a superior volumetric approximation by reducing reliance on geometric assumptions, whilst visual assessment retains clinical utility due to its high feasibility in unstable patients and its ability to integrate segmental motion and wall thickening in real time. However, none of these approaches completely eliminates the influence of preload, afterload, ventricular geometry or window limitations; consequently, the value of the finding depends both on the technique and on the pathophysiological context in which it is interpreted.

A second category comprises longitudinal function parameters, the clinical relevance of which in the ICU is becoming increasingly evident. GLS showed the strongest evidence for identifying subclinical systolic dysfunction, with clinical guidelines classifying a value more negative than −18% as normal, a value between −16% and −18% as borderline, and a value less negative than −16% as abnormal. Furthermore, the meta-analysis in severe sepsis and septic shock demonstrated an association between poorer GLS values and higher mortality, a finding that was not replicated with LVEF. These data reinforce the idea that longitudinal deformation can reveal significant mechanical impairment before the ejection fraction changes.

The enduring value of simple linear indices such as the shortening fraction and the EPSS. Both remain useful as screening tools or for initial assessment, particularly in targeted studies or POCUS. The shortening fraction remains interpretable in ventricles with relatively preserved geometry, whilst the EPSS has demonstrated reasonable diagnostic ability in identifying reduced LVEF, with a classic cut-off point above 7 mm and better performance reported around 9.5 mm for LVEF less than 50%. However, both parameters share a key limitation: they simplify ventricular mechanics to a linear dimension and, as a result, may fail when segmental dyssynchrony, remodelling, significant valvular disease or septal abnormalities are present.

An important finding is the conceptual utility of dP/dt derived from the mitral regurgitation jet. Unlike LVEF, this parameter attempts to approximate the rate at which the ventricle develops pressure during isovolumetric contraction. Normal values are typically above 1200 mmHg/s, whilst figures below 800 mmHg/s suggest significant impairment of systolic function. Its main strength lies in its closer approximation of the contractile component; its main weakness is technical: it requires a measurable, well-aligned mitral regurgitation with an adequate Doppler signal, which limits its universal applicability in the ICU.

With regard to geometric methods, the Teichholz, area-length and Dumesnil methods retain historical, educational and, on occasion, practical value, but must be interpreted with methodological caution. The Teichholz and area-length methods derive volumes or LVEF from geometric assumptions that lose their validity when the ventricle is dilated, remodelled or exhibits regional abnormalities. The Dumesnil method is of particular interest because it integrates anterograde systolic volume measured by Doppler with end-diastolic volume, but it has not displaced Simpson's or 3D methods in contemporary practice. Therefore, these methods should not be excluded from educational discourse, although they should not be presented as modern equivalents of comprehensive quantification.

The main finding of this review is that the available parameters are not redundant, but complementary. LVEF describes global ejection; GLS, MAPSE and S' explore longitudinal function; the shortening fraction and EPSS simplify analysis for rapid screening; dP/dt best approximates pressure generation; and geometric methods remind us that all quantification depends on anatomical and technical assumptions. This distinction is crucial in critical care medicine because it avoids the mistake of treating different indices as if they answered the same physiological question.

“Left ventricular systolic function: the heart is sometimes a witness, sometimes a culprit, but most of the time it is a victim of our bad decisions…”

Conclusion

In the modern ICU, true innovation does not lie in replacing physiology with technology, but in using technology to improve our understanding. Restoring the physiological interpretation of left ventricular systolic function is a step that benefits education, clinical practice and patient safety.

 Acknowledgements

Thanks to the coronary intensive care team, doctors, residents and nursing staff involved in the success cases of the ISSSTE High Specialty Hospital in Veracruz, Mexico.

Conflict of Interest

None.