Beyond the arterial line: less, more, or smarter use?

The case for “less”: rethinking routine invasive monitoring, even for the patient in shock

The assumption that invasive monitoring is mandatory in all patients with shock has recently been questioned by the EVERDAC trial [1]. This large pragmatic trial assessed whether deferring arterial catheterisation in patients with circulatory shock could be safe under controlled conditions. Surprisingly, the investigators reported no significant difference in mortality or major adverse events between patients managed with early invasive monitoring and those managed with non-invasive monitoring strategies.

These findings invite reflection on the potential overuse of arterial lines (AL) in critical or perioperative care in shocked patients. In addition to procedural complications, the presence of an AL may inadvertently reinforce data-driven rather than physiology-driven decision-making. With a high temporal resolution of Blood Pressure (BP) monitoring, the tendency to react to transient numerical variations can lead to overtreatment with fluids or vasopressors without necessarily improving perfusion. From this standpoint, “less” rather mean “only when necessary”.

The case for “more”: when continuous beat-to-beat monitoring may be of interest

Conversely, the argument for broader use of AL for accurate BP measurement persists in various clinical contexts. Non-invasive BP measurements are often inaccurate in some settings, particularly in critically ill patients [2, 3]. Moreover, patient discomfort related to repeated arterial punctures (blood gas analyses) or non-invasive cuff inflating frequently has to be considered when some patients experience this as painful [1].

In critical care settings, the 2025 European Society of Intensive Care Medicine (ESICM) guidelines on circulatory shock and haemodynamic monitoring reaffirm arterial catheterisation as the standard of care for patients in shock who are not responding to initial therapy, and or requiring vasopressor infusion [4, 5]. The 2026 Surviving Sepsis Campaign was written before the EVERDAC trial was published. They stated the following advice for AL in patients with shock: they are indicated for patients who require intermediate- to high-dose vasopressors, receive escalating doses of vasopressors, or need multiple vasopressors; are undergoing frequent arterial blood sampling; or have non-invasive BP measurements that are inconsistent on repeated assessments [6].

Continuous BP monitoring provides beat-to-beat fidelity that cannot be replicated by non-invasive devices, particularly in low-flow states, resuscitation of cardiac arrest or when vasoactive agents are being titrated. Delayed or inaccurate readings from non-invasive monitoring systems may contribute to under-resuscitation or excessive vasopressor dosing [7]. Furthermore, invasive monitoring allows the calculation of derived variables such as pulse contour cardiac output and pulse pressure variation (dynamic predictor of fluid responsiveness in mechanically ventilated patients) [8].

In complex conditions such as septic shock, cardiogenic shock, extracorporeal support, the ability to capture real-time trends in arterial waveform morphology remains crucial. Here, "more" is justified—not by tradition but by physiological necessity and patient safety. As recent guidelines emphasise, invasive BP monitoring should not be viewed strictly as a choice of technology but as an appropriate escalation of precision monitoring in high-risk contexts [6].

The case for “smarter”: from static targets to applying bedside physiology

Beyond quantity, contemporary evidence calls for a “smarter” approach to AL utilisation. For decades, the management of circulatory shock has been anchored to an apparently universal target: maintaining a mean arterial pressure (MAP) of approximately 65 mmHg. This threshold, adopted widely since the original 2002 Surviving Sepsis Campaign, was supported more by convention and expert consensus than by solid physiological evidence. However, large randomised clinical trials now challenge the notion that a single BP target is suitable for all patients.

The SEPSISPAM trial compared high versus low MAP targets in patients with septic shock and revealed that higher MAP goals (80–85 mmHg) did not improve survival compared with lower targets (65–70 mmHg) in the overall population, although certain subgroups—particularly those with chronic hypertension—showed a reduced incidence of renal dysfunction. Similarly, the OPTPRESS trial further nuanced this finding, suggesting that BP targets should be individualised according to patient phenotype and prior vascular physiology rather than a fixed value.

Ongoing research, including the REACT-SHOCK study, aims to further stratify patients on the basis of perfusion markers, microcirculatory assessments, and dynamic indices of tissue oxygenation. Parallel investigations, such as ANDROMEDA-SHOCK-2, are moving beyond static perfusion endpoints towards integrated multimodal monitoring that combines macrohaemodynamic and microcirculatory variables.

The move towards physiology-based care is already visible in initiatives such as “Applied Physiology at the Bedside,” which encourage clinicians to integrate haemodynamic principles into individualised management rather than relying exclusively on algorithmic thresholds [9].

“S.M.A.R.T.E.R.” AL use implies contextualised decision-making (Fig. 1): the Selection of appropriate high-risk patients or contexts for AL placement; close Monitoring of the MAP with attention to the risks of arterial waveform distortion, such as underdamping or overdamping; Attention to perfusion markers (e.g., capillary refill time, lactate kinetics, or sublingual microcirculation imaging) and the risk of haemodynamic incoherence, where the microcirculation remains severely impaired despite improvements in macrohaemodynamic parameters (BP). It implies also the assessment of fluid Responsiveness and changes in peripheral pulse pressure during dynamic tests (passive leg raising, sigh manoeuvre, positive end-expiratory pressure trials, and end-expiratory occlusion); the Titration of vasopressors; Estimate cardiac output, dP/dt, Pulse contour analysis; and Remove indwelling catheters.

Fig. 1.

SMARTER use of an arterial line as a true monitoring tool. dP/dt: represents the rate of pressure change over time; Eadyn: dynamic arterial elastance = PPV/SVV; CCO: continuous cardiac output monitoring; DSI: diastolic shock index = HR/EDP; EDP: end-diastolic pressure; MAP: mean arterial pressure; PCO2 gap: difference in carbon dioxide pressure between the central venous blood and the arterial blood; PPV: pulse pressure variation; SP: systolic pressure; SVV: stroke volume variation; VNERi: index of the vasomotor tone responsiveness to NE (EDP/[heart rate × norepinephrine dose])

Prospects and ecological challenges

Future directions could include the incorporation of artificial intelligence and machine learning algorithms capable of identifying subtle patterns in arterial waveform variability that are predictive of clinical deterioration. These systems hold promise, but they should amplify—not replace—clinical reasoning. This pragmatic, three-part “less, more, smarter” approach promotes a culture of justification rather than habit. Each AL placement (or lack thereof) becomes a deliberate act of clinical reasoning supported by evidence.

Continuous non-invasive BP technologies are emerging as promising alternatives to arterial catheters, particularly for early vasopressor titration or stable perioperative settings [10]. However, their accuracy remains debated in severe shock, where motion artefacts and periodic calibration limit beat-to-beat fidelity compared with invasive methods [11]. Moreover, the current emphasis on environmental sensitivity in Anaesthesiology and Critical care calls for assessing the ecological footprint of these sensors.

Conclusion

The true value beyond an AL lies not in mere pressure measurement but in its clinical interpretation based on a multiparameter approach. This “smarter” approach may transform the AL from a passive data source into an active diagnostic tool for physiology-guided individualised therapy and precision medicine.

Author contributions

“P.A. T.R and K.B. wrote the main manuscript text and P.A. prepared figure 1. All authors reviewed the manuscript.”

Funding

None.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.