EDITORIAL

A Less Invasive Approach to Intensive Care

Hayley B. Gershengorn

N Engl J Med Published October 29, 2025

DOI: 10.1056/NEJMe2512006

The intensive care unit (ICU) conjures images of unconscious patients surrounded by machines and penetrated by devices — endotracheal tubes, surgical drains, and urinary, venous, and arterial catheters. Although this scenario may be necessary for some patients, for many, high-quality studies have shown benefits of a different approach. The possibility of patients being awake, interactive, and able to walk even while receiving invasive mechanical ventilation is feasible and may improve outcomes.^1,2 Insertion of a pulmonary artery catheter has no benefit for the typical patient with shock (low blood pressure) or respiratory failure.³In other words, modern-day intensive care can be more humanizing and less invasive than we once presumed.

Muller and colleagues⁴ now provide in the Journal a strategy to move us another step closer to de-adoption of reflexive placement of invasive devices in the ICU. In a pragmatic, noninferiority, randomized, controlled trial, the authors evaluated the efficacy and safety of deferring arterial catheterization in adults with shock. Such catheters are typically inserted in peripheral arteries (e.g., radial, axillary, and femoral) and are used to continuously monitor blood pressure and to facilitate blood sampling. Despite the common use of these devices,⁵results of comparative effectiveness studies have generated uncertainty regarding their benefit,^6,7 and randomized trials have been lacking. In this trial, 1010 patients with shock in nine French ICUs were randomly assigned to treatment involving an invasive approach of early insertion (<4 hours after randomization) of an arterial catheter or a noninvasive approach of monitoring with an automated brachial cuff (and removal of an arterial catheter if one was already in place). Transition to an invasive approach was permitted in the noninvasive-strategy group; however, criteria to do so were strict (e.g., norepinephrine tartrate at a dose of >2.5 μg per kilogram of body weight per minute [norepinephrine base, >1.25 μg per kilogram per minute] plus epinephrine), allowing severe shock to be managed noninvasively.

In findings consistent with their hypothesis, Muller and colleagues showed that the noninvasive approach was noninferior to the invasive approach for the primary outcome of death by day 28 (34.3% and 36.9%, respectively; adjusted risk difference, −3.2 percentage points [95% confidence interval, −8.9 to 2.5]; P=0.006 for noninferiority), whereas notably fewer patients in the noninvasive-strategy group than in the invasive-strategy group received an arterial catheter (14.7% vs. 98.2%). The severity of organ failure on day 7 was similar in the two groups, as was the use of renal replacement therapy (i.e., dialysis) by day 28 and death by day 90. Device-related pain and discomfort (with cuff or catheter) was more common in the noninvasive-strategy group than in the invasive-strategy group (13.1% vs. 9.0%; P=0.05), although fewer arterial-catheter–related hematomas or hemorrhages occurred with the noninvasive strategy (1.0% vs. 8.2%; P<0.001). Limitations of the trial included the unavoidable lack of blinding to group assignment, several between-group baseline differences (e.g., more men, chronic hypertension, and mechanical ventilation and vasopressor use in the invasive-strategy group), and a fairly large noninferiority margin of 5%. In addition, although minimal exclusions allowed for a generalizable cohort (only 8.8% of the patients who met the inclusion criteria were excluded for clinical reasons other than moribund status, brain death, or palliative-care goals), most patients were medical rather than surgical.

Do these findings mean that ICU patients with shock should no longer receive arterial catheters? Proponents of catheter use may raise several concerns. First, they may note that arterial catheters provide continuous blood-pressure monitoring, which allows rapid identification of and response to clinical deterioration. However, delayed recognition and intervention only matter to the extent that they lead to worse clinical outcomes, which were not evident in this trial. Second, proponents may raise concerns about measurement error with a noninvasive approach. However, concordance of noninvasive and invasive devices for detecting clinically important hypotension is very good.⁸ Third, proponents of catheterization may point out that arterial catheters allow for blood sampling without arterial puncture, which is painful and occurred 2.76 times more often with the noninvasive strategy than with the invasive strategy in this trial. Often, however, venous samples (especially for pH⁹) and pulse oximeters (despite acknowledged limitations in patients with dark skin¹⁰) are reasonable alternatives. Finally, proponents may worry that reduced use of arterial catheters will lead to diminished clinician skill with regard to catheter insertion, management, and output interpretation. However, 15% of the patients in the noninvasive-strategy group received arterial catheters, a percentage that suggests that catheter use will remain sufficient to maintain skill — and, if not, simulation as a means to practice the procedure is increasingly available.

This trial should profoundly affect ICU practice by instigating a shift away from a knee-jerk invasive approach to blood-pressure monitoring in patients with shock. Some patients will receive invasive treatment, but this approach can be an exception rather than the rule. And guidelines, which address how to care for most patients, can evolve. Questions remain: Is the approach to postsurgical patients different from the approach to nonsurgical patients? Are arterial catheters beneficial in critical illnesses not involving shock (e.g., hypertension-associated intracranial hemorrhage)? Shared decision making may be needed to understand how individual patients weigh the risks of pain and discomfort against those of catheter-related complications. But a new day has dawned, and it is one in which many more ICU patients should be able to wave good morning without fear of ripping out an arterial catheter.