Intraoperative blood pressure management in noncardiac surgery: a narrative review based on current evidence

Abstract

Purpose

Blood pressure is closely monitored during anaesthesia, yet the optimal intraoperative target remains uncertain. This narrative review synthesizes contemporary observational and randomized evidence and explores emerging strategies for individualized haemodynamic management.

Methods

We reviewed major observational cohort studies, randomized controlled trials (RCTs), consensus statements, and recent technological developments addressing intraoperative hypotension, MAP thresholds, and strategies to prevent perioperative organ injury in adult noncardiac surgery.

Results

Large observational datasets consistently demonstrate graded, duration-dependent associations between intraoperative MAP 60–70 mmHg and postoperative myocardial injury, acute kidney injury, and mortality . These findings have informed international recommendations to avoid MAP below 60–65 mmHg. However, contemporary multicentre RCTs enrolling more than 13,000 patients show that targeting higher or individualized MAP thresholds does not improve patient-centred outcomes compared with routine care (typically MAP ≥ 65 mmHg) . Only one small trial reported benefit with individualized systolic targets. Emerging evidence suggests that hypotension reflects heterogeneous haemodynamic endotypes (vasodilation, hypovolaemia, myocardial depression, bradycardia), potentially explaining why uniform pressure targets fail to improve outcomes. Continuous blood pressure monitoring, proactive norepinephrine infusion, predictive analytics, and closed-loop vasopressor systems reliably reduce hypotension exposure, although definitive outcome benefits remain unproven.

Conclusions

Observational and randomized data are concordant: MAP ≥ 60–65 mmHg appears sufficient for most noncardiac surgical patients. Future progress will likely depend on mechanistic endotyping, integration of advanced monitoring, and precision-guided haemodynamic strategies rather than escalation of universal MAP targets alone.

Take-home message

Perioperative blood pressure control aims to maintain adequate organ perfusion while accounting for patient-specific physiological requirements. Maintaining mean arterial pressure ≥ 60-65 mmHg appears sufficient for most surgical patients. Trials targeting higher or individualized thresholds have not led to improved outcomes. Emerging strategies include consideration of pre-emptive continuous norepinephrine infusion, physiologic endotyping of hypotension, and integration of predictive analytics and closed-loop systems to provide stable individualized hemodynamic management.

Introduction

Blood pressure management remains a central yet persistently debated aspects of perioperative care [1]. Despite major advances in anesthetic and monitoring techniques [2], perioperative hypotension may be the most important modifiable risk factor associated with postoperative myocardial injury, acute kidney injury (AKI), and mortality. Large cohort studies across diverse noncardiac surgical populations consistently demonstrate graded associations between the duration and depth of intraoperative hypotension (often defined as mean arterial pressure [MAP] < 60–70 mmHg), thresholds that have informed contemporary consensus recommendations [3], and adverse postoperative outcomes [4–11]. However, observational analyses demonstrate associations which may or may not be causal.

The more important question is whether preventing hypotension improves meaningful outcomes. With just a single exception (N < 300) [12], trials that together randomized about 13,000 patients to routine care versus various higher intraoperative blood pressure targets report neutral results [13–17]. This narrative review synthesizes observational and trial evidence guiding perioperative blood pressure management and explores future direction for individualized hemodynamic management, including emerging predictive monitoring and automated systems. Although perioperative blood pressure management spans the preoperative, intraoperative, and postoperative periods, we focus on intraoperative hypotension where the majority of mechanistic data and randomized evidence are available. Postoperative hypotension is discussed separately given its distinct pathophysiology and management constraints; preoperative antihypertensive management is beyond our scope.

Physiological rationale of blood pressure control during anesthesia

Organ perfusion depends on the balance between systemic arterial pressure, venous or downstream pressure, organ-specific vascular resistance, and the integrity of autoregulatory mechanisms that maintain local blood flow across varying perfusion pressures [18, 19]. MAP is a practical surrogate for organ perfusion pressure, but the relationship between MAP and organ blood flow is neither linear nor uniform across organs or patient populations.

Autoregulation of organ blood flow

Vital organs, including the brain, heart, and kidneys, normally maintain relatively constant blood flow across a range of perfusion pressures through autoregulation. Within this autoregulatory plateau, changes in systemic arterial pressure are buffered by adjustments in vascular resistance, preserving stable tissue perfusion. Below the lower limit of autoregulation, blood flow becomes pressure-dependent, rendering organs vulnerable to hypoperfusion and ischemic injury. In most adults, cerebral autoregulation is maintained for MAPs between 60 and 150 mmHg, with renal and myocardial autoregulation typically occurring within slightly narrower, ranges. However, autoregulation limits vary among individuals and may be influenced by chronic disease, anesthetic drugs, and physiological perturbations. In practice, the lower limit of autoregulation is usually assumed rather than directly measured which makes population-based MAP thresholds attractive but oversimplified.

Effects of anesthesia on hemodynamics

Intravenous and volatile anesthetics produce dose-dependent vasodilation and myocardial depression which reduces systemic vascular resistance, and consequently MAP, typically without much changes in heart rate, stroke volume index, or cardiac index [20]. Anesthetic-induced vasodilation and myocardial depression are compounded by blunted autonomic responses which attenuate compensatory tachycardia or vasoconstriction. Hypotension can be amplified by surgical positioning, including reverse Trendelenburg and abdominal insufflation during laparoscopic procedures. Older patients, those with autonomic dysfunction, and those with chronic arterial hypertension may be particularly susceptible. Neuraxial techniques can cause abrupt sympathectomy and profound decreases in systemic vascular resistance. Combined, these factors make the perioperative period prone to transient but substantial hypotension. Fortunately, general anesthesia decreases whole-body metabolic rate by 25–30% which proportionately reduces tissue oxygen and perfusion requirements [21]. Moderate decreases in perfusion pressure during anesthesia therefore do not usually result in inadequate tissue oxygenation. Intraoperative hypotension is thus not a uniform phenomenon but rather reflects distinct hemodynamic mechanisms arising from fundamental cardiovascular physiology. Because MAP depends on cardiac output, heart rate, systemic vascular resistance, and central venous pressure, reductions in MAP may result from isolated or combined impairments in vascular tone, preload, contractility, or chronotropy (Fig. 1).

Fig. 1.

Main mechanisms of intraoperative hypotension

Evidence from observational studies

Unmodifiable preoperative comorbidities and baseline physiological risk factors are often stronger predictors of postoperative organ injury than intraoperative hypotension itself, as shown in large observational cohorts in which advanced age, cardiovascular disease, chronic kidney disease, frailty, and surgical complexity consistently account for a substantial proportion of outcome risk after multivariable adjustment. Over the last decade, analyses of large perioperative datasets conducted in diverse noncardiac surgical patients have consistently demonstrated a graded and time-dependent relationship between intraoperative hypotension and postoperative complications including myocardial injury, acute kidney injury, and death [4–11]. In at-risk patients, even short periods of modest hypotension measurably increase the risk of organ injury [11, 22]. Conversely, patients without major baseline comorbidities generally have a lower absolute risk of postoperative complications, even when exposed to intraoperative hypotension [23]. Thus, while intraoperative hypotension is an important modifiable risk factor, its impact must be interpreted within the broader context of underlying patient risk and overall perioperative vulnerability.

Absolute hypotension: universal thresholds

Initial studies identified MAP < 60 mmHg as the population harm thresholds for postoperative organ injury. For example, a cohort analysis of > 33,000 noncardiac surgical patients demonstrated a graded, duration-dependent relationship between intraoperative MAP < 55 mmHg and the risk of AKI and myocardial injury [11]. Even brief episodes of hypotension were associated with increased risk, with odds ratios rising as hypotension persisted. Subsequent studies confirmed these findings, consistently showing that sustained MAP < 60–65 mmHg, a commonly cited threshold, is associated with adverse outcomes across all ages.

Relative hypotension: individualized risk

In an analysis of more than 57,000 patients, hypotension relative to absolute thresholds such as MAP < 65 mmHg and hypotension relative to preoperative clinic blood pressures were each associated with myocardial injury and AKI. Considering either absolute or relative hypotension, longer periods and more extreme levels of hypotension increased risk. However, deviations from baseline blood pressure were no more predictive than area < 65 mmHg [9]. The investigators concluded that either keeping MAP > 65 mmHg or within 30% of baseline will be safe in most patients. In contrast, Wesselink et al., in a systematic review of observational studies, reported that chronically hypertensive patients appeared to experience postoperative complications at higher intraoperative MAPs than normotensive patients, suggesting that a fixed MAP target of 65 mmHg may be insufficient for this group [7]. This conclusion was derived from pooled evidence across multiple cohorts, encompassing approximately 167,000 patients, rather than from a single study.

Postoperative hypotension

Most myocardial injury occurs within the first 48 postoperative hours, and it is likely that much renal injury also accrues postoperatively. Myocardial injury, for example, is associated with postoperative hypotension, even after adjusting for intraoperative hypotension [5, 24]. The challenge is that intermittent ward vital sign assessments, say at 4-h intervals, miss much hypotension and hypoxemia, because perturbations occurs between assessments [25–28]. Continuous ward monitoring systems are available and will probably soon become routine [29]. Continuous ward monitoring may allow hospitals to move beyond “failure to rescue” (waiting for complications and then trying to save patients) to anticipating and preventing complications.

Unlike the intraoperative period, where hypotension is usually drug-induced, postoperative hypotension may reflect hypovolemia from bleeding or fluid shifts, sepsis, myocardial dysfunction, pulmonary embolism, arrhythmias, systemic inflammation or analgesic-related vasodilation. A recent prospective observational study of patients recovering from noncardiac surgery described the incidence, timing, and clinical context of new-onset postoperative hypotension on the surgical ward [30]. Hypotension was frequently sustained and often occurred in patients who subsequently developed postoperative complications, highlighting the clinical relevance of postoperative blood pressure monitoring. However, management of ward hypotension is constrained by practical and structural limitations that generally preclude invasive arterial monitoring, vasopressor infusions, and advanced diagnostics. Consequently, ward hypotension is frequently treated empirically with intravenous fluids, even when the underlying cause may not be fluid-responsive.

Limitations of observational data

Despite the consistency of findings across large cohorts, observational studies cannot establish causality and are subject to other limitations. First, confounding due to baseline patient vulnerability remains a major challenge: intraoperative hypotension is more likely to occur in patients who are physiologically fragile or have substantial comorbidities, characteristics that themselves augment the risk of postoperative complications. Second, residual confounding is unavoidable, because not all risk factors and relevant physiological variables—such as cardiac output, microcirculatory perfusion, or autonomic tone—are routinely measured or measured with sufficient accuracy to allow complete statistical adjustment. Third, the limited temporal granularity of blood pressure monitoring, particularly when non-invasive measurements are taken every 3–5 min, can miss brief but potentially important hypotensive episodes, thereby underestimating true exposure. Finally, reverse causation remains possible: for example, hypotension may be a consequence of myocardial ischemia rather than causing it. Nevertheless, the important volume and reproducibility of the data have made MAP 60–65 mmHg a widely accepted pragmatic threshold for intraoperative blood pressure management. The major observational studies examining the association between intraoperative hypotension and postoperative outcomes are reported in Table 1.

Table 1.

Major observational analyses examining the association between intraoperative hypotension and postoperative outcomes

Study (author, journal, year)	Population/N	Definition of IOH	Primary outcome	Effect size (RR/OR/HR)	Key notes
Mascha et al., Anesthesiology 2015 [5]	104,401 adult patients undergoing noncardiac surgery ≥ 60 min	TWA-MAP ARV-MAP SD-MAP	30-day mortality	Mortality > tripled as TWA-MAP decreased from 80 → 50 mmHg; low ARV-MAP OR 1.14 (1.03–1.25); high ARV-MAP OR 0.94 (0.88–0.99)	Lower MAP strongly linked with mortality; BP variability only weakly associated; duration of MAP < 50–70 mmHg also predictive (all P value < 0.001)
Wesselink et al., Br J Anaesth 2018 [6]	785,806 noncardiac surgeries (systematic review of 42 studies)	Various absolute and relative MAP thresholds (MAP < 80, 70, 65, 60, 55, 50 mmHg; durations ≥ 10 min or shorter)	Postoperative adverse outcomes: AKI, MI, stroke, overall organ injury	Reported risks increase with duration and depth of hypotension; prolonged MAP < 80 mmHg ≥ 10 min, MAP < 65–60 mmHg, or any MAP < 55–50 mmHg associated with higher risk	Observational studies only; large heterogeneity in populations, IOH definitions, and outcomes; dose–response relationship between severity/duration of hypotension and organ injury; recommendations for future studies provided
Wijnberge et al., BJS Open 2021 [7]	130,862 noncardiac surgeries (systematic review and meta-analysis of 29 studies)	Varying thresholds (e.g., MAP < 80 mmHg ≥ 10 min, MAP < 65 mmHg, etc.)	Postoperative morbidity and mortality	Morbidity OR 2.08 (95% CI 1.56–2.77); Mortality OR 1.94 (1.32–2.84)	Large meta-analysis; high heterogeneity but consistent direction of harm
Salmasi et al., Anesthesiology 2017 [8]	57,315 noncardiac surgical patients	Relative decrease (> 20–40%) or absolute MAP < 65 mmHg	MI and AKI	MAP < 65 mmHg for > 12 min → OR 1.20 (97.5% CI 1.02–1.40 for AKI, OR 1.34(97.5% CI 1.06–1.68) for MI	Dose–response effect between depth/duration of IOH and organ injury
Gregory et al., Anesth Analg 2021 [9]	368,222 noncardiac surgeries	Absolute MAP thresholds ≤ 75, ≤ 65, ≤ 55 mmHg; also relative thresholds 20–40% drop from baseline	MACCE at 30 days	MAP ≤ 75 →  + 12% (95% CI 11–14); ≤ 65 →  + 17% (15–19); ≤ 55 →  + 26% (22–29)	Large multicenter retrospective cohort; risk increases with severity of hypotension; consistent across all age groups; secondary outcomes included 30- and 90-day mortality, MI, and ischemic stroke

The table summarizes key studies on IOH and postoperative outcomes. Thresholds and duration of hypotension vary; effect sizes are reported

Some patients in Mascha and Salmasi overlap

AKI acute kidney injury, ARV-MAP average real variability of MAP, CI confidence interval, HR hazard ratio, IOH intraoperative hypotension, MACCE major adverse cardiac or cerebrovascular events, MAP mean arterial pressure, MI myocardial injury, OR odds ratio, RR risk ratio, SD-MAP standard deviation of MAP, TWA-MAP time-weighted average MAP

Evidence from randomized trials

Randomized trials provide the most rigorous evidence for determining whether hypotension prevention improves patient outcomes. Available trials evaluated various patient populations, considered different surgical risk categories, considered variable target definitions, and used different intervention strategies. Together, they nonetheless provide critical insights into the extent to which hypotension might cause serious postoperative complications. Table 2 summarizes major randomized trials examining intraoperative hypotension and postoperative outcomes.

Table 2.

Trials on intraoperative arterial blood pressure targets

Study (author, journal, year)	Population/N	Intervention/definition of target	Primary outcome	Effect size (RR/OR/HR)	Key notes
INPRESS, Futier et al., JAMA 2017 [11]	298 high-risk adults undergoing major surgery	Individualized systolic BP target within 10% of baseline vs. standard care (SBP < 80 mmHg or ↓ > 40%)	Composite of postoperative organ dysfunction or SIRS within 7 days	38% vs. 52%; RR: 0.73; (95% CI, 0.56 to 0.94); P = .02	Individualized BP management reduced organ dysfunction; effect mainly from less AKI stage 1 and altered consciousness
Wanner et al., J Am Coll Cardiol 2021 [12]	458 adults with cardiovascular risk factors undergoing noncardiac surgery	MAP ≥ 75 mmHg vs. ≥ 60 mmHg	Composite of major cardiovascular complications (MINS, MACE, AKI, and death) at 7 days	48% vs 52% (ARD 4.2%; 95% CI: − 13% to + 5%)	No benefit of higher MAP; secondary outcomes (AKI, delirium) also neutral; suggested maintain MAP ≥ 60 mmHg
POISE-3, Marcucci et al., Ann Intern Med 2023 [13]	7490 adults undergoing noncardiac surgery	Hypotension-avoidance (MAP ≥ 80 mmHg, continue antihypertensives) vs. hypertension-avoidance (hold antihypertensives, treat high BP)	Composite of major cardiovascular events and AKI at 30 days	13.9%vs 14% (hazard ratio, 0.99 [95% CI, 0.88 to 1.12]; P = 0.92)	Large pragmatic international trial; complex perioperative strategy yielded no clinical benefit
BP-CARES Zhao et al., JACC 2025 [14]	1477 high-risk adults undergoing major abdominal surgery	Intensive MAP ≥ 80 mmHg vs. conventional MAP ≥ 65 mmHg (or ≥ 60% baseline)	Composite of major cardiovascular events at 30 days	14.5% vs. 13.6% (RR 1.07; 95% CI 0.83–1.38; P = 0.61)	Intensive therapy reduced IOH duration (MAP < 65: 1 vs. 8 min) but not outcomes; similar secondary endpoints
IMPROVE-multi, Saugel et al., JAMA 2025 [15]	1142 patients ≥ 45 yrs undergoing major elective abdominal surgery with ≥ 1 high-risk criterion	Individualized target (MAP set to preoperative nighttime MAP) vs. standard target (MAP ≥ 65 mmHg)	Composite of AKI, MI, cardiac arrest, or death at 7 days	33.5% vs. 30.5% (RR 1.10 [95% CI 0.93–1.30]; P = 0.31)*	Adequately powered multicenter RCT; robust MAP separation but neutral outcomes; no difference in 90-day mortality
PRETREAT, Kant et al., JAMA 2025 [16]	3247 adults undergoing elective noncardiac surgery (stratified by IOH risk)	Risk-adapted MAP targets (≥ 70 /≥ 80/≥ 90 mmHg) vs. usual care (avoid MAP < 65 mmHg)	Functional disability at 6 months (WHODAS 2.0)	17.7 (20.1) vs 18.2 (20.5) (mean difference, − 0.5; 95% credible interval, − 1.9 to 0.9)	Stopped early for futility; no difference in 23 secondary outcomes; first large trial with patient-centered endpoint; neutral across risk strata

The table summarizes major randomized trials examining intraoperative hypotension and postoperative outcomes. Definitions of hypotension differ across studies, and effect sizes are reported as published

AKI acute kidney injury, BP blood pressure, CI confidence interval, GI gastrointestinal HR hazard ratio, IOH intraoperative hypotension, MACE major adverse cardiac events, MAP mean arterial pressure, MI myocardial infarction, MINS myocardial injury after noncardiac surgery, OR odds ratio, RCT randomized-controlled trial, RR relative risk, SBP systolic blood pressure, SIRS systemic inflammatory response syndrome, WHODAS 2.0 World Health Organization Disability Assessment Schedule 2.0

The INPRESS trial was a multicenter randomized trial in France that enrolled 298 high-risk patients having major abdominal surgery [12]. Patients were randomized to either an individualized systolic blood pressure target (using continuous norepinephrine infusion to maintain systolic blood pressure within 10% of the baseline resting value) or to routine care, in which ephedrine was administered only when systolic blood pressure fell below 80 mmHg or to less than 40% from the baseline. Norepinephrine was infused continuously in the individualized group to maintain targets, whereas the standard group was given ephedrine boluses. In both groups, additional fluids were given based on a protocolized hemodynamic algorithm to optimize blood flow. The primary outcome was a composite of systemic inflammatory response syndrome and at least one postoperative organ dysfunction within 7 days post-surgery. The primary outcome occurred in 38% of individualized-target patients vs. 51% of standard-care patients (relative risk reduction 27%; P = 0.03). The benefit was largely driven by reductions in stage 1 AKI, altered consciousness, and hypoxemia. Although the trial was small, its positive findings generated enthusiasm for personalized blood pressure management, especially in high-risk populations.

Wanner et al. conducted a randomized trial including 458 adults with cardiovascular risk factors undergoing noncardiac surgery [13]. Participants were assigned to intraoperative MAP targets of either ≥ 60 or ≥ 75 mmHg. The primary outcome was a composite of major cardiovascular complications within 7 days after surgery, including myocardial infarction, stroke, or death. The resultant median cumulative time with MAP < 65 mmHg was 9 min (IQR 3–24) in the MAP ≥ 75 mmHg group versus 23 min (IQR 8–49) in the MAP ≥ 60 mmHg group. The incidence of the primary outcome was 8.5% in the higher-target group and 9.6% in the control group (P = 0.68), indicating that there was no significant or meaningful benefit of targeting a higher MAP. Secondary outcomes, including AKI and postoperative delirium, were also similar between groups. These findings suggest that generally keeping intraoperative MAP above 75 mmHg does not reduce cardiovascular or renal complications in at-risk surgical patients.

POISE-3 was a large pragmatic international RCT that enrolled 7490 adults having noncardiac surgery in more than 100 trial sites [14]. The trial compared a hypotension-avoidance strategy—including an intraoperative target of MAP ≥ 80 mmHg and a perioperative antihypertensive medication algorithm (withholding renin–angiotensin–aldosterone system inhibitors on the day of surgery and for the first two postoperative days, and using other antihypertensives stepwise when systolic blood pressure was elevated)—with a hypertension-avoidance strategy, which targeted MAP ≥ 60 mmHg intraoperatively and generally continued chronic antihypertensive medications perioperatively. Separation in achieved blood pressure was largely confined to the intraoperative period and was modest. Perhaps consequently, there was no difference in the primary composite outcome (major cardiovascular events and acute kidney injury) at 30 days. POISE-3 therefore highlights both the challenges of achieving sustained perioperative blood pressure separation in pragmatic care pathways and the importance of considering perioperative antihypertensive medication management in addition to intraoperative MAP targets.

Individualized guidance was addressed in IMPROVE-multi, a multicenter trial that randomized 1142 patients aged ≥ 45 years having major elective abdominal surgery who had at least one high-risk criterion [16]. Patients were randomized to either individualized care (intraoperative MAP target set to their preoperative nighttime MAP measured by ambulatory blood pressure monitoring) or standard care (MAP ≥ 65 mmHg). Patients in the individualized group had significantly higher intraoperative MAPs and were given more vasopressors. However, there was no difference in the primary composite outcome of AKI, myocardial injury, non-fatal cardiac arrest, or death within the first 7 postoperative days (33.5% vs. 30.5%; RR 1.10 [95% CI 0.93–1.30]; P = 0.31). Secondary outcomes, including 90-day mortality and infectious complications, were also comparable. IMPROVE-multi was adequately powered, employed robust MAP separation, and targeted a physiologically plausible individualized threshold—yet reported no benefit from tight intraoperative pressure control.

PRETREAT was a pragmatic RCT that enrolled 3247 adults having elective noncardiac surgery [17]. Patients were stratified into low, intermediate, or high risk for intraoperative hypotension using a preoperative risk score. The intervention group had intraoperative MAP targets set at ≥ 70 mmHg for low-risk, ≥ 80 mmHg for intermediate-risk, and ≥ 90 mmHg for high-risk patients, with structured vasopressor protocols. The control group received usual care, aiming to avoid MAP < 65 mmHg without higher predefined targets. The primary outcome was functional disability at 6 months, measured by the World Health Organization Disability Assessment Schedule (WHODAS 2.0) that was assessed within clinical routine postoperative follow-up. The trial was stopped early for futility after interim analysis. There was no significant difference in WHODAS scores between proactive and usual care groups (mean difference − 0.5 points [95% CI − 1.9 to 0.9]), nor in any of 23 secondary outcomes. PRETREAT focused on a patient-centered functional outcome rather than surrogate biomarkers, and enrolled a broad, mostly moderate-risk population. Its neutral results further dampened expectations that higher MAP targets alone improve outcomes.

And finally, BP-CARES randomized 1477 high-risk adults having major abdominal surgery to an intensive MAP target of ≥ 80 mmHg or a conventional target of ≥ 65 mmHg (or ≥ 60% of baseline MAP) [15]. Continuous norepinephrine infusion was the primary vasopressor in both groups, and both had standardized goal-directed fluid therapy. Intensive blood pressure management achieved a higher time-weighted average MAP (87 ± 6 mmHg vs. 82 ± 7 mmHg) and markedly reduced the duration and severity of hypotension (median cumulative MAP < 65 mmHg: 1 min vs. 8 min; area under the curve < 65 mmHg: 3 vs 23 mmHg.min⁻¹). Despite substantial exposure differences, the incidence of the primary composite cardiovascular outcome within 30 days (myocardial injury or infarction, new arrhythmia, acute heart failure, stroke, cardiac arrest, or death) was similar between groups (14.5% vs. 13.6%; RR 1.07; 95% CI 0.83–1.38; P = 0.61). All secondary outcomes were also similar in each group. Thus, while intensive vasopressor use (median norepinephrine 0.037 µg kg⁻¹ min⁻¹ vs. 0.026 µg kg⁻¹ min⁻¹) effectively prevented intraoperative hypotension, it did not result in fewer cardiovascular events.

Available trials paint a consistent picture: higher or individualized MAP targets increase MAP and reduce hypotension exposure without improving patient-centered outcomes in a broad range of surgical populations including patients at cardiovascular risk and with baseline hypertension [14–17]. In fact, only one small trial, representing a tiny fraction of all patients randomized in hypotension prevention trials, reported benefit [12]. There is thus now convincing evidence that MAPs ≥ 65 mmHg, consistent with current consensus thresholds, are sufficient for most noncardiac surgical patients.

Several factors may contribute to the neutral results from all recent robust trials. First, intraoperative hypotension may be a marker rather than a mediator of harm. Low MAP often reflects underlying physiological vulnerability—such as impaired autoregulation, reduced cardiac reserve, or vasoplegia—rather than serving as the direct cause of injury. Simply raising MAP may therefore fail to reverse the underlying pathophysiological processes driving complications. Second, postoperative hypotension remains a major blind spot. Most trials focused almost exclusively on intraoperative blood pressure management while neglecting postoperative pressures, a period during which hypotension is common yet often unrecognized owing to intermittent ward monitoring. Third, maintaining higher MAP targets typically increases vasopressor exposure which might impair microcirculatory perfusion, offsetting potential benefit from augmenting perfusion pressure [31]. Fourth, even individualized MAP targets—such as those based on nighttime baseline pressures as in IMPROVE-multi—may fail to represent relevant autoregulatory thresholds which can vary substantially among individuals and fluctuate dynamically within individuals over time. Fifth, most trials lacked advanced hemodynamic monitoring which might help clinicians identify and treat underlying causes of arterial hypotension and possible inadequate organ perfusion. Finally, many trial reported only small differences in the amounts of hypotension. Small differences are unsurprising, since reference patients were all randomized to routine care or something similar, and many simply never because hypotensive. However, a consequence is that without a substantial exposure difference, outcome differences would not be expected.

Importantly, observational and trial results do not conflict. The observational analyses indicate that there are strong associations between MAP < 60–65 mmHg and organ injury. The trials show that targeting pressures well above 65 mmHg does not prevent organ injury. Together, the two types of study strongly suggest that MAP > 65 mmHg is safe in most patients.

How to prevent and correct hypotension

Maintaining an appropriate intraoperative MAP usually requires timely titrated vasopressor therapy integrated with vascular volume optimization and appropriate anesthetic depth, consistent with recent perioperative hemodynamic optimization consensus statements [32]. While many agents raise blood pressure, their hemodynamic profiles differ markedly, influencing tissue perfusion, particularly in patients with coronary disease, ventricular dysfunction, or chronic hypertension. Evidence favors proactive norepinephrine-based infusions over reactive bolus dosing, especially in high-risk surgical patients, reflecting a transition from corrective to anticipatory hemodynamic control [33–35]. The ongoing VEGA-2 trial (NCT 06802224) will further inform this approach by directly comparing different vasopressor strategies during the intraoperative period.

Hypotension prevention begins with early detection. Continuous arterial pressure monitoring, either invasively or noninvasively using advanced finger-cuff or photoplethysmographic technologies, roughly halves the incidence and duration of hypotension during anesthesia induction and maintenance [36–39] An ongoing trial will likely clarify its role in improving perioperative outcomes; for example, the niMON trial is an open-label, multicenter randomized study comparing continuous non-invasive blood pressure monitoring with intermittent monitoring in moderate- to high-risk patients undergoing noncardiac surgery, with postoperative organ failure as the primary outcome [40].

Emerging work on intraoperative hypotension endotypes reinforces the importance of advanced monitoring by showing that hypotension results from distinct mechanisms: vasodilation, hypovolemia, myocardial depression, or bradycardia. Each etiology suggests a different therapeutic response [41, 42]. Rather than simply pursuing a uniform MAP target, clinicians would ideally correct the underlying physiologic disturbance: vasodilation warrants a vasopressor, hypovolemia calls for fluid optimization, myocardial depression requires inotropic agents with or without afterload reduction, and bradycardia responds to positive chronotropic support. Recognizing these patterns may explain why trials that merely escalated MAP without addressing the underlying mechanism report neutral results.

A complementary line of research focuses on refining the concept of perfusion pressure itself. Rather than assuming a uniform MAP target, several investigators have proposed individualized MAP testing, as exemplified in the recent ANDROMEDA-II protocol in which clinicians transiently raise MAP to evaluate its effect on peripheral perfusion—assessed through capillary refill time [43]. Alternative metrics, such as peripheral perfusion index [44], mean perfusion pressure (MAP minus CVP) [45], or critical closing pressure–derived driving pressure [46], are tools that might help characterize perfusion adequacy. However, many tools rely on inputs—such as central venous pressure—that are rarely monitored in noncardiac surgical patients. Similarly, continuous estimation of critical closing pressure or dynamic perfusion indices requires specific monitors or algorithms that are not widely available in standard operating rooms. Consequently, patient-specific perfusion thresholds and individualized targets remain exploratory.

Emerging technologies: predictive analytics and closed-loop vasopressor control

Recent years have seen rapid development of decision-support and automation tools intended to prevent rather than merely react to intraoperative hypotension, representing a paradigm shift in perioperative hemodynamic management. Two complementary technological strategies have emerged: predictive analytics and closed-loop vasopressor control. Predictive analytics-guided management reduces the duration and severity of intraoperative hypotension compared with standard care, but without improving meaningful outcome [47, 48]. However, the studies were small and under-powered for detecting serious complications. Moreover, recent independent analyses and editorials have highlighted important methodological and conceptual limitations, including the difficulty of distinguishing prediction from early detection, uncertainty regarding actionable thresholds, and the risk of protocol-driven overtreatment in the absence of proven causal benefit [49–52].

Closed-loop systems automate vasopressor titration using feedback control algorithms (Fig. 2), continuously adjusting norepinephrine infusion rates to maintain a predefined MAP target. Randomized trials consistently demonstrate that these systems increase the proportion of intraoperative time spent within the target MAP range and reduce both hypotensive and hypertensive excursions compared with manual vasopressor titration [53–55]. To date, however, no randomized trial has evaluated whether these improvements in blood pressure control translate into better patient-centered outcomes. Furthermore, most studies in this field have been conducted by investigators involved in the development of these technologies, and independent replication is limited. As such, while closed-loop vasopressor systems reliably improve blood pressure stability, their clinical benefit beyond physiological control remains unproven and will require large, independently conducted trials powered for hard outcomes.

Fig. 2.

Closed-loop vasopressor representation and impact on blood pressure management compared to manual vasopressor management

Outside of healthcare, predictive algorithms already operate synergistically with autonomous systems to continuously adjust outputs, and examples include self-driving vehicles and industrial process control. Similar principles are now being adapted to medicine. Soon personalized drug titration will rely on intelligent systems that learn from large repositories of prior patient data, identifying patterns among individuals with comparable demographics, comorbidities, and physiological responses. Such systems could anticipate how a specific patient will respond to a given drug or hemodynamic intervention and dynamically optimize dosing within individualized treatment protocols. Predictive analytics represent a genuine step toward data-driven precision medicine where therapy is continuously adapted to the evolving physiologic state rather than being based on fixed algorithms. Furthermore, artificial intelligence and advanced analytics will increasingly be used to model and coordinate the interactions between multiple closed-loop systems. Systems in development include those regulating blood pressure, depth of anesthesia, and fluid balance which will ultimately enable fully integrated hemodynamic management and can potentially improve patient outcome [56].

Conclusion

Perioperative blood pressure management is central to maintaining organ perfusion, and preventing serious hypotension may prevent myocardial and renal injury. Both the magnitude and duration of intraoperative hypotension matter. Observational studies consistently link MAP < 60–70 mmHg to organ injury. In contrast, large randomized trials clearly indicate that targeting higher blood pressure does not improve outcomes across broad populations, including patients with pre-existing hypertension and cardiovascular disease. The observational analyses and trial results do not conflict: observational analyses identify strong associations between MAP < 60–65 mmHg and organ injury, whereas trials show that targeting higher pressures does not prevent organ injury. Together, the two types of study strongly suggest that MAP > 65 mmHg is sufficient in most patients, including those with pre-existing hypertension and known cardiovascular disease.

Proactive strategies, such as continuous norepinephrine infusion, continuous blood pressure monitoring, judicious fluid therapy, and careful control of anesthetic depth reduce the incidence, severity, and duration of hypotension. Whether prophylactic measures improve patient-centered outcomes remains to be established. Until more conclusive data are available, a pragmatic approach is to keep mean arterial pressure ≥ 60–65 mmHg, preferably basing treatment on underlying causes of hypotension. Artificial-intelligence-based-algorithms including predictive analytics and closed-loop systems, may facilitate continuous individualized adjustment of hemodynamic variables.

Data availability

Not applicable.

Declarations

Conflicts of interest

AJ: ownership interest in Perceptive Medical, Newport beach, CA, USA and is/has received research funding from Edwards Lifesciences, Irvine, USA. JLF: received honoraria and speaker’s fees from Edwards Lifesciences (Irvine, California, USA), AOP Orphan Pharmaceuticals France (Nanterre, France), and Masimo International (Irvine, California, USA). DIS: advisor and shareholder, Perceptive Medical (Irvine, CA); advisory board, Masimo (Irvine, CA); consultant, Dynacardia (Cambridge, Massachussets). MSC: received honoraria and speaker’s fees from Edwards Lifesciences (Irvine, California, USA), AOP Orphan Pharmaceuticals France (Nanterre, France), Philips Healthcare (Böbligen, Germany), Laboratoire Agguetant (Lyon, France). EF: received consulting fees from Draeger Medical, consulting fees from GE Healthcare, and lecture fees from Fisher & Paykel Healthcare. MJL: received honoraria and speaker’s fees from Edwards Lifesciences (Irvine, California, USA), AOP Orphan Pharmaceuticals France (Nanterre, France). BS: BS is a consultant for Edwards Lifesciences (Irvine, CA, USA), Philips North America (Cambridge, MA, USA), GE Healthcare (Chicago, IL, USA), Vygon (Aachen, Germany), Masimo (Neuchâtel, Switzerland), Retia Medical (Valhalla, NY, USA), Maquet Critical Care (Solna, Sweden), Pulsion Medical Systems (Feldkirchen, Germany), Dynocardia (Cambridge, MA, USA), RDS (Strasbourg, France). BS has received institutional restricted research grants from Edwards Lifesciences, Baxter (Deerfield, IL, USA), GE Healthcare, Masimo, Philips Medizin Systeme Böblingen (Böblingen, Germany), CNSystems Medizintechnik (Graz, Austria), Pulsion Medical Systems, Vygon, Retia Medical, Osypka Medical (Berlin, Germany). BS has received honoraria for giving lectures from Edwards Lifesciences, Philips Medizin Systeme Böblingen, Baxter, GE Healthcare, Masimo, CNSystems Medizintechnik, Getinge (Gothenburg, Sweden), Pulsion Medical Systems, Vygon, Ratiopharm (Ulm, Germany). BS is an Editor of the British Journal of Anaesthesia. DPV received institutional restricted research grants from Edwards Lifesciences (Irivine California) and Philips Medical B.V (Eindhoven, The Netherlands). MSC is a Section Editor for Intensive Care Medicine. She has not taken part in the review or selection process of this article.

Ethics approval

Non applicable for a review article.