ABSTRACT
Background
There is limited high‐quality evidence to guide haemodynamic and respiratory management during general anaesthesia. The goal of this trial was to test the feasibility of testing different ventilatory settings and blood pressure targets. Previous trials on blood pressure targets have suffered from poor adherence and limited between‐group separation. This manuscript focuses on the overall feasibility, as well as blood pressure targets.
Methods
This was a multicentre, randomised, factorial, clinical trial conducted at eight hospitals in Denmark. 483 patients ASA 3–5, scheduled for major surgery under general anaesthesia, were included. Patients were randomised to a minimum mean arterial blood pressure (MAP) target of 60, 70 or 80 mmHg, or 90% of baseline systolic blood pressure. The primary outcomes focused on feasibility, while secondary outcomes included intra‐ and postoperative complications, hospital length of stay, quality of life and recovery and mortality.
Results
Out of 1690 eligible patients, 483 (29%) were randomised, below the goal of 50%. Separation in blood pressure was achieved and blood pressure during vasoactive treatment was within predefined target ranges. There was no difference in most of the postoperative outcomes. Some complications, including acute kidney injury (OR 4.74, 95% CI 1.17; 31.8) and cardiac arrhythmias (OR 8.13, 95% CI 1.35; 156), were more common in the MAP‐80‐mmHg group compared to the MAP‐60‐mmHg group. The MAP‐80‐mmHg group had fewer days alive and out of hospital within 30 days (median difference −1.88 days, 95% CI −3.50; −0.26).
Conclusions
While not all feasibility goals were achieved, separation in blood pressure and adherence to targets were adequate. Larger trials are needed to determine the optimal blood pressure target during general anaesthesia.
Trial Registration: Clinicaltrials.gov: NCT06047119.
Keywords: blood pressure, clinical trial, feasibility, general anaesthesia, outcomes
1. Introduction
Although general anaesthesia is considered safe, postoperative complications still pose a significant health burden due to the large number of patients undergoing surgery every day [1, 2]. Postoperative complications are more common in high‐risk patients with American Society of Anesthesiologist Physical status (ASA) classification of 3 or higher [3]. Many complications are not clearly related or only partially related to the surgical procedure, suggesting that optimised perioperative care, including haemodynamic and respiratory management during general anaesthesia, could potentially reduce postoperative complications [4].
Our previous meta‐analyses of randomised clinical trials investigating different haemodynamic or respiratory targets during general anaesthesia found no clear evidence to support any specific target of blood pressure, positive end‐expiratory pressure (PEEP), tidal volume or fraction of inspired oxygen (FiO2) [5, 6, 7].
Recently, large, randomised trials on blood pressure targets have been published. However, poor compliance with assigned targets, especially in lower target groups, resulted in achieved mean arterial pressure approaching that of higher target groups. This was evident in the newly published BP‐Cares trial by Zhao et al., in which patients were randomised to a lower blood pressure target (MAP ≥ 65 mmHg or 60% of preoperative blood pressure) or a higher target (MAP ≥ 80 mmHg). Despite this design, the achieved mean map was 82 and 87 mmHg, respectively [8]. Similar tendency was seen in the PRETREAT trial by Kant et al., where patients were randomised to MAP ≥ 65 or risk prediction guided MAP targets (i.e., 70, 80, or 90 mmHg). The median MAP was 81 and 87 mmHg in the respective groups [9]. Adherence was explicitly mentioned as a key limitation in the POISE‐3 by Marcucci et al. This trial investigated strategies to avoid intraoperative hypertension and hypotension with MAP targets of ≥ 60 and ≥ 80 mmHg [10]. Collectively these trials demonstrate that adherence to blood pressure targets was challenging. It therefor remains unclear whether a truly low blood pressure target (60–70 mmHg) were effectively tested, which may explain the observed lack of differences in outcomes in these trials [8, 9, 10].
Multiple treatments and interventions are delivered during each general anaesthetic, suggesting that trial‐designs testing multiple interventions across several domains simultaneously could be possible and increase efficiency. Therefore, the ‘Respiratory and Hemodynamic Targets During General Anaesthesia—A Randomised Clinical Trial’ (‘GA‐Targets’ trial) was designed to assess the feasibility of a factorial design with multiple simultaneous interventions related to mechanical ventilation (PEEP, tidal volume and FiO2) and blood pressure targets during general anaesthesia. To improve adherence to blood pressure targets, enrolment was restricted to high‐risk patients. Prespecified blood pressure management was also specified to ensure titration of vasoactive drugs, when used, to achieve the assigned targets.
The current manuscript focuses on the overall feasibility of conducting a trial with multiple interventions. This manuscript will also provide a description of adherence to blood pressure targets as well as results on postoperative complications from different blood pressure targets. Results from the respiratory interventions, including the feasibility of delivering these, are presented in a separate manuscript [16].
2. Methods
2.1. Trial Design
The GA‐Targets trial was an investigator‐initiated, multicentre, randomised, factorial, single‐blinded trial of different PEEP, tidal volume, FiO2 and blood pressure targets during general anaesthesia. Patients were randomised within these four domains simultaneously, with a variable number of interventional arms, for a total of 72 possible allocation combinations (Figure S1).
An independent data and safety monitoring committee reviewed the trial after inclusion of approximately 80, 160 and 320 patients. There were no predefined stopping criteria for futility, efficacy or safety.
2.2. Ethics
The trial protocol was written by the steering committee and was approved by the Central Denmark regional ethics committee on 28 June 2023 (study number 1‐10‐72‐59‐23, chair Dr. Kronborg). All patients received written and oral trial information before giving written consent in accordance with Danish legislation. The trial was registered on clinicaltrials.gov (NCT06047119) on 14 September 2023.
2.3. Settings and Patients
The trial was conducted at eight hospitals in Denmark (Tables S1 and S2). Patients were eligible for inclusion if they were 18 years of age or older and were classified as ASA 3, 4 or 5. Patients with an expected duration of surgery of at least 90 min, under general anaesthesia with endotracheal intubation and positive pressure ventilation, were included (Figure S1). The types of surgeries included are described in Table S3.
Exclusion criteria included inability to obtain consent, known or suspected pregnancy, out‐patient surgery, patients intubated prior to operating room arrival or clinical judgement by the treating anaesthesiologist that any of the interventions might pose potential harm to the patient.
2.4. Randomisation and Blinding
Randomisation was centralised and conducted online in real‐time immediately before surgery through Research Electronic Data Capture (REDCap) [11, 12] ensuring allocation concealment. Patients were simultaneously randomised to targets of PEEP, tidal volume, FiO2 and blood pressure in a 3 × 3 × 2 × 4 fashion with equal allocation ratios within intervention domains (Figure 1). Patients were randomised in blocks of 72. Randomisation was stratified according to region (Central Denmark Region or other). Given the nature of the interventions, the medical team providing care for the patient during the period of the interventions was not blinded. Patients were not informed of their allocations. The research staff members who collected predefined data from the electronic medical journal were not informed of the allocations.
FIGURE 1.
Consort diagram. Patients were screened before the surgery by an anaesthesiologist or retrospectively by a site investigator. Reasons for not including eligible patients were provided. Lack of resources was the main reason for not including eligible patients. Two patients withdrew consent immediately after randomisation and did not receive interventions; these patients were excluded from analyses. One patient was inadvertently randomised twice, 2 months apart; data from this patient were analysed as two independent cases.
2.5. Intervention
Patients were randomised to interventions within four domains: PEEP of 5, 8 or 10 cm H2O, tidal volume of 6, 8 or 10 mL/kg predicted bodyweight [13], FiO2 of 30% or 80%, and a minimum blood pressure target of mean arterial pressure (MAP) 60, 70 or 80 mmHg or 90% of baseline systolic blood pressure.
The baseline blood pressure was measured in the hospital by medical personnel on the day of the pre‐operative risk evaluation. If multiple blood pressure measurements were available in the medical chart before randomisation, the lowest blood pressure was used. The allocated minimum blood pressure was targeted from administration of anaesthesia until 2 h after admission to the post anaesthesia care unit (PACU) or intensive care unit (ICU) (Figure S2).
To optimise adherence to blood pressure targets and better ensure separation between treatment groups, vasoactive drugs were allowed only when blood pressure fell below the allocated target and, when administered, were titrated to the allocated blood pressure target. The specific interventions used to reach the allocated blood pressure were not specified in the protocol. This also applied to the choice of anaesthetic agents and fluid administration. However, phenylephrine or norepinephrine was recommended as first choice vasopressors.
2.6. Procedures and Data Collection
Patients had an arterial line placed before anaesthesia induction and it remained in place until the end of the intervention period.
All patients had protocolised blood samples drawn before anaesthesia induction, 2 h after PACU/ICU admission, and at postoperative day 1. All blood samples were analysed for C‐reactive protein (CRP), leucocytes, creatinine, aminotransferase levels (ALT), troponin I (TnI) and albumin. A protocolised arterial blood gas was drawn and analysed before anaesthesia induction, every hour during general anaesthesia and 2 h after admission to the PACU/ICU (Figure S2).
Data on PEEP, tidal volume, FiO2 and dosing of vasopressor and inotropes were collected every 15 min from the electronic medical journal. Blood pressure measurements were recorded from the monitor by the treating clinician at 15‐min intervals, according to predefined time points relative to the time of induction.
2.7. Feasibility Outcomes
Our primary outcomes were all related to the feasibility of the trial and included the following feasibility objectives and goals: (1) At least 50% of all eligible patients were randomised. Eligible patients were defined as those meeting all inclusion criteria and no exclusion criteria except those excluded based on clinical judgement. (2) When vasopressors or inotropes were administered, we aimed for 70% of the registered blood pressures to be within ±10 mmHg of the target. (3) We aimed to achieve a minimum 5 mmHg separation in mean MAP for the blood pressure groups targeting MAP. (4) We aimed to collect outcome data on 90% of patients still alive at 30‐ and 90‐days. Feasibility outcomes related to the respiratory interventions are reported in a separate manuscript.
2.8. Clinical Outcomes
Clinical outcomes included intraoperative complications, postoperative complications, ICU admission, length of ICU stay, hospital length of stay and days alive and out of hospital within 30 days of surgery, health‐related quality of life and quality of recovery at 30 and 90 days and mortality at 30 and 90 days.
Intraoperative complications included intraoperative pneumothorax, cardiac arrest, arrhythmias, bleeding and need for blood transfusions.
Postoperative complications included a variety of organ‐specific complications, postoperative pain and nausea and vomiting (PONV).
Definitions for all outcomes are provided in the protocol (Supporting Information ‘Protocol GA‐Targets Trial’).
2.9. Statistical Analyses
Our sample size of 480 patients was based on a combination of expected eligible patients, expected feasibility of including all patients within an approximate 1‐year timeframe, and providing enough power to detect differences in blood pressure between groups. A 5‐mmHg difference in mean MAP between groups, with a common standard deviation of 10 mmHg, and 120 patients in each group (i.e., a total of 480 patients) would provide > 90% power to detect a statistically significant difference using an alpha of 0.05 with a t‐test.
We used descriptive statistics for baseline characteristics and the outcomes describing feasibility. Post hoc subgroup analyses were performed to assess feasibility based on trial sites and only considering elective patients.
Means of intraoperative blood gas values, as well as intra‐ and postoperative blood pressures within each blood pressure group, were estimated using generalised linear models with generalised estimating equations with an autoregressive correlation structure. Mean differences between groups with 95% confidence intervals (95% CI) were also estimated from these models. Each blood pressure group was compared to the reference group, MAP 60 mmHg.
During induction there is a high variability in blood pressure. Therefore, induction blood pressure was analysed separately from the remaining blood pressures. We chose to report the lowest blood pressure measured during induction. For the remaining intraoperative blood pressures, we used all blood pressures reported from 15 min after induction until end of general anaesthesia, with the primary analysis only including the first 180 min.
Binary outcomes are presented as counts and percentages, and between‐group differences are presented as odds ratios (OR) with 95% CIs using a logistic regression model. Continuous outcomes are presented as medians with interquartile ranges, and the between‐group differences are presented as geometric mean ratios (GMR) with 95% CIs or median differences with 95% CIs depending on the distribution of the data.
All models included blood pressure, PEEP, tidal volume and FiO2 interventions, as well as pre‐defined strong prognostic factors including ASA classification, age, and urgency of surgery (elective vs. acute), and the stratification variable. For analyses related to biomarkers, we also adjusted for preoperative values.
When outcome data were missing in less than 5% of patients, we performed complete case analyses. In cases where data were missing in more than 5% of patients, multiple imputation was performed (Section S1).
3. Results
3.1. Feasibility and Baseline Characteristics
From 9 October 2023 to 24 January 2025, 1752 patients were screened with 1690 considered eligible. Of these, 485 (29%) were randomised and allocated to the interventions, but two patients withdrew consent after randomisation but prior to anaesthesia induction (Figure 1). The proportion of included patients varied from 6% to 65% across sites (Table S2). Of the 1690 eligible patients, 1090 patients were not included due to either clinical or logistical reasons. Clinical reasons included concerns about applying the interventions in patients with significant comorbidities or patients undergoing spine or shoulder surgery in the prone or beach chair position (Figure 1; Table S4).
When only considering elective cases, 445 of 1354 eligible patients (33%) were included and this proportion varied from 8% to 64% across sites (Table S2).
Baseline characteristics are provided in Table 1. Surgical characteristics, reason for hospital admission, preoperative vital signs and preoperative biomarkers are provided in the supplement material (Tables S3 and S5–S8).
TABLE 1.
Baseline patient and surgical characteristics.
| Characteristics | Overall, N = 483 | MAP targets | 90% of systolic blood pressure, N = 120 | ||
|---|---|---|---|---|---|
| 60 mmHg, N = 119 | 70 mmHg, N = 122 | 80 mmHg, N = 122 | |||
| Age, years | 73 (66; 78) | 72 (67; 78) | 73 (64; 80) | 72 (62; 77) | 74 (68; 79) |
| Male sex | 294 (61%) | 64 (54%) | 72 (59%) | 76 (62%) | 82 (68%) |
| Body mass index, kg/m2 | 27 (24; 31) | 27 (24; 31) | 27 (24; 31) | 27 (24; 31) | 28 (24; 31) |
| Systolic blood pressure, mmHg | 133 (122; 147) | 139 (124; 154) | 134 (122; 146) | 130 (123; 146) | 130 (119; 140) |
| Diastolic blood pressure, mmHg | 77 (69; 85) | 78 (70; 85) | 75 (67; 84) | 80 (70; 86) | 75 (67; 82) |
| ASA classification a | |||||
| 3 | 456 (94%) | 116 (97%) | 110 (90%) | 118 (97%) | 112 (93%) |
| 4 | 27 (5.6%) | 3 (2.5%) | 12 (9.8%) | 4 (3.3%) | 8 (6.7%) |
| 5 | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| Frailty | |||||
| Very fit | 76 (16%) | 18 (15%) | 22 (18%) | 20 (16%) | 16 (13%) |
| Well/managing well | 130 (27%) | 35 (29%) | 28 (23%) | 30 (25%) | 37 (31%) |
| Vulnerable | 160 (33%) | 38 (32%) | 37 (30%) | 47 (39%) | 38 (32%) |
| Mild/moderate frailty | 113 (23%) | 27 (23%) | 33 (27%) | 24 (20%) | 29 (24%) |
| Severe frailty/terminally ill | 4 (0.8%) | 1 (0.8%) | 2 (1.6%) | 1 (0.8%) | 0 (0%) |
| Comorbidities | |||||
| Coronary artery disease | 116 (24%) | 22 (18%) | 44 (36%) | 22 (18%) | 28 (23%) |
| Chronic heart failure | 64 (13%) | 11 (9.2%) | 20 (16%) | 17 (14%) | 16 (13%) |
| Ejection fraction, % | 45 (40; 55) | 45 (45; 50) | 45 (40; 50) | 45 (45; 55) | 50 (45; 55) |
| Atrial fibrillation | 96 (20%) | 20 (17%) | 22 (18%) | 30 (25%) | 24 (20%) |
| Stroke | 75 (16%) | 21 (18%) | 19 (16%) | 25 (20%) | 10 (8.3%) |
| Venous thromboembolism | 33 (6.8%) | 8 (6.7%) | 12 (9.8%) | 6 (4.9%) | 7 (5.8%) |
| Hypertension | 352 (73%) | 91 (76%) | 92 (75%) | 85 (70%) | 84 (70%) |
| ≥ 3 antihypertensive drugs | 87 (25%) | 21 (23%) | 26 (28%) | 19 (22%) | 21 (25%) |
| Diabetes | 145 (30%) | 33 (28%) | 30 (25%) | 41 (34%) | 41 (34%) |
| Insulin dependent | 60 (41%) | 9 (27%) | 16 (53%) | 20 (49%) | 15 (37%) |
| Pulmonary disease | 121 (25%) | 26 (22%) | 27 (22%) | 38 (31%) | 30 (25%) |
| FEV1/FVC ratio | 0.61 (0.52; 0.71) | 0.66 (0.57; 0.70) | 0.62 (0.52; 0.73) | 0.61 (0.50; 0.71) | 0.60 (0.52; 0.72) |
| Renal disease | 60 (12%) | 10 (8.4%) | 16 (13%) | 13 (11%) | 21 (18%) |
| Liver disease | 11 (2.3%) | 2 (1.7%) | 2 (1.6%) | 2 (1.6%) | 5 (4.2%) |
| Cancer | 156 (32%) | 34 (29%) | 33 (27%) | 38 (31%) | 51 (43%) |
| Chronic pain | 65 (13%) | 23 (19%) | 21 (17%) | 14 (11%) | 7 (5.8%) |
| Type of surgery | |||||
| Acute | 40 (8.3%) | 8 (6.7%) | 7 (5.7%) | 10 (8.2%) | 15 (13%) |
| Abdominal/urological/gynaecological | 299 (62%) | 74 (62%) | 70 (57%) | 78 (64%) | 77 (64%) |
| Laparoscopic | 224 (75%) | 53 (72%) | 56 (80%) | 58 (74%) | 57 (74%) |
| Open | 75 (25%) | 21 (28%) | 14 (20%) | 20 (26%) | 20 (26%) |
| Orthopaedic/spine | 68 (14%) | 21 (18%) | 18 (15%) | 11 (9.0%) | 18 (15%) |
| Vascular | 111 (23%) | 22 (18%) | 32 (26%) | 33 (27%) | 24 (20%) |
| Ear, nose, and throat | 5 (1.0%) | 2 (1.7%) | 2 (1.6%) | 0 (0%) | 1 (0.8%) |
Note: Binary outcomes: no. (%). Continuous outcomes: medians (Q1, Q3). Reason for hospital admission (Table S5), preoperative vital signs (Table S6), preoperative laboratory variables (Tables S21 and S22) and additional surgical characteristics (Table S3) are provided in the Supporting Information.
Abbreviations: FEV1/FVC, forced expiratory volume in 1 s divided by forced vital capacity; MAP, mean arterial pressure.
Two patients were classified as ASA 2 in the electronic medical journal but classified as ASA 3 by the treating anaesthesiologist.
There were separations in blood pressure between the groups with a corresponding increase in total dose of vasoactive drugs with higher MAP targets during general anaesthesia (Figure 2; Figures S3 and S4; Table 2; Table S9). We saw no separation in mean MAP in the postoperative intervention period (Table S10). When vasoactive drugs were used, 71% of blood pressures were within 10 mmHg of the target during general anaesthesia (Table 2; Table S11). Few deliberate deviations from the assigned interventions were reported (Table S12).
FIGURE 2.
Mean arterial blood pressure according to allocation group. MAP, mean arterial pressure; SBP, systolic blood pressure. Mean MAP and 95% CI from induction (0 min) to 180 min after during general anaesthesia for each blood pressure target. Dotted lines and right‐side error‐bars represent the overall group mean and 95% CI across the 15–180 min of general anaesthesia, with numeral values provided below the graph. Systolic and diastolic blood pressures over time (0–180 min) are provided in Figures S3 and S4. Additional details are provided in Table 2 and Table S9.
TABLE 2.
Compliance to blood pressure target.
| MAP targets | 90% of systolic blood pressure, n = 120 | |||
|---|---|---|---|---|
| 60 mmHg, n = 119 Ref. | 70 mmHg, n = 122 | 80 mmHg, n = 122 | ||
| Mean MAP 0–180 min (95% CI), mmHg | 70 (69; 72) | 76 (75; 77) | 83 (82; 84) | 79 (78; 81) |
| Mean difference (95% CI) | 6 (4; 8) | 12 (11; 14) | 9 (7; 12) | |
| Lowest mean MAP during induction (95% CI), mmHg | 71 (67; 74) | 71 (68; 75) | 71 (68; 74) | 71 (68; 75) |
| Mean difference (95% CI) | 1 (−4; 5) | 0 (−4; 5) | 1 (−4; 5) | |
| Vasoactive drug use and fluid administration a | ||||
| Vasoactive drug used—no (%) | 112 (94) | 118 (97) | 119 (98) | 115 (96) |
| OR (95% CI) | 1.78 (0.51; 7.08) | 2.60 (0.69; 12.5) | 1.35 (0.41; 4.82) | |
| Median total vasoactive drug dose (Q1; Q3), μg b | 179 (49; 720) | 472 (168; 1090) | 688 (318; 1700) | 496 (205; 1325) |
| Median difference (95% CI) | 161 (25; 298) | 403 (179; 626) | 234 (67; 402) | |
| Blood pressures ±10 mmHg of target during vasoactive drug use, % c | 69% | 73% | 80% | 61% |
| Median total intraoperative IV fluids (Q1; Q3), mL | 1300 (950; 1800) | 1271 (1000; 1885) | 1550 (1072; 2310) | 1483 (1100; 2101) |
| Median difference (95% CI) | −29 (−182; 125) | 233 (10; 457) | 155 (−29; 338) | |
Abbreviation: MAP, mean arterial pressure.
Vasoactive drug use and fluid administration were missing in 2, 1, 2 and 2 patients in the MAP‐60‐, 70‐ and 80‐mmHg, and 90%‐of‐systolic‐blood‐pressure groups, respectively.
Vasoactive drug dose was missing in 3, 1, 3 and 3 patients in the MAP‐60‐, 70‐ and 80‐mmHg, and 90%‐of‐systolic‐blood‐pressure groups, respectively. Vasoactive drug dose is reported in norepinephrine equivalent dose: 0.01 × dopamine (μg), 0.06 × phenylephrine (μg), 1 × epinephrine (μg), and 1 × ephedrine (mg).
Additional details are provided in Table S11.
For postoperative outcomes, except those involving biomarkers and patient questionnaires, there was no missing data. 30‐ and 90‐day quality of life and quality of recovery follow‐up was completed in 380 (78%) and 387 (80%) patients, respectively (Figure 1; Tables S13 and S14).
3.2. Clinical Outcomes
Compared to the MAP‐60‐mmHg group, the MAP‐80‐mmHg group had a longer duration of surgery and anaesthesia and more frequent reporting of intraoperative bleeding (OR 1.88, 95% CI 1.05; 3.42) (Table 3; Table S15). No other differences in perioperative management or in the remaining perioperative complications, including intra‐ and postoperative arterial blood gas parameters, were observed for any blood pressure group compared to the MAP‐60‐mmHg group (Table S15–S17).
TABLE 3.
Intraoperative and postoperative complications.
| MAP targets | 90% of systolic blood pressure target, n = 120 | |||
|---|---|---|---|---|
| 60 mmHg, n = 119 Ref. | 70 mmHg, n = 122 | 80 mmHg, n = 122 | ||
| Intraoperative complications | ||||
| Any blood loss | 79 (67%) | 87 (71%), 1.24 (0.71; 2.18) | 96 (79%), 1.88 (1.05; 3.42) | 90 (76%), 1.65 (0.92; 2.98) |
| Estimated mL lost | 230 (100; 620) | 200 (50; 700), 0.91 (0.60; 1.38) | 275 (100; 798), 1.33 (0.89; 1.98) | 300 (100; 600), 1.27 (0.84; 1.91) |
| Intraoperative blood transfusion | 8 (6.7%) | 14 (11%), 1.72 (0.67; 4.64) | 14 (11%), 1.86 (0.74; 4.99) | 11 (9.2%), 1.22 (0.45; 3.38) |
| mL transfused | 325 (271; 791) | 600 (300; 1000), 1.18 (0.58; 2.38) | 475 (276; 1200), 1.31 (0.67; 2.55) | 900 (350; 1100), 1.31 (0.63; 2.72) |
| New onset intraoperative arrhythmia | 3 (2.5%) | 5 (4.1%), 1.30 (0.29; 6.73) | 5 (4.1%), 1.67 (0.40; 8.38) | 3 (2.5%), 0.78 (0.14; 4.39) |
| Postoperative complications | ||||
| 30‐day mortality | 1 (0.8%) | 5 (4.1%), 5.93 (0.82; 126) | 3 (2.5%), 3.57 (0.42; 77.7) | 3 (2.5%), 2.31 (0.26; 50.1) |
| 90‐day mortality | 4 (3.4%) | 6 (4.9%), 1.62 (0.41; 7.15) | 5 (4.1%), 1.23 (0.29; 5.58) | 5 (4.2%), 0.77 (0.17; 3.53) |
| Acute kidney injury | 2 (1.7%) | 2 (1.6%), 0.87 (0.10; 7.50) | 9 (7.4%), 4.74 (1.17; 31.8) | 6 (5.0%), 2.89 (0.63; 20.4) |
| Major adverse cardiac events | 6 (5.0%) | 5 (4.1%), 0.78 (0.22; 2.73) | 14 (11%), 2.42 (0.92; 7.19) | 9 (7.5%), 1.28 (0.43; 4.06) |
| Myocardial injury after non‐cardiac surgery | 7 (6.2%) | 15 (12%), 1.86 (0.73; 5.14) | 12 (10%), 1.76 (0.68; 4.94) | 12 (10%), 1.58 (0.60; 4.46) |
| Delirium | 5 (4.2%) | 7 (5.7%), 1.25 (0.37; 4.55) | 7 (5.7%), 1.32 (0.39; 4.72) | 3 (2.5%), 0.36 (0.06; 1.63) |
| Emergence delirium within 2 h | 5 (4.2%) | 6 (4.9%), 1.05 (0.28; 4.04) | 5 (4.1%), 0.93 (0.24; 3.64) | 6 (5.0%), 0.82 (0.22; 3.11) |
| Clavien Dindo grading of complications | ||||
| No deviation from normal postoperative course | 66 (55%) | 64 (52%), 0.85 (0.51; 1.43) | 51 (42%), 0.58 (0.34; 0.97) | 54 (45%), 0.66 (0.39; 1.11) |
| Grade I | 14 (12%) | 18 (15%), 1.29 (0.61; 2.81) | 13 (11%), 0.88 (0.39; 1.99) | 16 (13%), 1.16 (0.53; 2.55) |
| Grade II | 16 (13%) | 18 (15%), 1.13 (0.54; 2.38) | 24 (20%), 1.58 (0.79; 3.21) | 21 (18%), 1.34 (0.66; 2.78) |
| Grade III (IIIa + IIIb) | 21 (18%) | 16 (13%), 0.74 (0.36; 1.50) | 26 (21%), 1.23 (0.65; 2.37) | 23 (19%), 1.15 (0.59; 2.26) |
| Grade VI (VIa + Vb) | 1 (0.8%) | 1 (0.8%), 1.12 (0.04; 28.8) | 5 (4.1%), 5.13 (0.79; 100) | 3 (2.5%), 3.77 (0.45; 78.9) |
| Grade V | 1 (0.8%) | 5 (4.1%), 5.93 (0.82; 126) | 3 (2.5%), 3.57 (0.42; 77.7) | 3 (2.5%), 2.31 (0.26; 50.1) |
| Blood transfusion within 72 h | 15 (13%) | 15 (12%), 0.93 (0.42; 2.07) | 21 (17%), 1.47 (0.71; 3.12) | 23 (19%), 1.54 (0.75; 3.25) |
| ml transfused within 72 h | 554 (273; 1837) | 640 (543; 1093), 1.40 (0.71; 2.77) | 561 (483; 1092), 1.11 (0.63; 1.93) | 550 (299; 1094), 1.01 (0.57; 1.79) |
| Length of hospital stay (days) | 5 (2; 10) | 5 (2; 8), 1.01 (0.80; 1.28) | 6 (3; 13), 1.22 (0.97; 1.54) | 6 (3; 9), 1.06 (0.84; 1.35) |
| Days alive and out of hospital | 26 (21; 28) | 25 (22; 28), −0.88 (−2.22; 0.45) | 24 (17; 27), −1.88 (−3.50; −0.26) | 25 (21; 27), −1.05 (−2.20; 0.10) |
| PACU admission | 115 (97%) | 114 (93%), 0.59 (0.15; 2.04) | 114 (93%), 0.49 (0.12; 1.63) | 115 (96%), 1.08 (0.25; 4.54) |
| Length of PACU stay (hours) | 3.2 (2.6; 4.3) | 3.4 (2.7; 5.3), 1.06 (0.89; 1.27) | 3.6 (2.8; 5.7), 1.25 (1.05; 1.49) | 3.3 (2.8; 4.0), 1.01 (0.85; 1.20) |
| ICU admission | 15 (13%) | 13 (11%), 0.80 (0.35; 1.80) | 19 (16%), 1.27 (0.61; 2.73) | 13 (11%), 0.68 (0.29; 1.55) |
| Length of ICU stay (hours) | 41 (25; 94) | 46 (17; 92), 0.91 (0.33; 2.53) | 27 (13; 94), 0.63 (0.24; 1.68) | 48 (48; 62), 0.89 (0.32; 2.47) |
| Biomarkers | ||||
| Creatinine 2 h (μmol/L) | 71 (58; 88) | 78 (59; 100), 0.99 (0.95; 1.03) | 74 (61; 91), 1.02 (0.98; 1.06) | 75 (63; 99), 1.00 (0.97; 1.05) |
| Creatinine 24 h (μmol/L) | 75 (64; 92) | 80 (65; 101), 0.98 (0.93; 1.03) | 78 (64; 96), 1.01 (0.96; 1.06) | 82 (70; 108), 1.02 (0.97; 1.08) |
| TnI 2 h (ng/L) | 5 (3; 10) | 7 (4; 12), 1.04 (0.88; 1.22) | 7 (3; 11), 1.09 (0.92; 1.29) | 6 (3; 14), 1.16 (0.98; 1.37) |
| TnI 24 h (ng/L) | 8 (3; 12) | 8 (4; 16), 1.11 (0.86; 1.44) | 8 (4; 14), 1.00 (0.78; 1.29) | 8 (3; 15), 1.08 (0.84; 1.38) |
| ALT 2 h (U/L) | 21 (12; 36) | 22 (15; 39), 1.08 (0.92; 1.27) | 20 (12; 34), 1.01 (0.86; 1.19) | 17 (11; 30), 0.97 (0.82; 1.15) |
| ALT 24 h (U/L) | 21 (14; 33) | 22 (15; 35), 1.12 (0.95; 1.32) | 21 (13; 34), 1.05 (0.89; 1.23) | 18 (11; 30), 0.98 (0.82; 1.16) |
| Quality of life and quality of recovery | ||||
| QOL index 30 days | 0.85 (0.66; 0.93) | 0.86 (0.62; 0.92), 0.01 (−0.05; 0.08) | 0.83 (0.54; 0.95), −0.02 (−0.10; 0.06) | 0.83 (0.73; 0.95), −0.00 (−0.07; 0.07) |
| EQ VAS 30 days | 65 (50; 80) | 60 (50; 80), −3 (−15; 9) | 65 (50; 80), 1 (−10; 12) | 70 (50; 80), 2 (−9; 14) |
| QOR 30 days | 129 (107; 139) | 130 (107; 141), 1 (−6; 8) | 126 (106; 142), −1 (−8; 6) | 128 (111; 140), −1 (−8; 6) |
| QOL index 90 days | 0.85 (0.68; 0.95) | 0.88 (0.65; 0.96), 0.02 (−0.04; 0.08) | 0.90 (0.72; 0.97), 0.04 (−0.01; 0.09) | 0.87 (0.74; 0.95), 0.01 (−0.04; 0.07) |
| EQ VAS 90 days | 65 (50; 75) | 70 (50; 85), 5 (−5; 14) | 68 (50; 85), 2 (−8; 13) | 68 (50; 80), 1 (−8; 10) |
| QOR 90 days | 130 (117; 142) | 134 (113; 143), 1 (−6; 9) | 132 (111; 144), 1 (−7; 9) | 132 (117; 143), −1 (−8; 6) |
Note: Binary outcomes: no. (%). Odds ratio (95% CI). Days alive and out of hospital and quality of life and quality of recovery outcomes: Median difference (95% CI). The remaining continuous outcomes: medians (Q1, Q3). Multiplicative effects (geometric mean ratio) from log‐transformed models (95% CI). All biomarkers, quality of life and quality of recovery effect‐estimates (95% CI) are calculated from imputed data. Results from complete‐case data are provided in Tables S13 and S14. QOL index (EQ‐5D‐5L index) values are derived using the Danish value set and range from −0.757 (health states worse than death) to 1 (perfect health) [14]. EQ VAS (EuroQol visual analogue scale): A scale ranging from 0 (worst imaginable health) to 100 (best imaginable health) rated by the patient. QoR‐15 (Quality of Recovery‐15) scores range from 0 to 150, with higher scores indicating better postoperative recovery [15]. Missing data: Blood loss data were missing in one and two patients in the MAP‐60‐mmHg and 90%‐of‐systolic‐blood‐pressure groups, respectively. Intraoperative arrhythmia data were missing in one patient in the 90%‐of‐systolic‐blood‐pressure group. Outcomes: Intraoperative complication: No events of pneumothorax or cardiac arrest were observed. Acute kidney injury includes KDIGO stages 2 and 3. All KDIGO stages are reported in Table S26. Major adverse cardiac events, pulmonary complications and surgical site infections are reported in Tables S18–S20. ICU admissions are specified in Table S21. Pain and postoperative nausea and vomiting are reported in Tables S22–S24. Additional biomarker analyses are reported in Table S25.
Abbreviations: ALT, alanine aminotransferase; ICU, intensive care unit; KDIGO, Kidney Disease: Improving Global Outcomes guidelines; PACU, post anaesthesia care unit; TnI, troponin I.
For most postoperative outcomes including mortality, major adverse cardiac events, surgical site infection, pulmonary complications, delirium, intensive care admission, pain, PONV, blood transfusion, biomarker levels, quality of life and quality of recovery outcomes, no differences were observed between groups (Table 3; Tables S13, S14, S18–S25).
Patients in the MAP‐80‐mmHg group had lower odds of an uncomplicated postoperative course, compared to the MAP‐60‐mmHg group (OR 0.58, 95% CI 0.34; 0.97). Certain complications were more frequent in the MAP‐80‐mmHg group including severe acute kidney injury (OR 4.74, 95% CI 1.17; 31.8) and new onset of cardiac arrhythmias, all classified as atrial fibrillation/flutter (OR 8.13, 95% CI 1.35; 156), accompanied by fewer days alive and out of hospital (median difference −1.88 days, 95% CI −3.50; −0.26) (Table 3; Tables S18 and S26).
No differences in any clinical outcomes were observed for the MAP‐70‐mmHg or the individualised‐blood‐pressure group compared to the MAP‐60‐mmHg group.
3.3. Results on the Mechanical Ventilation Domains
Results are presented elsewhere [16]. In summary, adherence to the mechanical ventilation settings was adequate. No difference was observed for most outcomes within each mechanical intervention domain.
4. Discussion
In this trial, we primarily investigated the feasibility of conducting a randomised factorial trial comparing different minimum targets of blood pressure, PEEP, tidal volume and FiO2 during general anaesthesia in adult high‐risk patients. Although 483 patients were included from eight sites over a 16‐month period with a clear separation in blood pressure between groups, some of the pre‐defined feasibility outcomes were not met. There was no difference in most clinical outcomes between groups, but some outcomes were worse in the MAP‐80‐mmHg group.
The trial investigated both the feasibility of patient inclusion and compliance with the allocated targets. Only 29% of eligible patients completed randomisation, which is below the pre‐defined goal of 50%. The inclusion rate varied widely between sites, with some sites enrolling more than 50% of eligible patients. We attribute this primarily to differences in the prioritisation of research within routine clinical practice. At the highest‐performing sites, we observed clear support for research from departmental leadership, along with strong engagement from a small number of clinicians who carried a substantial share of the workload. As the trial was embedded in usual clinical care, lack of resources was the main reason for not including both acute and elective patients. The trial setup was complex, including both individual patient information and consent, and multiple blood draws, which might explain the low proportion of included patients.
Clear separations between blood pressure groups were observed, but similar to previous trials on intraoperative blood pressure management, the mean MAP was higher than the allocated minimum targets, particularly for the 60 mmHg target [10, 17]. Minimum blood pressure targets were used and treatment to reduce blood pressure was not recommended. Therefore, blood pressure above the allocated target was expected. While vasoactive drugs were administered, 71% of measured blood pressures were within 10 mmHg of the allocated target, which met our pre‐defined feasibility goal.
Compared with recent trials on intraoperative blood pressure targets, adherence to the assigned targets was better, and meaningful separation was achieved between the lower target group (MAP ≥ 60) and higher targets (MAP ≥ 80 mmHg or ≥ 90 of systolic blood pressure). In the GA‐Targets trials, only high‐risk patients (ASA 3–5) were eligible. This pragmatic strategy may have contributed to improved adherence to targets; however further restriction in patient selection could improve the feasibility of achieving blood pressure targets and separation. Previous trials have used restrictive inclusion criteria without these trials succeeding in achieving adequate adherence, indicating that this cannot be the sole strategy [8, 10]. Clearly, protocolised blood pressure management is necessary, and avoidance of unnecessary vasoactive treatment above the minimum blood pressure target should be emphasised in future trials.
This trial found no difference in most postoperative outcomes including mortality between groups. This is in concordance with results from recent trials and a systematic review on trials investigating different blood pressure targets during general anaesthesia [5, 8, 9, 18, 19]. The MAP‐80‐mmHg group had a higher risk of some intra‐ and postoperative complications including any intraoperative blood loss, postoperative cardiac arrhythmias and severe acute kidney injury (KDIGO stages 2 and 3).
Increased risk of severe acute kidney injury in the group targeting MAP‐80‐mmHg is in contrast with previous trials finding no difference [20, 21, 22]. Differences in trial populations, reporting of acute kidney injury within 30 days instead of within 7 days [21, 23], and lack of intervention compliance in previous trials could explain this discrepancy [8, 9, 19, 22]. An increased risk of intraoperative bleeding was evident in the MAP‐80‐mmHg group. However, in patients with reported blood loss, no difference was observed in the amount of blood loss nor in transfusion need, suggesting that the clinical impact of this finding was likely minimal. Previous trials reporting intraoperative bleeding have been conflicting; however, most trials have found no difference [5]. Myocardial injury and major adverse cardiac events were not different across blood pressure groups, whereas the incidence of cardiac arrhythmias was higher in the group targeting MAP 80 mmHg. To our knowledge, this has not previously been reported in trials investigating blood pressure targets during general anaesthesia for non‐cardiac surgery. However, a recent review of blood pressure targets in both cardiac and non‐cardiac surgery reported an association between higher blood pressure targets and atrial fibrillation [18]. Increased doses of vasoactive drugs could potentially explain the increased incidence of cardiac arrhythmias. Trials conducted in the ICU setting have reported a relationship between higher blood pressure targets, use of vasopressors and increased risk of cardiac arrhythmia [24].
The one‐size‐fits‐all approach to blood pressure targets has been questioned and individualising blood pressure has been suggested as a better approach. When comparing the minimum target of MAP 60 mmHg and 90% of baseline systolic blood pressure, no difference was observed in any outcomes. In a previous trial by Futier et al., an individualised approach, like the one tested in this trial, decreased the risk of renal dysfunction and a composite outcome of systemic inflammatory response syndrome and organ failure [17]. However, in the IMPROVE‐trial by Saugel et al., targeting MAP ≥ 65 or mean night time MAP, no difference was observed in multiple postoperative complications including acute kidney injury [19]. Both of the mentioned trials had problems in achieving separation between the low blood pressure group and the individualised blood pressure group.
Observational studies suggest that hypotension, variably defined as a MAP below 40 to 75 mmHg, increases the risk of morbidity and mortality [14, 25, 26]. However, to date no clinical randomised trial has identified a lower blood pressure target evidently causing harm. This trial used MAP 60 mmHg as the lowest target, as proposed in a PeriOperative Quality Initiative (POQI) statement [15]. Across outcomes, there was no indication that a MAP target of 60 mmHg caused harm.
The findings of this trial should be interpreted in context of some limitations. First, the trial was not powered to detect small differences in postoperative outcomes. Hence, there is a risk of false negative results (Type 2 error). Second, we report results from multiple outcomes and there is therefore a risk of false positive results (Type 1 error). 95% confidence intervals were not adjusted for multiplicity. Second, the anaesthetic team providing care for the patient was not blinded to the interventions. Data on blood pressure were documented every 15 min by the treating clinician; thus, there is a potential risk of detection bias. Beat‐to‐beat electronic capture of intraoperative blood pressure is feasible but unfortunately not widely implemented in the participating centres. Third, a limited number of patients undergoing acute surgery were included, limiting generalisability to this group.
In conclusion, this feasibility trial did not meet all predefined feasibility goals, that is, inclusion of 50% of eligible patients and achieving follow‐up data on 90% of patients. However, separation in blood pressure and adherence to targets were adequate end better than most previous trials. Severe postoperative outcomes, including mortality, were relatively rare. Combined with the assumptions that any intraoperative anaesthesiological intervention is likely to only have a small too modest effect on the postoperative course, future trials will need to be large. Our findings suggest that trial complexity should be minimised to achieve this goal. New trial designs, such as platform trials or cluster‐randomised trials, and simplified consent procedures might be needed.
Author Contributions
M.B.P. and L.W.A.: conception and trial design, statistical analyses, drafting the manuscript and approval of final version. B.H.R. and S.Ø.M.: database design, data collection, statistical analyses, reviewing manuscript critically for important intellectual content and approval of final version. A.G., M.F.V., M.K.J., J.A.K.P. and S.T.V.: conception and trial design, reviewing manuscript critically for important intellectual content and approval of final version. O.L.S., J.B., K.G.B., R.B., L.B., K.B.P., J.G.R.R., C.P., C.S., J.T.S., M.E., L.D.D.P., S.T.Z., N.M.B., A.M.R., I.A.G., M.A.J., U.S.E., L.D.B., C.F.E., A.G.K., P.D.D., M.L.A. and H.B.N.: patient inclusion, reviewing manuscript critically for important intellectual content and approval of final version. C.H.N., C.D.B., L.K.A., F.G.H., L.W.F. and S.W.: data collection, reviewing manuscript critically for important intellectual content and approval of final version.
Funding
This work was supported by the Independent Research Fund Denmark, the Danish Cancer Society, the KEE memorial fund, and the Else and Mogens Wedell Wedellsborg's fund.
Conflicts of Interest
Simon T. Vistisen reports consultancy for Philips Healthcare paid to institution. The other authors declare no conflicts of interest.
Supporting information
Data S1: aas70315‐sup‐0001‐Supinfo1.pdf.
Figure S1: Trial overview.
Figure S2: Timeline for interventions and follow‐up.
Figure S3: Mean diastolic blood pressure 0–180 min of general anaesthesia.
Figure S4: Mean systolic blood pressure 0–180 min of general anaesthesia.
Table S1: Sites and site‐specific inclusion.
Table S2: Sites and site‐specific inclusion of elective patients.
Table S3: Surgeries included in GA‐targets trial.
Table S4: Reasons provided by the clinicians why the interventions were judged to cause potential harm.
Table S5: Reason for hospital admission.
Table S6: Preoperative vital signs.
Table S7: Preoperative arterial blood gas analyses.
Table S8: Preoperative biomarkers.
Table S9: Mean MAP during general anaesthesia.
Table S10: Mean MAP during postoperative intervention period.
Table S11: Compliance to blood pressure target when vasoactive drugs were used during general anaesthesia.
Table S12: Documented deviations from blood pressure targets.
Table S13: Biomarkers excluding patients with missing data.
Table S14: Quality of life and quality of recovery excluding patients with missing data.
Table S15: Intra‐ and perioperative data.
Table S16: Intraoperative blood gas analysis (0–180 min).
Table S17: Postoperative blood gas analysis.
Table S18: Major adverse cardiac events (MACE).
Table S19: Pulmonary complications.
Table S20: Surgical site infections.
Table S21: ICU admission.
Table S22: Postoperative pain and treatment of pain.
Table S23: Intraoperative antiemetic administration.
Table S24: Postoperative nausea and vomiting and antiemetic administration.
Table S25: Postoperative biomarkers.
Table S26: Acute kidney injury, KDIGO stages.
Acknowledgements
We gratefully acknowledge the statistical guidance from Anais Charles‐Nelson, Australian and New Zealand Intensive Care Research Centre, School of Public Health and Preventive Medicine, Monash University, Melbourne, Victoria, Australia.
Data Availability Statement
Six months after the publication of the last results, all de‐identified individual patient data will be made available for data sharing. Data will be available for any research purpose to all interested parties who have approval from an independent review committee and who have a methodologically sound proposal as determined by the steering committee of the current trial.
References
- 1. Liu L. L. and Larson M. D., “Patient Safety During Anesthesia: 100 Years of Progress Documented in Anesthesia & Analgesia,” Anesthesia & Analgesia 135 (2022): S37–S47. [DOI] [PubMed] [Google Scholar]
- 2. Armellini A., Chew S., Johnston S., Muralidharan V., Nikfarjam M., and Weinberg L., “The Hospital Costs of Complications Following Major Abdominal Surgery: A Retrospective Cohort Study,” BMC Research Notes 17 (2024): 59. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Hackett N. J., De Oliveira G. S., Jain U. K., and Kim J. Y., “ASA Class Is a Reliable Independent Predictor of Medical Complications and Mortality Following Surgery,” International Journal of Surgery 18 (2015): 184–190. [DOI] [PubMed] [Google Scholar]
- 4. Jammer I., Wickboldt N., Sander M., et al., “Standards for Definitions and Use of Outcome Measures for Clinical Effectiveness Research in Perioperative Medicine: European Perioperative Clinical Outcome (EPCO) Definitions: A Statement From the ESA‐ESICM Joint Taskforce on Perioperative Outcome Measures,” European Journal of Anaesthesiology 32 (2015): 88–105. [DOI] [PubMed] [Google Scholar]
- 5. Bolther M., Henriksen J., Holmberg M. J., Granfeldt A., and Andersen L. W., “Blood Pressure Targets During General Anaesthesia for Noncardiac Surgery: A Systematic Review of Clinical Trials,” European Journal of Anaesthesiology 39 (2022): 903–905. [DOI] [PubMed] [Google Scholar]
- 6. Høybye M., Lind P. C., Holmberg M. J., et al., “Fraction of Inspired Oxygen During General Anesthesia for Non‐Cardiac Surgery: Systematic Review and Meta‐Analysis,” Acta Anaesthesiologica Scandinavica 66 (2022): 923–933. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Bolther M., Henriksen J., Holmberg M. J., et al., “Ventilation Strategies During General Anesthesia for Noncardiac Surgery: A Systematic Review and Meta‐Analysis,” Anesthesia and Analgesia 135 (2022): 971–985. [DOI] [PubMed] [Google Scholar]
- 8. Zhao B., Zhang J., Xie Y., et al., “Intensive vs. Conventional Intraoperative Blood Pressure Management on Cardiovascular Events After Major Abdominal Surgery: The BP‐CARES Randomized Trial,” Journal of the American College of Cardiology 86 (2025): 892–906. [DOI] [PubMed] [Google Scholar]
- 9. Kant M., van Klei W. A., Hollmann M. W., et al., “Proactive vs. Reactive Treatment of Hypotension During Surgery: The PRETREAT Randomized Clinical Trial,” JAMA 334 (2025): 1905–1914. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Marcucci M., Painter T. W., Conen D., et al., “Hypotension‐Avoidance Versus Hypertension‐Avoidance Strategies in Noncardiac Surgery: An International Randomized Controlled Trial,” Annals of Internal Medicine 176 (2023): 605–614. [DOI] [PubMed] [Google Scholar]
- 11. Harris P. A., Taylor R., Minor B. L., et al., “The REDCap Consortium: Building an International Community of Software Platform Partners,” Journal of Biomedical Informatics 95 (2019): 103208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Harris P. A., Taylor R., Thielke R., Payne J., Gonzalez N., and Conde J. G., “Research Electronic Data Capture (REDCap)—A Metadata‐Driven Methodology and Workflow Process for Providing Translational Research Informatics Support,” Journal of Biomedical Informatics 42 (2009): 377–381. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Brower R. G., Matthay M. A., Morris A., et al., “Ventilation With Lower Tidal Volumes as Compared With Traditional Tidal Volumes for Acute Lung Injury and the Acute Respiratory Distress Syndrome,” New England Journal of Medicine 342 (2000): 1301–1308. [DOI] [PubMed] [Google Scholar]
- 14. Weinberg L., Li S. Y., Louis M., et al., “Reported Definitions of Intraoperative Hypotension in Adults Undergoing Non‐Cardiac Surgery Under General Anaesthesia: A Review,” BMC Anesthesiology 22 (2022): 69. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Saugel B., Fletcher N., Gan T. J., et al., “PeriOperative Quality Initiative (POQI) International Consensus Statement on Perioperative Arterial Pressure Management,” British Journal of Anaesthesia 133 (2024): 264–276. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Pælestik M. B., Risager B. H., Mortensen S. Ø., et al., “Positive End‐Expiratory Pressure, Tidal Volume, and Fraction of Inspired Oxygen During General Anaesthesia in High‐Risk Patients‐A Randomised Clinical Feasibility Trial,” Acta Anaesthesiol Scandinavica 70, no. 8 (2026): e70309, 10.1111/aas.70309. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Futier E., Lefrant J. Y., Guinot P. G., et al., “Effect of Individualized vs Standard Blood Pressure Management Strategies on Postoperative Organ Dysfunction Among High‐Risk Patients Undergoing Major Surgery: A Randomized Clinical Trial,” JAMA 318 (2017): 1346–1357. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. D'Amico F., Pruna A., Putowski Z., et al., “Low Versus High Blood Pressure Targets in Critically Ill and Surgical Patients: A Systematic Review and Meta‐Analysis of Randomized Controlled Trials,” Critical Care Medicine 52 (2024): 1427–1438. [DOI] [PubMed] [Google Scholar]
- 19. Saugel B., Meidert A. S., Brunkhorst F. M., et al., “Individualized Perioperative Blood Pressure Management in Patients Undergoing Major Abdominal Surgery: The IMPROVE‐Multi Randomized Clinical Trial,” JAMA 334 (2025): 1893–1904. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Wu X., Jiang Z., Ying J., Han Y., and Chen Z., “Optimal Blood Pressure Decreases Acute Kidney Injury After Gastrointestinal Surgery in Elderly Hypertensive Patients: A Randomized Study: Optimal Blood Pressure Reduces Acute Kidney Injury,” Journal of Clinical Anesthesia 43 (2017): 77–83. [DOI] [PubMed] [Google Scholar]
- 21. Wanner P. M., Wulff D. U., Djurdjevic M., Korte W., Schnider T. W., and Filipovic M., “Targeting Higher Intraoperative Blood Pressures Does Not Reduce Adverse Cardiovascular Events Following Noncardiac Surgery,” Journal of the American College of Cardiology 78 (2021): 1753–1764. [DOI] [PubMed] [Google Scholar]
- 22. Garg A. X., Marcucci M., Cuerden M. S., et al., “A Sub‐Study of the POISE‐3 Randomized Trial Examined Effects of a Perioperative Hypotension‐Avoidance Strategy Versus a Hypertension‐Avoidance Strategy on the Risk of Acute Kidney Injury,” Kidney International 107 (2025): 155–168. [DOI] [PubMed] [Google Scholar]
- 23. Hu A. M., Qiu Y., Zhang P., et al., “Higher Versus Lower Mean Arterial Pressure Target Management in Older Patients Having Non‐Cardiothoracic Surgery: A Prospective Randomized Controlled Trial,” Journal of Clinical Anesthesia 69 (2021): 110150. [DOI] [PubMed] [Google Scholar]
- 24. Richards‐Belle A., Hylands M., Muttalib F., et al., “Lower Versus Higher Exposure to Vasopressor Therapy in Vasodilatory Hypotension: A Systematic Review With Meta‐Analysis,” Critical Care Medicine 51 (2022): 254–266. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Gregory A., Stapelfeldt W. H., Khanna A. K., et al., “Intraoperative Hypotension Is Associated With Adverse Clinical Outcomes After Noncardiac Surgery,” Anesthesia and Analgesia 132 (2021): 1654–1665. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Wesselink E. M., Kappen T. H., Torn H. M., Slooter A. J. C., and van Klei W. A., “Intraoperative Hypotension and the Risk of Postoperative Adverse Outcomes: A Systematic Review,” British Journal of Anaesthesia 121 (2018): 706–721. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data S1: aas70315‐sup‐0001‐Supinfo1.pdf.
Figure S1: Trial overview.
Figure S2: Timeline for interventions and follow‐up.
Figure S3: Mean diastolic blood pressure 0–180 min of general anaesthesia.
Figure S4: Mean systolic blood pressure 0–180 min of general anaesthesia.
Table S1: Sites and site‐specific inclusion.
Table S2: Sites and site‐specific inclusion of elective patients.
Table S3: Surgeries included in GA‐targets trial.
Table S4: Reasons provided by the clinicians why the interventions were judged to cause potential harm.
Table S5: Reason for hospital admission.
Table S6: Preoperative vital signs.
Table S7: Preoperative arterial blood gas analyses.
Table S8: Preoperative biomarkers.
Table S9: Mean MAP during general anaesthesia.
Table S10: Mean MAP during postoperative intervention period.
Table S11: Compliance to blood pressure target when vasoactive drugs were used during general anaesthesia.
Table S12: Documented deviations from blood pressure targets.
Table S13: Biomarkers excluding patients with missing data.
Table S14: Quality of life and quality of recovery excluding patients with missing data.
Table S15: Intra‐ and perioperative data.
Table S16: Intraoperative blood gas analysis (0–180 min).
Table S17: Postoperative blood gas analysis.
Table S18: Major adverse cardiac events (MACE).
Table S19: Pulmonary complications.
Table S20: Surgical site infections.
Table S21: ICU admission.
Table S22: Postoperative pain and treatment of pain.
Table S23: Intraoperative antiemetic administration.
Table S24: Postoperative nausea and vomiting and antiemetic administration.
Table S25: Postoperative biomarkers.
Table S26: Acute kidney injury, KDIGO stages.
Data Availability Statement
Six months after the publication of the last results, all de‐identified individual patient data will be made available for data sharing. Data will be available for any research purpose to all interested parties who have approval from an independent review committee and who have a methodologically sound proposal as determined by the steering committee of the current trial.