Individualized Perioperative Blood Pressure Management in Patients Undergoing Major Abdominal Surgery: The IMPROVE-multi Randomized Clinical Trial

 B Saugel. JAMA 2025; doi: 10.1001/jama.2025.17235

Clinical Question

• In patients undergoing major intra-abdominal surgery, does an individualised blood pressure target, compared to standard care, improve rates of acute kidney injury, acute myocardial infarction, non-fatal cardiac arrest or death within the first 7 days? 

Background

• Intra-operative hypotension is associated with increased organ injury and death 
• Mean blood pressures < 55mmHg, or SBP < 90mmHg may be associated with myocardial and acute kidney injury
• This risk increases further in vulnerable patients with pre-existing organ dysfunction or undergoing a high risk operation
• Standard BP targets are non-specific and do not take into account chronic patient compensations and arterial/cardiac modelling
• It is proposed that by personalising blood pressure targets during prolonged surgery, that outcomes of organ dysfunction or death would be improved compared to the standard arm

Design

• Multi-centre randomised single blinded clinical superiority trial 
• Included those undergoing high risk intra-abdominal surgical operations 
• Interim analysis by an independent safety committee 
• Sample size was calculated from a historical composite incidence of the primary outcome being 25% in the standard care and 17% in the treatment arm with an absolute risk reduction of 8% 
• Randomised 1:1 using a block strategy (2 and 4) using a central online data management system 
  • Stratified by centre

Setting

• 15 German Medical Centres
• Study period included Feb 2023 to April 2024 and last follow-up July 2024

Population

• Inclusion:
  • > 45 years old
  • Elective abdominal surgery 
  • Expected to last > 90 mins 
  • Presence of >1 high risk criteria (see list below)
• Exclusion:
  • Undergoing nephrectomy 
  • Prior liver, heart or lung transplant 
  • Sepsis 
  • On renal replacement therapy
• Participant Numbers
  • 1272 enrolled and 1142 randomised
    • 571 in the individualised treatment group and 571 in the routine treatment group
    • 7621 patients fulfilled inclusion criteria, of which 6349 excluded:
      • Main reasons for exclusion included: 1130 met exclusion criteria, 3263 excluded for ‘other reasons’ and 843 excluded for logistical reasons 
• Comparing baseline characteristics of individual vs. standard care group
  • Age: 67 vs 66
  • Male: 67.2 vs 64.6%   
  • BMI: 26.2 vs 25.9 
  • ASA status
    • I: 2.1 vs 1.6 % 
    • II: 41.4 vs 41.4% 
    • III: 55.7 vs 55.7%
    • IV: 0.7 vs 1.2%
  • High risk criteria (not mutually exclusive): 
    • Expected Anaesthesia duration >180min: 79 vs 80.4% 
    • Immunodeficiency: 74.6 vs  72.1%
    • Kidney impairment: 69.8 vs 66.5%
    • Age >65: 59.8 vs 56.1% 
    • BMI > 30: 23.5 vs 21.7% 
    • Current smoking OR pack years >15: 23.5 vs 25%
    • Medicated diabetes: 18.5 vs 15.2% 
    • Coronary artery disease: 12.5 vs 11.8% 
    • Exercise tolerance <4 METS: 6.9 vs 6.7%  
    • COPD or pulmonary fibrosis: 5.5 vs 5.1% 
    • Stroke history: 4.8 vs 4.8% 
    • Chronic heart failure: 3.7 vs 3.0%  
    • Valvular heart disease: 3.7 vs 1.8%
    • Peripheral arterial disease: 1.4 vs 2.8% 
    • Liver cirrhosis: 1.2 vs 2.8% 
  • Other significant co-morbidities 
    • Solid organ cancers: 75.8 vs 75.3%
    • Chronic arterial hypertension: 54.1 vs 46.7% 
    • eGFR <60: 10.2 vs 9.2%

Intervention

• Randomisation occurred prior to induction of anaesthesia 
• Calculating blood pressure targets: 
  • At home or at hospital (~95% completed within hospital) 
    • 94.4 vs 95.2% in hospital
  • Mean MAP calculated between midnight and 0600 (30 minute intervals)
    • 12 vs 12 measurements
    • Mean MAP 84 vs 82mmHg
  • MAP limits of 65mmHg and 110mmHg in the setting of low or high measured night time MAP 
• Targets from induction until 2 hours after surgery
• Treatment to maintain MAP was determined by treating anaesthetist

Control

• MAP 65mmHg treatment determined by treating anaesthetist 

Management common to both groups

• Type of surgery 
  • General: 45.7 vs 47.6% 
  • Urology 40.9 vs 37%
  • Gynaecology: 9 vs 11.3%
  • Liver: 4.4 vs 3.5%   
  • Vascular 0 vs 0.5
• Arterial Line: 74.3 vs 74.6%
• Duration of surgery: 207 vs 211 minutes 
• Post operative ICU admission: 47.3 vs 49.9% 
• Crystalloid: 3 vs 3L 
• Colloids: 750 vs 500ml
• Blood products: 2 vs 2 units
• Vasopressors: 94.5 vs 87.3%
  • No significant difference between the type of vasopressor used (noradrenaline vs adrenaline vs ephedrine vs phenylephrine or akrinor)
• Area under a MAP of 65mmHg (mmHg x min): 6 vs 48
• Proportion of time with MAP > preoperative MAP: 79 vs 57% 

Outcome

• Primary outcome (comparing individual vs standard care):
  • Composite of AKI (> 50% increase in creatinine), Myocardial Injury (according to Fourth Universal Definition of Myocardial Infarction), Cardiac Arrest or Death within first 7 post operative days:
    • 33.5%  vs 30.5% events in the standard group
    • RR 1.10, 95% CI 0.93 – 1.30, p = 0.31
  • AKI: 12.7 vs 10.2%
  • Acute myocardial injury: 27.4 vs 24.6%
  • Non-fatal cardiac arrest: 0.4 vs 0.0%
  • Death: 0 vs 0.7%
• Secondary outcomes:
  • Multiple secondary outcomes (22)
  • No significant difference in any of the secondary outcomes
    • Primary outcome and its components at 30 or 90 days
    • Infectious complications 
    • Unplanned hospital readmissions at 30 days

Authors’ Conclusions

• Among patients at high risk of postoperative complications undergoing major abdominal surgery, individualized perioperative blood pressure management did not decrease the composite outcome of acute kidney injury, acute myocardial injury, nonfatal cardiac arrest, or death within the first 7 postoperative days

Strengths

• Large, multi-site randomised controlled trial 
• Baseline characteristics balanced between groups 
• Included high risk patients with multiple co-morbidities who are most likely to benefit from this intervention with attempts to assess for heterogeneity of treatment effect by baseline severity of pre-operative hypertension (eTable3)
• Separation achieved with higher median MAPs in individual group
• Very few protocol deviations
• Plausibility for intervention with physiological bases compared to standard practice 
• Builds on other trials and expands evidence in this area (e.g. PRETREAT)

Weaknesses

• Composite outcome and attendant issues such as component events of hugely varying clinical importance and patient centred  
• Limited information on other perioperative factors 
• Anaesthetists unblinded
• Approximately 25% of patients did not have arterial lines – whilst similar between groups this may be less reliable especially if the patient is not in supine position
• Night time blood pressures may not be reflective of optimal blood pressure to target
• Difficult to tease out the effect of post operative management of patients  compared to intra-operative management alone with respect to 7 day outcomes
• There may be heterogeneity of treatment effect in which some groups (e.g. GI surgery) may suffer harm from greater doses of vasoconstrictors (splanchnic vasoconstriction)
• Noradrenaline used most commonly – some animal bypass models suggest that renal medullary oxygenation is worse with noradrenaline compared to other vasopressors such as vasopressin

The Bottom Line

• This study, in conjunction with other recently published work, suggests that targeting a MAP of 65mmHg is safe during the intra-operative period 
• This may limit the use of vasoactive medications, colloids and monitoring, all which cary a patient risk along with environmental and healthcare cost  

External Links

• article Individualized Perioperative Blood Pressure Management in Patients Undergoing Major Abdominal Surgery
• editorial Perioperative Outcomes—The Limits of Blood Pressure–Centered Strategies

Metadata

Summary author: Jake Reeve
Summary date: 19th November 2025
Peer-review editor: George Walker

Picture by: CottonBro Studio/Pexels