Effect of Individualized vs Standard Blood Pressure Management Strategies on Postoperative Organ Dysfunction Among High-Risk Patients Undergoing Major Surgery

Futier E et al. JAMA 2017; doi:10.1001/jama.2017.14172

Clinical Question

  • In high-risk patients undergoing surgical procedures, does a strategy of targeting individualised systolic blood pressure compared to standard practice reduce organ dysfunction?


  • The number of patients undergoing surgical procedures worldwide is growing
  • More recent trials of early goal-directed therapy pathways have not been shown to be superior to standard therapy
  • SEPSISPAM trial found no mortality difference in patients with septic shock who underwent resuscitation targeting a mean BP of either 65 to 70mmHg or 80 to 85mmHg


  • Multicenter, stratified, parallel-group, randomised controlled trial
  • 1:1 ratio to controlled and intervention group
  • Randomisation using minimisation algorithm and stratified according to study site, urgency or surgery and surgical site
  • Sample size of 268 patients for 90% power to detect an absolute difference of 20% in primary outcome, a 2-sided α-level of 0.05, assuming an event rate of 40% in control arm
  • Recruitment target of 300 to account for potential protocol deviations and withdrawal of consent
  • Intention to treat analysis
  • Intervention could not be blinded but attempts to reduce bias by online randomisation to ensure concealment of group assignment and blinding post-op carers


  • 9 university and non-university hospitals in France
  • December 2012 – August 2016


  • Inclusion:
    • Aged above 50
    • Undergoing surgery under general anaesthesia with an expected duration of more than 2 hours
    • American Society of Anesthesiology score of 2 or higher
    • Preoperative Acute Kidney Injury Risk Index of class 3 or higher
  • Exclusion:
    • Uncontrolled hypertension (systolic BP > 180mmHg or diastolic BP > 110mmHg)
    • Chronic kidney disease (GFR < 30ml/min/1.73m2 or requiring renal replacement therapy)
    • Acute or decompensated heart failure
    • Acute coronary syndrome
    • Preoperative sepsis
    • Already receiving norepinephrine infusion
    • Pregnant or breastfeeding
  • 1494 patients screened, 298 randomised; 149 in each arm
  • Baseline characteristics similar with 84% of procedures being elective procedures and the vast majority of surgery in both groups involving the abdomen (93.9% and 96.6%). Significant proportion of patients in both groups had pre-existing chronic arterial hypertension (82%)


  • Continuous infusion of norepinephrine used to target a systolic blood pressure (SBP) within 10% of reference value
  • Infusion rate reduced if heart rate < 40 bpm


  • Intravenous ephedrine at 6 mg boluses (up to 60mg) administered as recommended for SBP < 80 mmHg or if 40% lower than reference value
  • If SBP remained below target after maximum dose of 60mg ephedrine administered, norepinephrine permitted as rescue therapy

Management common to both groups

  • Resting blood pressure from preoperative assessment used as reference value; if not available, the blood pressure was measured on the surgical ward, the day before surgery
  • Invasive blood pressure measurement through a radial catheter required
  • Ringers Lactate solution infused at 4ml/kg/hr as maintenance fluid
  • Additional fluid were given based on a protocolised haemodynamic algorithm (stroke volume index optimisation) using 6% hydroxyethyl starch solution in 250ml boluses
  • Intervention period lasted from anaesthetic induction to 4 hours after the completion of surgery
  • Decisions regarding all other aspects of care was left at the discretion of the attending physician


  • Primary outcome: Composite of systemic inflammatory response syndrome (SIRS) and at least 1 organ system dysfunction for renal, respiratory, cardiovascular, neurological and coagulation system occurring by day 7 after surgery – Lower event rate in intervention group
    • 38.1% in intervention group compared to 51.7% in control
    • Absolute risk reduction (ARR): 14% (95% CI -25% to -2%; P = 0.02)
    • Number needed to treat (NNT): 8
    • Fragility index (FI): 3
  • Secondary outcome: Individual components of the primary outcome and adverse events
    • Renal dysfunction lower:
      • Intervention group 32.7% vs Control 49.0%
      • ARR: 16% (95% CI -27% to -5%; P = 0.01)
      • NNT: 7
      • FI: 7
    • Altered consciousness lower:
      • Intervention group 5.4% vs Control 15.9%
      • ARR: 10% (95% CI -17% to -3%; P = 0.007)
      • NNT: 10
      • FI: 5
    • No significant difference in:
      • SIRS score
      • Coagulation SOFA score
      • Hypoxaemia
      • Pneumonia
      • ARDS
      • Reintubation rates
      • SOFA scores on day 1, 2 and 7
      • Incidence of sepsis
      • Surgical complications
      • Death at 30 days

Authors’ Conclusions

  • Management targeting an individualised systolic blood pressure, compared with standard management, reduced the risk of postoperative organ dysfunction


  • Relevant and important clinical question
  • Targeted higher risk population
  • Clear separation of systolic, mean and diastolic BP between groups at the end of intervention (no difference between groups pre-induction of anaesthesia)
  • Although a composite score was used as primary outcome, the components are clinically relevant


  • Slow recruitment despite high number of patients screened
  • Majority of patients were excluded due to AKI risk index of < 3
  • The routine use of invasive blood pressure monitoring may not be reflective of current practice

The Bottom Line

  • Targeting individualised BP reduces the rate of post-operative organ dysfunction in patients undergoing surgery
  • The increased used of arterial cannulation and norepinephrine infusion does not seem to increase adverse outcomes

External Links


Summary author: Adrian Wong
Summary date: 5/10/2017
Peer-review editor: Segun Olusanya

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.