Continuous vs Intermittent β-Lactam Antibiotic Infusions in Critically Ill Patients with Sepsis The BLING III Randomized Clinical Trial

Dulhunty et al. JAMA 2024. 10.1001/jama.2024.9779

Clinical Question

• In critically ill adult patients with presumed sepsis, does the administration of β-lactam antibiotics via a continuous infusion compared to an intermittent infusion reduce 90-day mortality?

Background

• β-lactam antibiotics have time-dependent bactericidal activity
• It is hypothesised that improved antibiotic exposure with continuous infusion will lead to longer time above MIC, with potentially improved bacterial eradication
• Additionally, underdosing may occur in sepsis due to:
  • Increased cardiac output with potentially increased drug clearance
  • Capillary leak and increased volume of distribution
• BLING II (2015) showed no difference in alive ICU-free days at day 28 between continuous and intermittent administration of β-lactam antibiotics
• MERCY (2023) showed a non-significant 2% reduction in mortality in with continuous infusions of Meropenem

Design

• Multicentre, multinational randomised controlled trial
• Randomisation 1:1 via an online minimisation algorithm
  • Stratified by study site
• Open label
• Power calculation guided by BLING II
  • Sample size of 7000 required to achieve 90% power to detect 3.5% difference in all-cause mortality at 90 days including 5% loss to follow up
  • Assumption of baseline mortality of 27.5%, at α of 0.05
• Modified intention-to-treat analysis for primary end point
• Prespecified adjusted analysis accounting for sex, APACHE II score, source of admission (emergency vs elective surgery vs other), type of β-lactam antibiotics used before randomisation
• 5 pre-specified subgroups: Pulmonary infection, β-lactam antibiotic, Age (< or > 65), Sex, APACHE II score (< or > 25)
• Clinical cure defined as completion of β-lactam antibiotic by day 14 without recommencement of antibiotics within 48hrs of cessation for same infective episode (investigator assessment of clinical response)
• Independent DSMB conducted pre-specified midpoint safety analysis

Setting

• 104 intensive care units in 7 countries:
  • Australia (25) / New Zealand (6) / UK (55) / Belgium (8) / France (5) / Sweden (3) / Malaysia (2)
• Recruitment from March 2018 – January 2023

Population

• Inclusion:
• Adults in ICU with
  • Documented site/strong suspicion of infection
  • ≥1 organ dysfunction criteria met in previous 24hrs
  • Treatment with piperacillin-tazobactam or meropenem commenced within prior 24hrs
  • Expected to remain in ICU for at least the next calendar day
• Exclusion:
  • Receipt of piperacillin-tazobactam or meropenem for more than 24hrs during current infectious episode
  • Patient known or suspected to be pregnant
  • Known allergy to piperacillin-tazobactam, meropenem or penicillin
  • Requiring renal replacement therapy (acute or chronic)
  • Previously enrolled in BLING III
  • Palliative treatment only, limits to care, death is imminent, or death is likely within 90 days due to underlying process
• 29042 patients screened → 7202 randomised
  • 3595 to receive continuous antibiotics
  • 3607 to receive intermittent antibiotics
• Comparing baseline characteristics of continuous vs. intermittent group
  • Age: 59.3 vs 59.6 years
  • Male: 66 vs 65 %
  • APACHE II: 19.6 vs 19.5
  • Inotropes/vasopressors in 24hr prior to randomisation: 71 vs 70 %
  • Received antibiotics in 24hr prior to randomisation: 67 vs 70 %
  • Source of admission:
    • ED: 34 vs 33 %
    • Wards: 28 vs 30 %
    • OT post emergency surgery: 21 vs 20 %
  • Primary site of infection:
    • Pulmonary 59 vs 60 %
    • Intra-abdominal 13 vs 13 %
    • Blood 8 vs 8 %

Intervention

• Continuous infusion of β-lactam antibiotic (for clinician prescribed duration OR until ICU discharge)
• Infusion over 24 hours began with a bolus of intermittent dose as a load
  • Duration: median 5.8 days
  • Cumulative Dose: median 72 g (40.5 – 108) of Piperacillin-Tazobactam, and 17g (9 – 29) of meropenem

Control

• Intermittent infusion of β-lactam antibiotic
• Infusion over 30mins
  • Duration: median 5.7 days
  • Cumulative Dose: median 70.8 g (40.5 – 103.5) of Piperacillin-Tazobactam, and 16.0g (8.0 – 28.0) of meropenem

Management common to both groups

• Dose of antibiotic determined by clinician
  • Defined daily dose was 14g for Piperacillin-Tazobactam and 3g for Meropenem
  • Dose could be modified in response to patient changes
• Switch between either was permitted as long as remained in same treatment group
• Reversion to intermittent dosing after 14 days or ICU discharge
• All other management at discretion of treating clinician

Outcome

• Primary outcome
  • No significant difference in all-cause mortality at day 90:
    • 24.9% (continuous) vs 26.8% (intermittent)
      • Absolute difference -1.9% (95% CI -4.9 to 1.1)
      • OR 0.91 (95% CI 0.81 to 1.01), p = 0.08
    • Note that for adjusted analysis absolute difference -2.2% (95% CI -5.5 to 1.1)
      • OR 0.89 (95% CI 0.79 – 0.99), p = 0.04
• Secondary outcomes:
• Comparing continuous vs intermittent infusion
• Significantly greater in intervention group:
  • Clinical cure: 55.7% vs 50.0%
    • Absolute difference 5.7% (95% CI 2.4 to 9.1), p < 0.001
    • OR 1.26 (95% CI 1.15 – 1.38)
• No significant difference in:
  • New acquisition/colonisation/infection with MRO or C Difficile: 7.2 vs 7.5%
    • Absolute difference -0.3% (95% CI -1.9 to 1.4) p = 0.65
  • All-cause ICU mortality: 17.1 vs 18.4%
    • Absolute difference -1.3% (95% CI -4 to 1.4) p = 0.35
  • All-cause hospital mortality: 23.3 vs 25 %
    • Absolute difference -1.8% (95% CI -4.8 to 1.2) p = 0.27
  • Days alive and free of ICU, hospital, mechanical ventilation or RRT by day 90
• Subgroups:
  • All point estimates favoured continuous infusion but none statistically significant
• Adverse Events:
  • 0.3% in continuous group vs 0.2% in intermittent group

Authors’ Conclusions

• The observed difference in 90-day mortality between continuous vs intermittent infusions of β-lactam antibiotics did not meet statistical significance in the primary analysis. However, the confidence interval around the effect estimate includes the possibility of both no important effect and a clinically important benefit in the use of continuous infusions in this group of patients

Strengths

• Extremely large RCT with robust internal and external validity
  • Detailed screening log with a majority of those screened but not enrolled meeting exclusion criteria or declining to take part
  • Large number of centres across multiple countries
  • Balanced baseline characteristics – similar organisms cultured, baseline antibiotic use
  • Doses and duration of antibiotics similar between groups (eTable 7, eFigure 1 and eFigure 2)
• Detailed collection of infection site and antimicrobial data (eTable 3 – 6, 9)
• Minimal loss to follow up

Weaknesses

• Unblinded
• Dosing reverted to intermittent pre and post ICU – this may bias results towards the null
• No consensus as to definition of clinical cure:
  • Whilst the definition used in this trial is pragmatic and objective, is 48 hours long enough following cessation of antimicrobials to say clinical cure has been achieved?
• Potential that patients with non-infectious causes of organ dysfunction are included
• No reporting on therapeutic dose monitoring, or sensitivities / susceptibilities
• Approximately 1/3rd of patients had a protocol violation (eTable 12):
  • 34% (continuous) vs 32% (intermittent) of patients
  • A large proportion were related to administration or dosing, of which the largest contributor was a pause or delay for > 1 hour:
    • 15.3% (continuous) and 22.6% (intermittent)

The Bottom Line

• In adult ICU patients with sepsis, continuous infusions should be commenced if IV access is available, given potential for benefit alongside no evidence for harm
• I will be interested to see the results of any economic analyses or work into the impact of continuous infusions on bedside workload

External Links

• article Continuous vs Intermittent β-Lactam Antibiotic Infusions in Critically Ill Patients With Sepsis: The BLING III Randomized Clinical Trial
• further reading Resolving the Dilemma on Continuous vs Intermittent β-Lactam Antibiotics in Sepsis – JAMA editorial
• further reading Prolonged vs Intermittent Infusions of β-Lactam Antibiotics in Adults With Sepsis or Septic Shock – A Systematic Review and Meta-Analysis

Metadata

Summary author: Audrey Guo
Summary date: July 20th 2024
Peer-review editor: George Walker

Picture by: Jordan Rushton / Pexels