Empirical Meropenem Versus Piperacillin/Tazobactam for Critically Ill Adults With Sepsis: Feasibility of a Randomised Trial.

Abstract

BACKGROUND: Meropenem and piperacillin/tazobactam are commonly used empirical antibiotics in critically ill adults with sepsis, but whether one is superior to the other is uncertain. METHODS: The Empirical Meropenem versus Piperacillin/Tazobactam for Adult Patients with Sepsis (EMPRESS) trial is an ongoing investigator-initiated, randomised, open-label, adaptive clinical trial with an integrated feasibility phase comparing empirical treatment with meropenem versus piperacillin/tazobactam in critically ill adults with sepsis. The integrated feasibility phase enrolled 200 participants across 10 intensive care units (ICUs) in Denmark between 28 June and 12 December 2025. Five pre-specified feasibility criteria were evaluated; if all feasibility criteria were met, the trial would proceed unaltered, whereas failure to meet one or more criteria would require intervention and re-evaluation. RESULTS: We randomised 200 of 284 screened patients (70.4%). The median age was 70&#x2009;years (interquartile range (IQR): 60-77), 65.5% were males. At randomisation, 80.0% received vasopressors or inotropes, and 43.5% were on invasive mechanical ventilation. Four of five pre-specified feasibility criteria were met: time to completion of the feasibility phase (5.5&#x2009;months vs. threshold <&#x2009;12.0&#x2009;months), recruitment proportion (70.4% vs. threshold &#x2265;&#x2009;50.0%), proportion of participants without consent to the continued collection of data (2.5% vs. threshold <&#x2009;5.0%) and protocol adherence (81.0% vs. threshold &#x2265;&#x2009;75.0%). The proportion of participants with timely primary outcome data availability (30-day mortality) within 45&#x2009;days was 85.5% and below the pre-specified threshold of &#x2265;&#x2009;95.0%. The proportions were low in the first 3&#x2009;months (33.3%, 22.2% and 30.8%, respectively), increasing to 95.8% in the last month of the feasibility phase. All-cause mortality at 30&#x2009;days was 30.5%, and specific serious adverse reactions occurred in 4.0% of participants. CONCLUSIONS: In this integrated feasibility evaluation of the EMPRESS trial comparing empirical meropenem versus piperacillin/tazobactam in critically ill adults with sepsis, four of five pre-specified feasibility criteria were met. The unmet criterion, timely primary outcome data availability, improved substantially during the feasibility phase. We consider the trial feasible and will proceed without modifications. EDITORIAL COMMENT: This feasibility study assessed recruitment, randomised allocation and data collection for the multicentre EMPRESS trial. For adaptive trials on trial platforms, careful interim checking of trial design functions is an important and necessary process. TRIAL REGISTRATION: Clinical Trials Information System EUCT number: 2023-509703-33-00; ClinicalTrials.gov identifier: NCT06184659; Universal Trial Number: U1111-1301-6379.