Introduction
Optimizing antibiotic therapy in the intensive care unit (ICU) is a cornerstone of managing sepsis, yet it remains one of the most complex challenges in clinical pharmacology. For critically ill patients requiring renal replacement therapy (RRT), the challenge is magnified. The physiological shifts associated with critical illness—such as capillary leak, altered volume of distribution, and fluctuating organ function—are compounded by the extrinsic clearance provided by RRT. Historically, dosing for meropenem and piperacillin/tazobactam in this population has been based on small-scale studies or manufacturer recommendations that often fail to account for the diversity of RRT modalities and patient-specific factors. The SMARRT Study, recently published in Intensive Care Medicine, provides much-needed clarity through a large-scale, prospective, multinational pharmacokinetic (PK) analysis.
Highlights
1. RRT intensity, duration, and residual urine output are the primary determinants of meropenem and piperacillin/tazobactam clearance in critically ill patients.
2. Extended or continuous infusions consistently outperform short-duration bolus infusions in achieving therapeutic targets while using lower total daily doses.
3. The study provides validated dosing nomograms that allow clinicians to personalize antimicrobial therapy based on specific RRT settings and patient characteristics.
The Clinical Dilemma: Pharmacokinetics in the RRT Patient
Beta-lactam antibiotics like meropenem and piperacillin exhibit time-dependent killing, meaning their efficacy is linked to the percentage of the dosing interval during which the unbound drug concentration remains above the minimum inhibitory concentration (fT > MIC). In the ICU, the target is often 100% fT > MIC or even higher (e.g., 100% fT > 4x MIC) to ensure efficacy against less susceptible pathogens like Pseudomonas aeruginosa and to prevent the emergence of resistance.
However, RRT adds a layer of complexity. Whether a patient is on continuous veno-venous haemodialysis (CVVHD), haemofiltration (CVVH), or sustained low-efficiency dialysis (SLED), the drug clearance is highly variable. Over-dosing can lead to neurotoxicity (particularly with meropenem), while under-dosing risks treatment failure and mortality. Until now, clinicians lacked a robust, generalizable framework to navigate these variables.
Study Design and Methodology
The SMARRT study was a prospective, multinational pharmacokinetic investigation involving 300 patients across 22 ICUs in 12 countries. This diversity is a significant strength, as it captures various clinical practices and RRT protocols globally.
Patient Population and RRT Modalities
The cohort included patients receiving various forms of RRT: continuous veno-venous haemodialysis (13.0%), haemofiltration (23.3%), haemodiafiltration (48.4%), and sustained low-efficiency dialysis (15.3%). This broad representation ensures that the findings are applicable to most modern ICU settings.
Analytical Approach
Researchers developed independent population PK models using data from 234 patients (comprising 8,322 blood samples) and externally validated these models using data from an additional 66 patients (560 samples). Monte Carlo simulations were then performed to determine the probability of target attainment (PTA) for various dosing regimens against Enterobacterales and Pseudomonas aeruginosa. The primary targets were unbound concentrations exceeding the MIC for 100% of the dosing interval.
Key Findings: Factors Driving Drug Clearance
The study identified three critical factors that dictate the necessary dose of meropenem and piperacillin/tazobactam in RRT patients: RRT intensity, RRT duration, and urine output.
The Impact of RRT Intensity and Duration
Clearance was directly proportional to the intensity of the RRT (e.g., the effluent flow rate). Higher intensity RRT naturally requires higher doses to maintain therapeutic levels. Interestingly, the duration of RRT also played a role, reflecting the stabilization of the patient’s volume status and drug distribution over time.
The Role of Residual Urine Output
One of the most clinically relevant findings was the significant impact of residual urine output. Even in patients requiring RRT, native renal clearance can contribute meaningfully to drug elimination. The study demonstrated that failing to account for urine output often leads to under-dosing in patients who are not completely anuric.
Superiority of Extended Infusions
In virtually every simulation scenario, extended or continuous infusions (e.g., infusions lasting 3-4 hours or continuous 24-hour administration) were superior to short (30-minute) infusions. Extended infusions achieved higher PTA with lower total daily doses. For example, a lower dose given as a continuous infusion was more likely to hit the 100% fT > MIC target than a higher dose given as a bolus. This has significant implications for both cost-effectiveness and the reduction of dose-related toxicities.
Results: Predictive Performance and Validation
The predictive performance of the developed models was exceptionally high. The mean prediction errors were remarkably low (-5.2% for meropenem and -16.9% for piperacillin), suggesting that the models are highly reliable for clinical application. The researchers utilized these models to create dosing nomograms, which provide specific dosage recommendations based on RRT modality, effluent rate, and the patient’s urine output.
Expert Commentary and Clinical Application
The SMARRT study represents a shift toward precision medicine in the ICU. The finding that residual renal function (measured via urine output) must be factored into dosing even during RRT is a vital takeaway for bedside clinicians. Many traditional dosing tables assume zero renal function once RRT is initiated, which this study proves is a flawed assumption.
Furthermore, the advocacy for extended infusions aligns with a growing body of evidence in the broader ICU population but provides specific evidence for the RRT subgroup. By using continuous infusions, clinicians can maintain steady-state concentrations, avoiding the peaks that contribute to toxicity and the troughs that allow for bacterial regrowth.
Study Limitations
While the study is robust, it is important to note that PK targets are surrogates for clinical outcomes. While achieving 100% fT > MIC is biologically plausible and supported by observational data, prospective randomized controlled trials focusing on hard clinical endpoints (like mortality) specifically in the RRT population are still the ‘gold standard’ toward which the field is moving. Additionally, the study focused on two specific antibiotics; while these are workhorses in the ICU, the results cannot be directly extrapolated to other classes like aminoglycosides or glycopeptides.
Conclusion
The SMARRT study provides a definitive framework for dosing meropenem and piperacillin/tazobactam in critically ill patients receiving RRT. By integrating RRT intensity and urine output into clinical decision-making and prioritizing extended or continuous infusions, clinicians can significantly improve the likelihood of therapeutic success. The provided nomograms serve as a practical tool to move away from ‘one-size-fits-all’ dosing and toward a personalized approach that accounts for the dynamic physiology of the critically ill patient.
References
Roberts JA, Ulldemolins M, Liu X, et al. Meropenem and piperacillin/tazobactam optimised dosing regimens for critically ill patients receiving renal replacement therapy. Intensive Care Med. 2025 Sep;51(9):1628-1640. doi: 10.1007/s00134-025-08067-w. Epub 2025 Aug 13. PMID: 40801954.
