Therapeutic Drug Monitoring in the Intensive Care Unit: A Practical Guide

Dr Neeraj Manikath , claude.ai

Abstract

Therapeutic drug monitoring (TDM) is an essential component of precision medicine in the intensive care unit (ICU), where altered pharmacokinetics and pharmacodynamics significantly impact drug exposure and efficacy. This review provides a practical, evidence-based approach to implementing TDM in critically ill patients, highlighting key principles, common pitfalls, and actionable strategies for optimizing drug therapy. We discuss the physiological derangements affecting drug disposition in critical illness, identify medications requiring monitoring, and provide guidance on interpretation and dose adjustment. Clinical pearls and common fallacies are highlighted to enhance the practical utility of this review for postgraduate trainees and critical care practitioners.

Introduction

The critically ill patient presents a moving target for pharmacotherapy. Pathophysiological alterations including altered protein binding, increased volume of distribution (Vd), augmented renal clearance (ARC), hepatic dysfunction, and the use of extracorporeal therapies fundamentally alter drug pharmacokinetics.[1,2] Standard dosing regimens derived from healthy volunteers or stable patients often result in subtherapeutic or toxic drug concentrations in ICU patients. TDM—the measurement of drug concentrations to guide dosing—has evolved from a supportive tool to a cornerstone of individualized therapy in critical care.

Pharmacokinetic Alterations in Critical Illness

Volume of Distribution

Critically ill patients commonly experience increased Vd due to capillary leak, aggressive fluid resuscitation, hypoalbuminemia, and third-spacing.[3] Hydrophilic antibiotics (beta-lactams, aminoglycosides, vancomycin) are particularly affected, often requiring higher loading doses than anticipated.

Pearl: Always calculate loading doses based on actual body weight or adjusted body weight in obese patients. The loading dose is independent of renal or hepatic function—it depends solely on Vd.

Fallacy: "Obese patients always need higher maintenance doses." While loading doses should be weight-based, maintenance doses depend primarily on clearance, which may not increase proportionally with weight.

Renal Function and Augmented Renal Clearance

ARC, defined as creatinine clearance >130 mL/min/1.73m², occurs in 30-65% of critically ill patients, particularly young trauma patients, burn victims, and those with sepsis without established acute kidney injury.[4] This phenomenon leads to enhanced elimination of renally cleared drugs, resulting in subtherapeutic concentrations.

Pearl: Serum creatinine is an unreliable marker of renal function in the ICU. Consider measuring 8- or 24-hour urine creatinine clearance when ARC is suspected. Alternatively, use biomarkers or calculated formulas specifically validated in critical illness.

Oyster: A "normal" or low-normal serum creatinine in a young, muscular trauma patient often masks ARC. These patients may require 2-3 times the standard antibiotic dose.

Protein Binding

Hypoalbuminemia, acute phase reactants, and competitive binding from endogenous substances alter protein binding.[5] For highly protein-bound drugs (phenytoin, valproic acid), total concentrations become misleading, necessitating free (unbound) drug measurement.

Hack: For phenytoin, use the Sheiner-Tozer equation to estimate corrected total phenytoin when albumin <3.2 g/dL: Corrected phenytoin = Measured phenytoin / (0.2 × albumin + 0.1)

However, directly measuring free phenytoin concentration is preferred when available.

Hepatic Dysfunction

Hepatic clearance is unpredictable in critical illness, affected by blood flow, enzyme function, and biliary excretion. Unlike renal function, no single laboratory marker reliably predicts hepatic drug clearance.

Pearl: In patients with significant liver dysfunction, start with reduced doses of hepatically cleared drugs, monitor closely, and titrate based on clinical response and TDM when available.

Indications for Therapeutic Drug Monitoring

TDM is most valuable for drugs with:

Narrow therapeutic indices
Significant interpatient pharmacokinetic variability
Established concentration-effect relationships
Available, rapid, and reliable assays
Potential for significant toxicity

Antimicrobials

Vancomycin

Current guidelines recommend AUC/MIC (area under the curve to minimum inhibitory concentration ratio) targeting 400-600 for serious MRSA infections rather than trough-based monitoring.[6] However, practical implementation varies.

Best Practice Approach:

Loading dose: 25-30 mg/kg actual body weight
Target AUC₀₋₂₄: 400-600 mg·h/L
Use Bayesian software or two-level sampling (peak and trough) for AUC estimation
If using trough-only monitoring: Target 15-20 mg/L for serious infections, but recognize this is a surrogate marker

Fallacy: "Vancomycin troughs of 15-20 are always necessary." This dogma has been challenged. For many infections, AUC/MIC of 400 may suffice, and aggressive trough targeting increases nephrotoxicity without proven benefit.[7]

Hack: If Bayesian software is unavailable, obtain levels at 1-hour post-infusion and at trough. Use these with pharmacokinetic equations or online calculators to estimate AUC.

Aminoglycosides

Once-daily dosing is preferred in most ICU patients (7 mg/kg for gentamicin/tobramycin; 15-20 mg/kg for amikacin).[8]

Monitoring Strategy:

Check random level 6-14 hours after first dose
Use the Hartford nomogram or institution-specific protocol
Target peak concentrations: gentamicin/tobramycin 20-30 mg/L, amikacin 60-80 mg/L
Target trough <1 mg/L (gentamicin/tobramycin) or <5 mg/L (amikacin)

Pearl: In patients with ARC, extended-interval dosing may fail. Consider dividing doses or increasing frequency based on levels.

Oyster: In patients receiving continuous renal replacement therapy (CRRT), aminoglycoside clearance is highly variable and protocol-driven approaches often fail. Individual TDM is essential.

Beta-Lactams

Emerging evidence supports TDM for beta-lactams in critically ill patients.[9] Time above MIC (T>MIC) is the key pharmacodynamic parameter.

Target Concentrations:

For severe infections: Target free drug concentration >4× MIC for 100% of the dosing interval
For difficult-to-treat organisms: Consider continuous or extended infusions

Pearl: Beta-lactam TDM is increasingly available and should be considered in patients with:

Life-threatening infections (endocarditis, meningitis, osteomyelitis)
Infections with less susceptible organisms
Altered pharmacokinetics (ARC, CRRT, obesity, burns)
Clinical failure despite appropriate antimicrobial selection

Hack: When TDM is unavailable, consider empiric extended infusions (3-4 hours) or continuous infusions for critically ill patients with severe sepsis or altered pharmacokinetics.

Antiepileptics

Phenytoin

Total therapeutic range: 10-20 mg/L; Free therapeutic range: 1-2 mg/L

Critical Points:

Non-linear (Michaelis-Menten) kinetics: Small dose increases can cause disproportionate concentration increases
Free fraction increases significantly in hypoalbuminemia, uremia, and critical illness
Always order free levels in ICU patients

Fallacy: "Phenytoin is safe because it's been used for decades." Phenytoin has a narrow therapeutic index, complex kinetics, and significant drug interactions, making it challenging in the ICU. Consider alternatives (levetiracetam, valproic acid) for acute seizure management when appropriate.

Valproic Acid

Total therapeutic range: 50-100 mg/L; Free therapeutic range: 5-10 mg/L

Similar to phenytoin, valproic acid is highly protein-bound and affected by hypoalbuminemia.

Pearl: For status epilepticus, higher concentrations (100-150 mg/L total) may be required. Monitor free levels if available.

Immunosuppressants

Tacrolimus, cyclosporine, mycophenolate, and sirolimus require meticulous monitoring in critically ill transplant recipients.

Key Principles:

Drug interactions are extensive (azoles, antibiotics, proton pump inhibitors)
Target ranges vary by organ, time post-transplant, and indication
Trough levels are standard, but C2 (2-hour post-dose) may be monitored for cyclosporine
Always check levels 3-5 days after dose adjustment or drug interactions

Oyster: In patients on CRRT or ECMO, immunosuppressant levels can be unpredictable due to drug adsorption to circuits. Increase monitoring frequency.

Antiarrhythmics

Digoxin

Therapeutic range: 0.5-0.9 ng/mL (heart failure); 0.8-2.0 ng/mL (atrial fibrillation rate control)

Critical Considerations:

Reduced volume of distribution in elderly patients
Significant drug interactions (amiodarone, verapamil, quinidine)
Toxicity risk increased by hypokalemia, hypomagnesemia, hypercalcemia, and hypothyroidism
Check levels at least 6 hours after dose (12-24 hours is preferred for steady-state)

Fallacy: "A digoxin level within the therapeutic range excludes toxicity." Toxicity can occur at therapeutic or even subtherapeutic levels in the presence of electrolyte abnormalities or drug interactions.

Amiodarone

Therapeutic range: 1.0-2.5 mg/L

Pearl: Amiodarone has a very long half-life (up to 100 days) and extensive tissue distribution. It takes weeks to months to reach steady state. Loading strategies are essential for acute use, and toxicity can persist long after discontinuation.

Lithium

Therapeutic range: 0.6-1.2 mEq/L (acute treatment); 0.4-0.8 mEq/L (maintenance)

Critical in the ICU:

Narrow therapeutic index with potentially fatal toxicity
Entirely renally eliminated; adjust for renal function
Dehydration, NSAIDs, ACE inhibitors, and diuretics increase levels
Check levels 12 hours post-dose

Pearl: In acute lithium toxicity, a single level is insufficient. Check serial levels as redistribution from tissues can cause rebound after initial clearance.

Sampling Considerations

Timing

Incorrect sampling timing is a leading cause of TDM misinterpretation.

General Principles:

Trough levels: Immediately before the next dose (within 30 minutes)
Peak levels: Timing varies by drug and infusion duration
Steady state: Generally achieved after 4-5 half-lives
Loading doses: Can check levels sooner to verify achievement of target concentration

Hack: Create an ICU-specific TDM ordering guide with timing recommendations to standardize practice and reduce errors.

Sample Collection

Best Practices:

Avoid drawing from lines used for drug administration (wait 2-3 hours if necessary)
Discard the first 5-10 mL if drawing from a line
Specify sampling time clearly in orders and documentation
Communicate with laboratory about urgent processing needs

Fallacy: "Any blood draw will do." Improper sample collection is a major source of erroneous results, leading to inappropriate dose adjustments.

Dose Adjustment Strategies

Pharmacokinetic Principles

Two parameters govern maintenance dosing:

Clearance (CL): Determines the dose rate needed to achieve steady-state concentration
Volume of distribution (Vd): Determines loading dose and concentration fluctuation

First-Order Kinetics:

Maintenance dose = Target Css × CL

Where Css = steady-state concentration

Pearl: When concentration is below target, calculate the proportional increase needed:

New dose = Current dose × (Target concentration / Measured concentration)

Then round to practical doses and consider pharmacokinetic changes that may have occurred.

Bayesian Dosing Software

Computer-assisted dosing using Bayesian algorithms integrates:

Population pharmacokinetic parameters
Patient-specific covariates (weight, renal function, age)
Measured drug concentrations
Dosing history

Advantages:

Improved accuracy compared to nomograms
Can accommodate complex dosing histories
Provides confidence intervals
Useful for drugs with complex kinetics (vancomycin AUC estimation)

Available Tools:

DoseMeRx
InsightRX
PrecisePK
MwPharm

Pearl: Bayesian software is most valuable when measurements deviate significantly from expected values or when dosing history is complex. For straightforward cases, clinical judgment and simple calculations often suffice.

Special Populations

Obesity

Use adjusted body weight for renally eliminated drugs: ABW = IBW + 0.4(TBW - IBW)
Consider actual body weight for loading doses of lipophilic drugs
Increased monitoring frequency may be needed due to unpredictable pharmacokinetics

Renal Replacement Therapy

Drug removal during RRT depends on:

Membrane characteristics (high-flux vs. low-flux)
Modality (intermittent HD vs. CRRT)
Drug properties (molecular weight, protein binding, Vd)

Key Points:

Continuous therapies: Provide steady clearance; dose as for moderate renal impairment (CrCl 30-50) as a starting point
Intermittent HD: Dose post-dialysis for dialyzable drugs
CRRT effluent rates: Higher rates increase clearance
TDM is essential—predictive equations are unreliable

Hack: Create institution-specific protocols for common drugs during CRRT based on local practices and typical effluent rates.

Extracorporeal Membrane Oxygenation (ECMO)

Drug pharmacokinetics are profoundly altered by:

Drug sequestration in circuit (particularly lipophilic drugs)
Increased Vd
Altered protein binding
Potential for hemolysis affecting drug measurement

Pearl: For patients on ECMO, assume standard dosing will be inadequate. Increase loading doses and monitoring frequency, particularly in the first 72 hours.

Common Pitfalls and How to Avoid Them

Pitfall 1: Assuming Steady State

Problem: Checking levels before steady state is reached leads to misinterpretation.

Solution: Calculate time to steady state (4-5 half-lives). For drugs with long half-lives (amiodarone, phenobarbital), steady-state monitoring is impractical; use clinical response and serial levels.

Pitfall 2: Ignoring Free vs. Total Concentrations

Problem: Relying on total concentrations for highly protein-bound drugs in patients with hypoalbuminemia.

Solution: Measure free concentrations for phenytoin and valproic acid in ICU patients. Correct total concentrations only when free levels are unavailable.

Pitfall 3: Single-Point Problem-Solving

Problem: Making dramatic dose changes based on a single aberrant level without considering clinical context.

Solution: Verify unexpected results. Consider:

Was sampling timed correctly?
Were there recent dose changes?
Are there new drug interactions?
Has renal/hepatic function changed?

Pitfall 4: Cookbook Dosing in Dynamic Patients

Problem: Using fixed nomograms without considering individual patient trajectories.

Solution: Reassess pharmacokinetic assumptions regularly. A patient transitioning from ARC to AKI requires completely different dosing.

Pitfall 5: Ignoring Pharmacodynamic Monitoring

Problem: Focusing solely on concentrations while ignoring clinical response and toxicity.

Solution: TDM guides dosing but doesn't replace clinical assessment. Monitor for:

Efficacy endpoints (infection resolution, seizure control)
Toxicity markers (nephrotoxicity, ototoxicity, neurotoxicity)
Biomarkers when available (procalcitonin for antibiotics)

Emerging Concepts and Future Directions

Precision Dosing and Pharmacogenomics

Genetic polymorphisms affect drug metabolism (CYP enzymes, drug transporters). While not yet routine, pharmacogenomic testing may become standard for select drugs (warfarin, clopidogrel, thiopurines).

Point-of-Care Testing

Rapid TDM at the bedside could enable real-time dose adjustments. Technologies under development include biosensors and microfluidic devices, though widespread implementation remains limited by cost and regulatory challenges.

Model-Informed Precision Dosing (MIPD)

Integration of population pharmacokinetic models, machine learning, and real-time patient data may optimize dosing decisions. Clinical decision support systems incorporating MIPD show promise in early studies.[10]

Practical Implementation: Building an ICU TDM Program

Essential Components

Written protocols: Standardized approaches for commonly monitored drugs
Education: Regular teaching for physicians, pharmacists, and nurses
Clinical pharmacy integration: Dedicated ICU pharmacists with TDM expertise
Laboratory coordination: Reliable turnaround times and after-hours availability
Documentation systems: Electronic order sets with timing prompts
Quality assurance: Regular audit of TDM practices and outcomes

The Role of the Clinical Pharmacist

Clinical pharmacists are invaluable in ICU TDM programs:

Reviewing levels daily and recommending dose adjustments
Identifying drug interactions
Calculating patient-specific pharmacokinetic parameters
Using Bayesian software for complex dosing
Educating the team on TDM principles

Pearl: Establish regular interdisciplinary rounds including pharmacists. Studies consistently show improved outcomes, reduced toxicity, and cost savings when pharmacists are integrated into ICU teams.[11]

Conclusion

Therapeutic drug monitoring is both an art and a science, requiring integration of pharmacokinetic principles, clinical judgment, and individualized patient assessment. The critically ill patient challenges our assumptions about drug dosing at every turn. Success requires vigilance, humility, and adaptability. By understanding the physiological derangements that alter pharmacokinetics, recognizing when TDM is indicated, sampling appropriately, and interpreting results in clinical context, we can harness TDM as a powerful tool to optimize outcomes while minimizing toxicity.

The future of ICU pharmacotherapy lies in precision medicine—using all available data to deliver the right drug at the right dose to the right patient at the right time. TDM, enhanced by emerging technologies and decision support systems, will remain central to this endeavor.

Key Takeaways

Loading doses depend on Vd; maintenance doses depend on clearance
Serum creatinine is unreliable in critical illness; suspect ARC in young patients with "normal" creatinine
Always measure free concentrations for highly protein-bound drugs in ICU patients
Incorrect sampling timing is the most common cause of TDM errors
Bayesian software improves dosing accuracy for complex drugs like vancomycin
Clinical context trumps isolated concentration values
Integrate clinical pharmacists into ICU teams for optimal TDM implementation

References

Roberts JA, Abdul-Aziz MH, Lipman J, et al. Individualised antibiotic dosing for patients who are critically ill: challenges and potential solutions. Lancet Infect Dis. 2014;14(6):498-509.
Blot SI, Pea F, Lipman J. The effect of pathophysiology on pharmacokinetics in the critically ill patient—concepts appraised by the example of antimicrobial agents. Adv Drug Deliv Rev. 2014;77:3-11.
Udy AA, Roberts JA, Lipman J. Implications of augmented renal clearance in critically ill patients. Nat Rev Nephrol. 2011;7(9):539-543.
Bilbao-Meseguer I, Rodríguez-Gascón A, Barrasa H, et al. Augmented renal clearance in critically ill patients: a systematic review. Clin Pharmacokinet. 2018;57(9):1107-1121.
Ulldemolins M, Roberts JA, Rello J, Paterson DL, Lipman J. The effects of hypoalbuminaemia on optimizing antibacterial dosing in critically ill patients. Clin Pharmacokinet. 2011;50(2):99-110.
Rybak MJ, Le J, Lodise TP, et al. Therapeutic monitoring of vancomycin for serious methicillin-resistant Staphylococcus aureus infections: a revised consensus guideline and review by the American Society of Health-System Pharmacists, the Infectious Diseases Society of America, the Pediatric Infectious Diseases Society, and the Society of Infectious Diseases Pharmacists. Am J Health Syst Pharm. 2020;77(11):835-864.
Aljefri DM, Avedissian SN, Rhodes NJ, et al. Vancomycin area under the curve and acute kidney injury: a meta-analysis. Clin Infect Dis. 2019;69(11):1881-1887.
Nicolau DP, Freeman CD, Belliveau PP, et al. Experience with a once-daily aminoglycoside program administered to 2,184 adult patients. Antimicrob Agents Chemother. 1995;39(3):650-655.
Abdul-Aziz MH, Alffenaar JC, Bassetti M, et al. Antimicrobial therapeutic drug monitoring in critically ill adult patients: a position paper. Intensive Care Med. 2020;46(6):1127-1153.
Darwich AS, Ogungbenro K, Vinks AA, et al. Why has model-informed precision dosing not yet become common clinical reality? Lessons from the past and a roadmap for the future. Clin Pharmacol Ther. 2017;101(5):646-656.
MacLaren R, Bond CA, Martin SJ, Fike D. Clinical and economic outcomes of involving pharmacists in the direct care of critically ill patients with infections. Crit Care Med. 2008;36(12):3184-3189.

Conflicts of Interest: None declared.

Funding: No external funding received for this review.

Friday, December 26, 2025