Few drugs get looked up as often, or dosed as nervously, as vancomycin. You reach for it constantly for suspected MRSA and serious gram-positive coverage, and it punishes you at both ends: underdose and you risk treatment failure, overdose and you risk the kidneys. The therapeutic window is narrow, and for years the way we monitored it was making the safety problem worse. The 2020 consensus guideline fixed the target we aim for. This is the bedside version of what changed and how to dose it well.

The old 2009 guidance told us to chase troughs of 15 to 20 mg/L as a stand-in for total drug exposure. The problem is that aggressive trough targeting drove nephrotoxicity without actually improving outcomes. Patients were getting hurt to hit a number that was only ever a surrogate. The 2020 ASHP/IDSA/PIDS/SIDP consensus retired trough-only monitoring for serious MRSA infections and moved the target to an AUC/MIC of 400 to 600 mg-h/L, assuming an MIC of 1 mg/L by broth microdilution. AUC/MIC tracks real exposure far better, and it matters at both ends: exposure above roughly 600 is where acute kidney injury risk climbs.

If your patient is seriously ill, load them. Maintenance dosing alone takes too long to reach therapeutic exposure, and the early hours of a serious infection are exactly when you cannot afford to be subtherapeutic. Give 20 to 25 mg/kg based on actual body weight. Note that word: actual. The 2020 guideline moved away from ideal or adjusted body weight for the initial dose. Most institutions cap the loading dose somewhere around 3,000 mg, and you should keep the infusion rate at or below 1 g/hour to avoid infusion reactions.

For a patient with normal renal function, empiric maintenance usually lands at 15 to 20 mg/kg of actual body weight every 8 to 12 hours. That is a starting point, not a final answer. You refine from there using AUC/MIC monitoring, ideally Bayesian, off one or two levels. Many institutions watch a daily ceiling around 4,500 mg. The dose is always individualized, and the sicker or more complex the patient, the earlier you want pharmacy in the loop.

Vancomycin is cleared by the kidneys, so as renal function falls, the interval stretches out. This is where AUC-guided monitoring earns its keep, and where guessing gets people into trouble. Do not forget the other direction either: patients with augmented renal clearance, often young and critically ill, can blow through a standard dose and sit subtherapeutic. Dialysis and impaired renal function are the scenarios where Bayesian software and a pharmacist are not optional extras, they are the standard of care. Specific intervals vary by institution, so anchor to your local protocol.

Treat those as reference ranges, not orders. Your institution protocol governs the specifics, especially loading caps and renal intervals.

Monitor AUC, not just a trough. The preferred approach is Bayesian software using one to two serum concentrations, and for serious infections you want to check early, within the first 24 to 48 hours, rather than waiting for a steady state that may never behave the way the textbook promises. Follow renal function throughout the course so you catch nephrotoxicity while it is still reversible.

Load by actual body weight, titrate maintenance to an AUC/MIC of 400 to 600, stretch the interval as renal function drops, monitor with Bayesian AUC instead of chasing troughs, and pull pharmacy in for the complicated patients. None of this is hard once the target is clear, and getting it right is the difference between a treated infection and an injured kidney. DUK-C keeps the targets and the math at the bedside, so you can dose confidently without stepping away from the patient.