Step by Step Guide to Colistin Dosing and Therapeutic Drug Monitoring
Colistin is often requested when the microbiology report leaves few comfortable options. It may be active against multidrug-resistant Gram-negative bacteria, including some carbapenem-resistant Enterobacterales, Pseudomonas aeruginosa and Acinetobacter baumannii. Yet it is also one of the easiest antibiotics to dose badly.
The main reasons are familiar to anyone who has prescribed it on a difficult ward round. Colistin is given intravenously as colistimethate sodium, a prodrug. Different products describe the amount in different units. Renal function changes quickly in sepsis. Toxicity is real. The therapeutic window is narrow. Therapeutic drug monitoring can help, but only if the samples are taken and interpreted properly.
This guide sets out a practical UK microbiology approach to colistin dosing and therapeutic drug monitoring. It is written for clinical use, but it is informational only. Always follow local Trust policy, the product information, specialist antimicrobial advice, renal pharmacy guidance and the patient’s clinical context.
Why colistin dosing needs extra care
The word “colistin” is often used loosely, but the distinction matters.
Intravenous treatment uses colistimethate sodium, often shortened to CMS. CMS is converted in the body to the active drug, colistin. Microbiology susceptibility testing reports colistin activity, while pharmacy products may display CMS in international units, milligrams of CMS, or colistin base activity.
This creates a common and dangerous source of error.
Unit used | Approximate relationship |
1 million international units of CMS | About 80 mg CMS |
1 million international units of CMS | About 30 mg colistin base activity |
9 million international units of CMS | About 720 mg CMS, or about 270 mg colistin base activity |
These conversions are approximate and should not replace the product information. The key point is simpler: never write “mg colistin” unless the exact meaning is clear. In UK practice, intravenous colistimethate sodium is commonly prescribed in million international units, written as MIU.
Colistin also behaves differently from many other antibiotics. The prodrug takes time to convert to active colistin. Without a loading dose, critically ill patients may spend the first day or more below the intended exposure. At the same time, active colistin is cleared mainly through non-renal pathways, while CMS clearance is strongly affected by kidney function. This is why renal function, timing and monitoring all matter.
Start with a clear indication
Do not begin with the dose. Begin with the reason for treatment.
Colistin should usually be reserved for infections where safer or more effective agents are unsuitable. It may be used when the organism is resistant to standard agents, when allergy limits options, or when susceptibility results support it as part of a planned regimen.
Before prescribing, confirm four points.
The patient has a true infection, not colonisation.
The organism is likely to be covered by colistin.
The infection site is plausible for colistin therapy.
Source control has been addressed or is actively planned.
Examples where the distinction matters include ventilator-associated pneumonia, catheter-associated bloodstream infection, infected collections, complicated urinary infection, and chronic airway colonisation. A positive culture alone is not enough. If the patient is clinically stable and the isolate is from a non-sterile site, review carefully before committing to colistin.
Colistin is rarely a “set and forget” antibiotic. It should trigger a daily review of indication, renal function, microbiology results and toxicity.
Follow a step by step prescribing process
Step 1. Confirm the microbiology result and susceptibility
Check the species, specimen source and susceptibility method. Colistin susceptibility testing can be technically difficult, and unexpected results should be discussed with the laboratory.
Pay close attention to the minimum inhibitory concentration, or MIC, if reported. Colistin exposure targets are usually discussed in relation to the MIC. If the MIC is high, trying to force exposure upwards may increase nephrotoxicity without reliably improving outcome. In that situation, review alternative agents, combination therapy, drainage, device removal and surgical options.
If the isolate is Proteus, Serratia, Morganella, Providencia or Burkholderia cepacia complex, remember that intrinsic resistance to colistin is common or expected. Do not assume activity without expert confirmation.
Step 2. Record baseline patient factors
Before the first dose, collect the information that will change the plan.
Actual body weight, height and an estimate of ideal or adjusted body weight if needed
Serum creatinine and recent trend
Estimated creatinine clearance, preferably for dosing rather than only eGFR
Urine output and haemodynamic status
Renal replacement therapy status
Albumin level if clinically relevant
Infection site and severity
Current and recent nephrotoxic medicines
Baseline neurological symptoms, if present
In critically ill patients, creatinine can lag behind true renal function. A low creatinine in a patient with low muscle mass can be misleading. Rapidly improving renal function can cause underexposure if the dose is not reviewed. Acute kidney injury can cause accumulation if the pre-planned dose continues unchecked.
Step 3. Decide whether intravenous colistin is the right route
Intravenous CMS is used for systemic infection. Nebulised colistin may have a role in selected respiratory cases, usually as adjunctive therapy and according to local respiratory or intensive care guidance. It should not be treated as a simple substitute for intravenous treatment in bloodstream infection or deep-seated sepsis.
For urinary tract infection, colistin can achieve useful urinary concentrations when renal function allows renal excretion of CMS. In severe renal impairment, that advantage may be reduced. Discuss complicated cases early with microbiology and renal pharmacy.
Step 4. Prescribe in million international units
For adult intravenous prescribing in the UK, use MIU where local systems permit.
Write the drug clearly as colistimethate sodium rather than only “colistin”. Include the dose, route, frequency and infusion instructions from local policy. Avoid mixing units in the same prescription.
A safe prescription line should make these points obvious.
The product is colistimethate sodium.
The dose is in MIU.
The route is intravenous.
The dosing interval is stated.
The review plan is documented.
This is one of the most effective ways to prevent ten-fold errors.
Step 5. Give a loading dose when rapid treatment is needed
For serious systemic infection in adults, a commonly used loading dose is 9 MIU intravenously once, subject to local guidance and patient-specific review. The loading dose is usually not reduced simply because renal function is poor, because the purpose is to fill the volume of distribution and reach active concentrations promptly.
There are exceptions. Frailty, very low body weight, recent colistin exposure, unusual risk of toxicity or uncertainty about the diagnosis may change the decision. Still, in septic shock or severe infection caused by a susceptible multidrug-resistant organism, omitting the loading dose is a common cause of early underexposure.
Document the loading dose time. It affects when maintenance dosing and levels are interpreted.
Step 6. Choose the maintenance dose from renal function and illness severity
A commonly used adult regimen for patients with normal renal function is 4.5 MIU every 12 hours after the loading dose, but maintenance dosing must be adjusted for kidney function, renal replacement therapy and local policy.
Use creatinine clearance rather than eGFR when your Trust dosing table specifies it. For extremes of body size, ask pharmacy which body weight to use in the calculation.
Practical principles are:
Normal renal function often needs twice-daily dosing after the loading dose.
Reduced renal function usually needs dose reduction, interval extension, or both.
Augmented renal clearance may lead to underexposure.
Intermittent haemodialysis needs careful post-dialysis planning.
Continuous renal replacement therapy can remove CMS and colistin variably, so standard “renal failure” reductions may underdose.
In critical care, renal function may change within hours. A dose chosen on Monday morning may be wrong by Tuesday.
Step 7. Review the full antimicrobial regimen
Colistin is often used with another active or potentially active agent, especially in severe infection. The partner drug should not be chosen at random. Base it on the organism, susceptibility profile, infection site and patient factors.
Possible partner agents may include a carbapenem, tigecycline, fosfomycin, aminoglycoside, ceftazidime-avibactam, cefiderocol or others, depending on susceptibility and local access. Some combinations are used for stewardship or resistance-suppression reasons, but evidence varies by organism and infection type.
Avoid unnecessary overlapping nephrotoxicity. If an aminoglycoside is used, make the reason explicit and monitor both drugs carefully.
Step 8. Arrange therapeutic drug monitoring at the start
Do not wait until day four to discover that nobody can process the sample.
Before or soon after starting treatment, confirm:
Whether your laboratory offers colistin therapeutic drug monitoring.
Which assay is used and what sample type is required.
Whether samples must be chilled or separated rapidly.
Which times the lab accepts samples.
What information must accompany the request.
Whether the result reports CMS, formed colistin, or an exposure estimate.
This matters because CMS can convert to colistin after blood is taken. Poor sample handling may falsely raise the measured colistin concentration. That can lead to inappropriate dose reduction.
Step 9. Take the right samples at the right time
The best sampling strategy depends on the assay and the dosing model used locally. Many services prefer samples once maintenance therapy has started and early steady exposure can be assessed, often around 24 to 48 hours in serious infection. In unstable renal function, earlier discussion is sensible.
A practical plan may include:
A pre-dose sample just before the next maintenance dose
A post-infusion sample at a locally specified time
Repeat levels after dose changes or renal function changes
A single trough concentration may be easier, but it does not always describe total exposure well. Bayesian or pharmacokinetic interpretation using timed samples is more informative where available.
Label samples with exact times. “Morning level” is not good enough. Record the dose time, infusion duration, sample time, renal function and dialysis status.
Step 10. Adjust the dose using exposure, MIC and toxicity
Therapeutic drug monitoring is not just a number. Interpret it alongside the patient’s response.
Many services aim for an average steady-state plasma colistin concentration around 2 mg/L for susceptible organisms, while avoiding higher exposures linked with nephrotoxicity. Local targets may differ, and the relationship between plasma concentration and lung or tissue exposure is not perfect.
When the result returns, ask five questions.
Was the sample taken and handled correctly?
Has renal function changed since the sample?
Is the organism’s MIC within a range where colistin is likely to help?
Is the patient improving clinically?
Are there signs of kidney injury or neurotoxicity?
If exposure is low and renal function is good or improving, increase the maintenance dose according to local guidance and repeat monitoring. If exposure is high, reduce or hold doses depending on severity, infection status and toxicity. If the target cannot be reached safely, reassess whether colistin remains the right drug.
Monitor toxicity every day
The main toxicity concern is acute kidney injury. Neurotoxicity can also occur, including paraesthesia, dizziness, weakness, confusion and, rarely, neuromuscular blockade. Risk rises with higher exposure, longer duration, critical illness and concurrent nephrotoxins.
Daily review should include:
Serum creatinine and urea
Urine output
Fluid balance
Potassium and magnesium
Neurological symptoms
Current nephrotoxic medicines
Need for ongoing colistin
Common nephrotoxic or kidney-stressing exposures include aminoglycosides, vancomycin, amphotericin B, calcineurin inhibitors, intravenous contrast, dehydration and haemodynamic instability. Do not stop necessary treatment without a plan, but remove avoidable risks.
If acute kidney injury develops, do not simply hold colistin and move on. Decide whether the infection still needs treatment, whether another agent is now available, whether drainage or line removal is incomplete, and whether TDM can guide a safer revised dose.
Handle special situations carefully
Acute kidney injury
Acute kidney injury is common in severe Gram-negative sepsis. It may be due to shock, inflammation, nephrotoxins, obstruction, contrast or the antibiotic itself. The dose should be reviewed whenever creatinine or urine output changes.
If the patient received a loading dose and then develops reduced clearance, the next maintenance dose may need delay or reduction. Use TDM if available. If the patient is deteriorating because of uncontrolled infection, dose reduction alone may not be safe. That is a decision for a senior infection, renal and critical care discussion.
Continuous renal replacement therapy
Continuous renal replacement therapy does not mean “no clearance”. Depending on the modality, filter, flow rates and interruptions, clearance may be substantial or inconsistent. Patients on continuous renal replacement therapy often still need a full loading dose. Maintenance dosing should follow local critical care guidance and TDM whenever possible.
Record the modality and settings on the TDM request. A level taken during uninterrupted treatment may not reflect exposure after a filter change or several hours off circuit.
Intermittent haemodialysis
Intermittent haemodialysis creates peaks and troughs in clearance. Dosing is often planned around dialysis sessions, with supplementary dosing after dialysis when recommended locally. The exact plan should involve renal pharmacy or nephrology, because dialysis timing often changes.
TDM interpretation must include whether the sample was pre-dialysis, post-dialysis or between sessions.
Obesity and low body weight
Colistin dosing is not as simple as using total body weight for every patient. In obesity, using actual body weight without thought can overestimate dose. In very low body weight, standard adult dosing may be excessive. Follow local guidance on ideal or adjusted body weight and involve pharmacy early.
Respiratory infection
Colistin is used in some respiratory infections caused by resistant Gram-negative organisms. Intravenous exposure in lung tissue can be variable. Nebulised therapy may be added in selected ventilated patients, but delivery depends on nebuliser type, ventilator settings and airway factors. It should be prescribed and administered using a local critical care protocol.
Do not assume that a plasma concentration alone proves adequate lung exposure.
Use a simple decision framework for TDM results
The following framework is not a replacement for local policy, but it reflects how results are commonly assessed in practice.
TDM finding | Clinical interpretation | Typical action |
Below local target and patient not improving | Possible underexposure or poor source control | Check timing, renal function, MIC and source control, then consider dose increase |
Below local target but patient improving | Exposure may still be adequate for this case | Discuss before changing, repeat if renal function changes |
Within local target and patient improving | Current regimen is likely acceptable | Continue daily review and repeat levels if treatment continues |
Above local target without toxicity | Risk is increased, especially if renal function worsens | Reduce dose or extend interval according to local guidance |
Above local target with kidney injury or neurotoxicity | Likely harm risk | Hold or reduce, seek urgent specialist advice and review alternatives |
A common mistake is to change the dose before checking whether the sample was valid. If a post-dose sample is mislabelled as a trough, the result may look dangerously high. If the sample sat unprocessed and CMS converted after collection, the apparent colistin level may be misleading.
Avoid common prescribing and monitoring errors
Error | Why it matters | Safer approach |
Prescribing “colistin 300 mg” without specifying the unit basis | The dose may mean CMS or colistin base activity | Prescribe colistimethate sodium in MIU |
Omitting the loading dose in severe infection | Active concentrations may be delayed | Give a documented loading dose unless there is a clear reason not to |
Using a fixed dose despite changing renal function | Exposure may become too low or too high | Review dose daily and after creatinine or urine output changes |
Sending untimed TDM samples | The result may be hard to interpret | Record dose time, infusion time and sample time |
Ignoring dialysis details | Clearance varies widely | Include modality, timing and interruptions |
Treating colonisation as infection | Unnecessary toxicity and resistance pressure | Confirm clinical infection before starting |
Continuing after new susceptibilities return | Safer agents may become available | De-escalate as soon as a better option is clear |
Document the plan clearly
Good documentation prevents errors during handover. A complete plan should include:
Indication and suspected or confirmed organism
Infection site
Date and time of loading dose
Maintenance dose and renal basis
Current renal function and dialysis status
TDM plan and sample timing
Toxicity monitoring plan
Review date or stop date
Criteria for de-escalation
For example, a microbiology note might state that the isolate is susceptible, that the patient has severe ventilator-associated pneumonia, that intravenous colistimethate sodium is being used with a full loading dose, that maintenance dosing follows the Trust renal dosing table, and that levels should be taken at specified times with renal function checked daily.
That level of detail makes overnight prescribing safer.
Know when to stop colistin
The safest colistin course is the shortest effective one.
Review daily for de-escalation. Stop or switch when:
Susceptibility results show a safer effective agent.
The patient has improved and a planned course is complete.
Cultures suggest colonisation rather than infection.
Toxicity outweighs likely benefit.
Source control has resolved the problem and ongoing therapy is unnecessary.
Duration should be based on infection site, source control, immune status and response. Avoid automatic prolonged courses for resistant organisms. Resistance does not always mean longer treatment.
A practical bedside checklist
Before signing the first prescription, run through this checklist.
Confirm infection
Treat the patient, not just the culture result.
Confirm likely activity
Check organism, MIC where available, and intrinsic resistance.
Use the correct name and units
Prescribe colistimethate sodium in MIU.
Give a loading dose when indicated
Common adult severe infection practice uses 9 MIU once, subject to local policy.
Adjust maintenance dosing
Use renal function, dialysis status, body size and Trust guidance.
Plan TDM before samples are due
Confirm timing, handling and reporting arrangements.
Review toxicity daily
Track creatinine, urine output, electrolytes and neurological symptoms.
Reassess the whole treatment plan
Check source control, partner antibiotics and de-escalation options.
Colistin can be life-saving, but only when prescribed with discipline. The step that makes the most difference is not a clever equation. It is a reliable process: clear indication, correct units, early loading dose, renal-adjusted maintenance, valid therapeutic drug monitoring, daily toxicity review and prompt de-escalation when a better option appears.
Comments