Tci pumps: marsh vs schnider models

At-a-glance comparison (what differs clinically)

Both are pharmacokinetic (PK) models used by TCI pumps to achieve a chosen target concentration by calculating infusion rates using a multi-compartment model.
- They do not measure drug concentration; they estimate it from population PK parameters.
Marsh model: targets and displays plasma concentration (Cp) in most implementations; weight-based scaling; no age term.
- Commonly used for propofol TCI in the UK historically; simple inputs (often weight only, plus sex depending on pump).
Schnider model: can target effect-site concentration (Ce) directly; uses age, height, weight, sex to derive lean body mass (LBM); includes age-related changes in parameters.
- Often preferred for effect-site targeting and smoother titration to clinical effect in adults.
Key practical implication: for the same clinical endpoint, Marsh Cp targets and Schnider Ce targets are not numerically interchangeable.
- Example: A Schnider Ce target of ~3–4 µg/mL may correspond to a higher initial Cp during the bolus phase; the pump handles this automatically but the displayed numbers differ by model/target type.

How a TCI pump achieves the target (core mechanism)

Pump gives an initial bolus (or high initial infusion) to fill the central compartment, then reduces to maintain target and compensate for distribution and clearance.
- Subsequent rate changes reflect predicted redistribution to peripheral compartments and metabolic clearance.
Effect-site targeting adds a ke0 link model between plasma and effect site; the pump may transiently overshoot Cp to drive Ce to target quickly.
- This is why Ce targeting can produce a larger initial bolus than Cp targeting for the same displayed target value.

Choosing a model/target in practice (adult propofol)

Induction: Schnider Ce targeting is commonly used (e.g., start 3–4 µg/mL and titrate) because it aligns with clinical effect and avoids chasing a delayed effect.
- Marsh Cp targeting can be used effectively but requires awareness of hysteresis (delay between Cp and effect).
Maintenance: either model can be used; titrate to effect (clinical signs ± processed EEG) and reduce for age/frailty, co-administered opioids, and regional techniques.
Obesity: Schnider uses LBM and may avoid extreme dosing based purely on total body weight; Marsh weight-based scaling can risk excessive bolus/infusion if total body weight is entered without thought.
- Always check the pump’s allowed weight/height range and whether it expects total body weight; follow local policy/manufacturer guidance.

Definitions and concepts examinable in FRCA

TCI (target-controlled infusion): a computer-controlled infusion where the user sets a target concentration and the device calculates infusion rates using a PK model.
Cp vs Ce: plasma concentration (Cp) is the modelled central compartment concentration; effect-site concentration (Ce) is the modelled biophase concentration linked to effect by ke0.
Context-sensitive decrement time: time for concentration to fall by a given percentage after stopping an infusion; depends on duration (context) and model parameters.
Hysteresis: delay between Cp and clinical effect due to distribution to effect site; Ce targeting aims to account for this.

Marsh model (propofol) — key points

Structure: 3-compartment PK model; parameters scaled primarily to body weight.
Inputs (typical): weight (and sometimes sex depending on device); does not explicitly include age or height in the classic implementation.
Targeting: classically plasma targeting (Cp). Some pumps may allow effect-site targeting using Marsh PK with an added ke0, but the classic comparison is Marsh Cp vs Schnider Ce.
Clinical behaviour: for a given displayed target Cp, initial bolus and early infusion are determined by central volume (V1) and clearances; because scaling is weight-based, large weights can produce large boluses.
Strengths: simplicity; long-standing familiarity; robust for average-sized adults when used with appropriate clinical titration.
Limitations: less individualisation for age/frailty and body composition; Cp target may lag behind effect (hysteresis) leading to overshoot if titrated too quickly.

Schnider model (propofol) — key points

Structure: 3-compartment PK model with covariates; incorporates age-related changes and uses LBM rather than total body weight for key parameters.
Inputs: age, sex, height, weight (to calculate LBM); more patient-specific than Marsh.
Targeting: commonly used with effect-site targeting (Ce), which accounts for plasma–effect site delay using ke0.
Clinical behaviour: Ce targeting may deliver a higher initial Cp (overshoot) to reach the desired Ce quickly; subsequent infusion adapts as Ce approaches target.
Strengths: improved titration to effect in adults; incorporates age and body composition; often more conservative dosing in obesity compared with total-body-weight scaling.
Limitations: requires accurate height/weight/age entry; LBM formula can behave unexpectedly at extremes (very obese/very short/very tall), and model validity is limited outside studied ranges.

Practical pump considerations (equipment focus)

User must select: drug concentration (e.g., propofol 1% vs 2%), model (Marsh vs Schnider), and target type (Cp vs Ce if available).
Data entry errors (wrong weight, height, age, or drug concentration) can cause major dosing error because the pump assumes inputs are correct.
Line setup: use anti-siphon/anti-reflux valves as per local policy; ensure dedicated line where possible; minimise dead space and be aware of bolus from flushing.
When changing targets: allow time for effect-site equilibration (especially if targeting Cp) before making repeated rapid increases.
Co-administered drugs: opioids, benzodiazepines, volatile agents, and regional anaesthesia reduce propofol requirement; models do not account for pharmacodynamic synergy.

Typical adult target ranges (contextual, titrate to effect)

Propofol TCI induction/maintenance targets vary widely with age, opioids, and stimulus; use local practice and clinical endpoints rather than fixed numbers.
Common ballpark: Schnider Ce ~2–6 µg/mL (lower in elderly/frail; higher for very stimulating phases) and Marsh Cp often in a similar numeric range but not directly comparable.

Explain what a TCI pump does and what it does not do.

Core points expected in an equipment/pharmacology viva.

User selects a target concentration (Cp or Ce depending on model/pump); pump calculates infusion rates using a population PK model to achieve/maintain that target.
It does not measure drug concentration; it assumes the patient behaves like the model population and that the entered demographics and drug concentration are correct.
It does not account for pharmacodynamic variability or drug interactions; clinician must titrate to effect and monitor depth of anaesthesia and physiology.

Define Cp and Ce and describe hysteresis in the context of propofol TCI.

Often asked to test understanding of why effect-site targeting exists.

Cp: modelled plasma/central compartment concentration; Ce: modelled effect-site (biophase) concentration linked to effect.
Hysteresis: for a given Cp, effect changes with time because drug takes time to equilibrate between plasma and effect site.
Ce targeting uses a ke0 link model to account for this delay, aiming to match the displayed target more closely to clinical effect.

Compare Marsh and Schnider models: what patient variables do they use and why does it matter?

A common FRCA comparison question.

Marsh: primarily weight-based scaling; classic model does not explicitly include age or height.
Schnider: uses age, sex, height, weight to compute LBM and incorporates age-related parameter changes.
Matters because extremes of age/body composition can lead to different predicted volumes/clearances and therefore different bolus/infusion requirements and risk of over/underdosing.

Why can Schnider effect-site targeting produce a larger initial bolus than plasma targeting?

Tests understanding of the ke0 link and pump behaviour.

To raise Ce quickly, the pump may transiently raise Cp above the eventual steady-state Cp (an intentional Cp overshoot) because Ce lags behind Cp.
As Ce approaches target, the pump reduces infusion to avoid excessive Ce once equilibration occurs.

A patient is 78 years old. How might Marsh vs Schnider differ in dosing behaviour and what would you do clinically?

Assesses safe application rather than exact numbers.

Schnider includes age as a covariate and often predicts lower requirements; Marsh does not explicitly adjust for age, so a weight-based approach may overestimate dose if not manually reduced.
Clinically: start with a lower target, increase slowly allowing time for effect-site equilibration, use opioid-sparing where appropriate, and monitor closely (BP, EEG-based depth if used).

An obese patient: what are the risks of entering total body weight in Marsh, and how does Schnider attempt to address this?

Common safety-themed viva.

Marsh scales parameters to body weight; entering a very high total body weight can lead to a large calculated bolus and high early infusion rates, risking hypotension and excessive depth.
Schnider uses LBM derived from height/weight/sex, which may reduce the tendency to dose purely to total body weight (though model validity still limited at extremes).
Clinically: consider lower starting targets, slow titration, careful haemodynamic management, and ensure entered demographics are accurate; follow local guidance for obesity and TCI.

What are the main sources of error with TCI pumps (equipment + human factors)?

Often asked in equipment stations and safety discussions.

Wrong model/target type selected (e.g., Cp vs Ce confusion) leading to unexpected bolus/infusion behaviour.
Incorrect patient data entry (weight/height/age/sex) or wrong drug concentration selected (1% vs 2%).
Infusion line issues: occlusion, disconnection, siphoning, backflow, dead-space bolus after flushing, non-dedicated line with variable carrier flow.
Clinical: assuming the displayed number equals the patient’s true concentration/effect; failing to adjust for co-administered drugs and physiological changes (shock, low CO).

If the pump is targeting Cp, why might repeated rapid increases in target cause overshoot of clinical effect?

Tests understanding of equilibration delay.

Effect lags behind Cp (hysteresis). If you increase Cp targets rapidly before the effect-site has equilibrated, you may keep escalating based on apparent under-effect, then Ce catches up and the patient becomes too deep.
Practical response: make incremental changes, wait for effect-site equilibration, and use clinical/processed EEG guidance if available.

Describe what happens to infusion rates when you stop a propofol TCI and what determines wake-up time.

Links TCI to context-sensitive decrement concepts.

On stopping, infusion rate becomes zero but drug redistributes and is cleared; Cp and Ce fall according to the model’s clearances and compartment sizes.
Wake-up depends on effect-site concentration falling below a patient-specific threshold, influenced by duration of infusion (context), age, co-drugs (opioids), and physiological state.

Give a structured explanation of why Marsh and Schnider target numbers should not be directly compared.

A frequent conceptual pitfall tested in orals.

They are different PK parameter sets derived from different datasets and covariate structures; therefore the same numeric target does not imply the same predicted dosing history or concentrations over time.
They may target different compartments (Cp vs Ce). Even if both display µg/mL, the site differs and so does the time-course.
Clinical approach: use the chosen model consistently, titrate to effect, and avoid cross-model ‘equivalence tables’ as a substitute for monitoring.

You are asked to set up a propofol TCI for a 65-year-old, 170 cm, 95 kg male. Talk through your setup and how model choice (Marsh vs Schnider) changes what you enter and what you expect to see.

Key is safe setup + demonstrating understanding of inputs/outputs and expected behaviour.

Checks: correct patient, allergies, IV access, monitoring, resus drugs/airway ready; confirm propofol concentration (1% vs 2%) and syringe size/brand compatibility with pump.
Model selection: Marsh typically requires weight entry and targets Cp; Schnider requires age/sex/height/weight and commonly targets Ce.
Expectations: Schnider Ce may deliver a more pronounced initial bolus (Cp overshoot) to reach Ce; Marsh Cp changes may show delayed clinical effect if titrated quickly.
Administration: start with conservative target, titrate in steps, allow time for equilibration, manage haemodynamics, and adjust for opioids/adjuncts.

Discuss the limitations of TCI models and how you would mitigate risk when using them.

This is often examined as ‘limitations + safety strategies’.

Population model mismatch: extremes of age, obesity, pregnancy, severe illness (low cardiac output, hepatic failure) may not fit the model well.
Pharmacodynamic variability and synergy not modelled: opioids and other sedatives reduce propofol requirement; stimulus varies over time.
Mitigation: careful patient selection, conservative starting targets, incremental changes, close monitoring (including depth monitoring where appropriate), and readiness to switch to manual infusion/bolus if needed.
Equipment mitigation: double-check entries (weight/height/age/sex, concentration), use anti-siphon measures, dedicated line, and document model/target type.

A patient becomes hypotensive shortly after starting propofol TCI. How do you analyse whether this is model-related and what immediate actions do you take?

Assesses prioritisation and understanding of bolus/overshoot behaviour.

Immediate actions: assess ABC, reduce/stop propofol temporarily, treat hypotension (fluids/vasopressor), check depth and other causes (bleeding, anaphylaxis, arrhythmia).
Model-related considerations: effect-site targeting can deliver a larger initial bolus; Marsh with high entered weight can also produce large bolus/infusion.
Check pump: correct concentration, correct model/target type, correct demographics, line patency and whether a flush/line change delivered an unintended bolus.