Infusion Devices (Pumps, Syringes)

Foundations and mechanisms

Purpose and fundamental physics of infusion

Infusion devices deliver a prescribed volume or mass of drug over time, replacing the imprecision of gravity flow with controlled displacement, pressure generation and alarmed monitoring. For anaesthesia, their relevance is greatest for potent drugs with narrow therapeutic indices—vasopressors, insulin, opioids, neuromuscular blockers, sedatives and total intravenous anaesthesia. A seemingly small volumetric error may be clinically important: noradrenaline 8 mg in 50 mL gives 160 micrograms/mL, so an unintended 1 mL bolus delivers 160 micrograms, enough to cause marked hypertension or arrhythmia.

Flow through an infusion system is governed by pressure gradient and resistance. For laminar flow in a rigid cylindrical tube:

Q = ΔPπr⁴ / 8ηl

Thus flow is proportional to the fourth power of cannula radius and inversely proportional to viscosity and length. Clinically, this explains why a short, wide-bore cannula is superior for rapid fluid administration, and why syringe pumps delivering via long, narrow, compliant tubing may have delayed onset at low rates. Typical peripheral venous pressure is 5–15 mmHg; central venous pressure is approximately 2–8 mmHg. Infusion pumps commonly generate pressures far exceeding venous pressure, often with occlusion alarm thresholds in the range 300–1000 mmHg depending on device and setting.

Classification of infusion devices

Device type	Primary mechanism	Typical accuracy	Common anaesthetic use	Key limitations
Gravity infusion	Hydrostatic pressure; roller clamp alters resistance	Poor; affected by height, cannula resistance and venous pressure	Maintenance crystalloid where precision is not critical	No true flow control, no occlusion detection, free-flow risk
Volumetric pump	Peristaltic fingers, cassette piston or diaphragm displacing fluid from a bag	Usually ±5%; some modern devices ±3%	Fluids, blood, antibiotics, parenteral nutrition	Requires compatible giving set; risk of free flow if anti-free-flow clamp bypassed
Syringe driver	Stepper motor advances lead screw and syringe plunger	Often ±2–3% once steady state achieved	Vasoactive drugs, TIVA, paediatric infusions	Start-up delay, compliance effects, syringe size dependence, bolus after occlusion release
Patient-controlled analgesia	Programmable syringe or cassette pump with lockout logic	Device dependent, generally similar to volumetric/syringe systems	Postoperative opioid delivery	Programming error, wrong drug concentration, respiratory depression
Target-controlled infusion	Microprocessor-driven variable infusion using pharmacokinetic model	Depends on pump accuracy and model performance	Propofol and remifentanil anaesthesia	Population model error; not a closed-loop measure of effect

Syringe pump mechanics and performance

A syringe pump converts rotary stepper motor motion into linear plunger displacement. The delivered flow depends on plunger travel and syringe internal cross-sectional area. A 50 mL syringe has a larger barrel diameter than a 10 mL syringe; therefore the same linear plunger error produces a larger volume error in the 50 mL syringe. Conversely, at very low rates, smaller syringes often improve start-up responsiveness because less plunger movement is required to generate clinically relevant flow and the mechanical system has less dead volume.

Important performance concepts include accuracy, precision, continuity of flow and response time. At low rates such as 0.1–1 mL/h, delivery may be pulsatile because discrete motor steps are separated by long intervals. This is particularly relevant for neonates, vasoactive infusions and concentrated anaesthetic drugs. Start-up delay is caused by mechanical backlash, stiction of the plunger, syringe compliance and expansion of tubing before pressure equilibrates. At 1 mL/h through compliant extension tubing, meaningful delivery may be delayed for minutes, especially if the system is not primed under pressure.

Pressure, compliance, occlusion and bolus phenomena

Infusion systems are not rigid. Pressure generated against downstream resistance distends tubing, compresses gas bubbles and deforms the syringe. The stored elastic energy becomes clinically important during occlusion. If an infusion continues against an occluded cannula until the alarm threshold is reached, the fluid accumulated in the compliant system may be delivered as an inadvertent bolus when the obstruction is relieved. The bolus volume depends on occlusion pressure, compliance of syringe and tubing, presence of air, and alarm delay. Air is highly compressible and markedly increases stored volume; meticulous priming is therefore a safety as well as an accuracy issue.

• Occlusion pressure limits: commonly selectable as low, medium or high; approximate ranges are 300–500 mmHg for low settings and up to 900–1000 mmHg for high settings.
• Alarm delay: increases at low flow rates and with compliant tubing; a partial occlusion at 1 mL/h may take substantially longer to alarm than at 50 mL/h.
• Post-occlusion bolus: clinically relevant for catecholamines, insulin, opioids and local anaesthetic infusions.
• Siphonage/free flow: occurs when hydrostatic pressure or gravity permits uncontrolled flow, particularly if the syringe is positioned above the patient or if anti-siphon valves/clamps fail.

Target-controlled infusion and pharmacokinetic foundations

TCI pumps use compartmental pharmacokinetic models to calculate a time-varying infusion profile: an initial loading bolus to fill the central compartment, a declining infusion to replace redistribution, and a maintenance component to match elimination. They do not measure plasma or effect-site concentration. Common models include Marsh and Schnider for propofol, and Minto for remifentanil. Propofol has a context-sensitive half-time that increases with infusion duration, whereas remifentanil has organ-independent esterase metabolism and a context-sensitive half-time of approximately 3–5 minutes, even after prolonged infusion.

Drug/model context	Typical adult target range	Key examination point
Propofol TCI	Plasma/effect-site target commonly 2–6 micrograms/mL	Model performance affected by age, weight, cardiac output and co-administered opioids
Remifentanil TCI	Effect-site target commonly 2–8 nanograms/mL	Rapid offset; no residual analgesia after stopping
Noradrenaline infusion	Often 0.02–1 microgram/kg/min clinically	Dead space and line changes can abruptly alter delivered dose

For Primary FRCA purposes, infusion devices should be understood as electromechanical pressure generators constrained by fluid mechanics, compliance, dead space, software logic and human factors. Many serious incidents arise not from motor failure but from concentration error, wrong syringe selection, line misconnection, unrecognised occlusion, or uncontrolled free flow.

Clinical assessment and investigations

Clinical presentation of infusion-device related incidents

Infusion pump and syringe driver problems usually present as a drug effect inconsistent with the prescribed infusion, an alarm state, or unexpected fluid balance discrepancy. In anaesthesia and critical care the highest-risk presentations are abrupt haemodynamic change during vasoactive infusions, inadequate hypnosis during total intravenous anaesthesia, opioid or sedative toxicity, insulin-related dysglycaemia, local anaesthetic toxicity, anticoagulant misdelivery, and extravasation of vasoactive or cytotoxic drugs. Assessment must distinguish patient pathology from infusion system error; both may coexist.

Presentation	Infusion-related differential	Key clinical interpretation
Hypotension, bradycardia, low cardiac output	Excess propofol, remifentanil, epidural/local anaesthetic, vasodilator overdose; cessation of noradrenaline/adrenaline	Correlate timing with syringe change, line disconnection, occlusion release, pump programme change, bolus delivery or wrong concentration.
Hypertension, tachycardia, sweating, movement, awareness signs	Underdelivery of hypnotic/analgesic during TIVA; empty syringe; occluded cannula; wrong line; TCI target not achieved	Particularly important because propofol has rapid redistribution; remifentanil offset half-time is approximately 3–5 min.
Respiratory depression, miosis, reduced consciousness	Opioid PCA overdose, basal infusion error, lockout failure, wrong drug in syringe	Concerning thresholds include respiratory rate <8 min-1, SpO2 fall, sedation score 3–4, or rising PaCO2.
Seizures, arrhythmias, circumoral tingling, metallic taste	Local anaesthetic systemic toxicity from epidural, wound catheter or regional infusion	Clinical diagnosis; plasma levels are rarely timely. Toxicity risk rises with intravascular delivery and cumulative dosing.
Severe hypo-/hyperglycaemia	Insulin infusion wrong concentration, pump rate error, line interruption, dextrose co-infusion failure	Check capillary/arterial glucose urgently; ICU targets commonly 6–10 mmol L-1, with hypoglycaemia <4.0 mmol L-1.
Swelling, pain, blanching, tissue necrosis	Extravasation, especially vasopressors, calcium, potassium, hyperosmolar infusions	Peripheral vasopressor extravasation may occur despite apparently normal pump function because pumps sense pressure, not tissue injury.

Structured bedside assessment

1. Patient first: ABCDE assessment, haemodynamic trend, conscious level, neuromuscular signs, pupillary size, temperature, urine output and fluid balance. In anaesthetised patients use end-tidal agent concentration if applicable, processed EEG where used, capnography, arterial pressure waveform and neuromuscular monitoring.
2. Identify the drug, concentration and intended dose: calculate dose in pharmacological units, not merely mL h^-1. For example, noradrenaline 4 mg in 50 mL gives 80 microgram mL^-1; at 5 mL h^-1 this is 400 microgram h^-1.
3. Inspect the complete system: syringe size and seating, plunger engagement, anti-siphon valve, line clamps, Luer locks, filters, one-way valves, cannula patency, pressure bag, manifold dead space, and whether the infusion enters a high-flow carrier line.
4. Interrogate the pump: programmed rate, volume infused, volume to be infused, bolus history, alarm log, battery/mains status, drug library selection, concentration field, patient weight, TCI model and target if applicable.
5. Preserve evidence: do not discard syringe, giving set or pump after serious incident; document pump serial number, settings, residual volume and line configuration.

Investigations and interpretation

Investigations are directed by the suspected drug effect and by pump physics. Compare prescribed dose, programmed delivery, volume actually infused and clinical effect. A discrepancy between pump-displayed volume and syringe residual volume suggests programming error, free flow, siphoning, leakage or inaccurate syringe recognition.

Investigation	Use	Interpretation / thresholds
Blood gas, lactate, electrolytes	Shock, respiratory depression, insulin/potassium infusions	Hyperkalaemia >6.0 mmol L-1, severe acidaemia pH <7.20, lactate >2 mmol L-1 support significant physiological consequence.
Glucose monitoring	Insulin or dextrose infusion incidents	Urgent confirmation if glucose <4.0 or >15 mmol L-1; arterial analyser preferred in shocked patients.
Coagulation tests	Heparin infusion error	APTT ratio and anti-Xa; therapeutic unfractionated heparin anti-Xa is typically 0.3–0.7 IU mL-1.
Drug levels	Selected drugs: aminoglycosides, vancomycin, lithium, digoxin, phenytoin	Rarely useful acutely for anaesthetic agents or local anaesthetic toxicity; treat clinically.
Pump bench testing / biomedical engineering download	Suspected device failure	Formal flow accuracy, occlusion alarm, battery and event-log analysis; remove from service pending testing.

Device-performance thresholds relevant to clinical interpretation

Most infusion pumps are assessed against IEC 60601-2-24 principles using short-term and long-term accuracy curves, including the trumpet curve. Clinically, syringe pumps are generally more accurate at low flows than volumetric pumps, but performance is degraded by very low rates, large syringes, compliant tubing, vertical height changes and start-up delay. Typical accuracy is approximately ±2% for well-calibrated syringe pumps and ±5% for volumetric pumps, although the manufacturer’s specification is definitive.

• Dead space delay: drug arrival time equals internal volume divided by flow. A 1 mL extension line at 1 mL h^-1 produces an approximately 60 min delay; this is critical for vasoactive drugs and TIVA.
• Hydrostatic pressure: a 1 m vertical height difference generates about 74 mmHg. Raising a syringe above the patient can cause siphoning/free flow unless anti-siphon mechanisms are intact.
• Occlusion alarms: thresholds are device-dependent, commonly selectable around 300, 500 and 900 mmHg. Alarm time increases as flow falls and system compliance rises; therefore absence of an alarm does not exclude extravasation or partial occlusion.
• Occlusion release bolus: accumulated elastic energy in compliant tubing may deliver a bolus when obstruction is relieved, particularly at high pressure thresholds and with long, soft tubing.

Guideline-based safety assessment should consider local governance, MHRA safety alerts and national standards for injectable medicines, including double-checking of high-risk infusions, standard concentrations, labelled dedicated lines, anti-free-flow administration sets, and smart-pump drug libraries. For examination purposes, the key diagnostic principle is that infusion incidents are usually systems failures: prescription, preparation, programming, line configuration, patient access and device mechanics must all be assessed systematically.

Management, pharmacology and procedures

Selection and prescription of infusion devices

Choice of device is determined by required accuracy, flow range, drug potency, viscosity, vascular access and clinical risk. Volumetric pumps are preferred for maintenance fluids, blood products and large-volume infusions, whereas syringe drivers are used for potent vasoactive, anaesthetic and analgesic drugs where low dead space and accurate low-flow delivery are required. Gravity administration is inappropriate for vasoactive drugs, paediatric infusions, neuraxial infusions and most critical care sedatives because flow varies with hydrostatic pressure, venous pressure, tubing resistance and patient position.

Application	Typical device	Key management point
Noradrenaline, adrenaline, insulin, heparin	Syringe pump or dedicated smart pump channel	Use concentration standardisation, anti-reflux/non-return valves, minimal dead space and clear labelling
Propofol/remifentanil TCI	Target-controlled infusion pump	Select correct pharmacokinetic model, patient covariates and effect-site/plasma target
PCA opioids	PCA pump with lockout	No background infusion in opioid-naïve adults except specialist indications; monitor sedation and respiratory rate
Epidural/local anaesthetic infusion	Dedicated epidural pump with non-Luer/NRFit connectors where available	Never co-administer with intravenous lines; mandatory sensory, motor and haemodynamic surveillance

Set-up procedure and risk reduction

1. Prescription: prescribe drug, concentration, dose units, route, maximum dose and review time. Avoid ambiguous rates such as “ml/h” alone for high-risk drugs.
2. Preparation: use aseptic non-touch technique; employ prefilled or pharmacy-prepared syringes where possible. Standardisation reduces calculation error.
3. Priming: prime the entire administration set before connection. Air-in-line alarms are not a substitute for priming; clinically relevant venous air embolism is possible with pressurised infusions, particularly via central venous access.
4. Connection: use dedicated lumens for vasoactive drugs; anti-siphon valves prevent free flow from height difference, and anti-reflux valves reduce retrograde drug accumulation when multiple pumps share a line.
5. Programming: independently check patient weight, concentration, dose, rate, VTBI and library limits. Smart pumps with dose-error reduction systems provide soft and hard limits but do not eliminate human error.
6. Monitoring: document line site, pump settings, delivered volume and remaining volume at handover. For vasoactive infusions, observe haemodynamics continuously during syringe changes and rate adjustments.

Pharmacological infusions commonly examined

Drug	Usual adult infusion range	Important pharmacokinetic/clinical point
Noradrenaline	0.02–1 microgram/kg/min	Short context-insensitive effect; interruption may cause abrupt hypotension. Central administration preferred; peripheral use requires close site observation.
Adrenaline	0.01–0.5 microgram/kg/min	β then α effects; lactate may rise from β2-mediated glycolysis.
Propofol anaesthesia	4–12 mg/kg/h initially; sedation often 0.3–4 mg/kg/h	Context-sensitive decrement increases with duration; avoid prolonged high-dose ICU use >4 mg/kg/h for >48 h because of propofol infusion syndrome risk.
Remifentanil	0.05–2 microgram/kg/min	Metabolised by nonspecific esterases; context-sensitive half-time approximately 3–5 min independent of infusion duration.
Insulin variable-rate infusion	Commonly 0.5–6 units/h, titrated to capillary glucose	Use substrate fluid and potassium replacement protocols; hypoglycaemia is a pump-related and prescription-related hazard.
Unfractionated heparin	Often 12–18 units/kg/h after bolus when indicated	Titrate to anti-Xa or APTT ratio per local protocol; check platelets for HIT during ongoing therapy.

Target-controlled infusion and procedural use

TCI pumps use compartmental pharmacokinetic models to calculate infusion rates achieving a target plasma or effect-site concentration. Propofol models include Marsh, which uses weight and fixed central volume assumptions, and Schnider, which incorporates age, height, weight and lean body mass; remifentanil commonly uses the Minto model. Effect-site targeting produces an initial overshoot in plasma concentration to accelerate equilibration, governed by the ke0 constant. Typical adult propofol targets are 3–6 microgram/ml for induction/maintenance and 0.5–2 microgram/ml for sedation, adjusted to age, comorbidity, co-administered opioids and processed EEG where used. Remifentanil effect-site targets commonly range from 2–8 ng/ml for anaesthesia and 0.5–2 ng/ml for procedural analgesia.

PCA prescriptions should specify bolus dose, lockout interval, maximum hourly dose and background infusion status. Morphine PCA commonly uses 1 mg bolus with 5 min lockout; fentanyl 10–20 microgram bolus with 5 min lockout may be used when morphine is unsuitable. Monitoring focuses on respiratory rate, oxygen saturation, sedation score, pain score, nausea, pruritus and cumulative dose. Sedation usually precedes opioid-induced respiratory depression; increasing somnolence mandates stopping the infusion, oxygen, stimulation, airway assessment and naloxone titration, typically 20–40 microgram IV increments to restore ventilation while avoiding severe pain or sympathetic surge.

Complications, troubleshooting and follow-up

Important pump complications include occlusion, infiltration/extravasation, free-flow, siphonage, incorrect syringe recognition, wrong concentration, wrong route, battery failure, electromagnetic or mechanical malfunction, and delayed onset after syringe change. Occlusion alarms are pressure-threshold alarms, commonly set around 300–900 mmHg depending on device and clinical area; at low flow rates, time to alarm may be prolonged because pressure rises slowly, and post-occlusion bolus may occur when obstruction is relieved. Management is to stop the infusion, clamp appropriately, assess the patient and line, aspirate if possible, replace the cannula if uncertain, and never simply flush a line containing concentrated vasoactive drug.

Extravasation of vasopressors requires immediate cessation, aspiration through the cannula, limb elevation, warm compresses for catecholamine extravasation, senior review and local protocol treatment. Phentolamine infiltration, for example 5–10 mg diluted in 10–15 ml 0.9% saline infiltrated subcutaneously around the affected area as early as possible, is a recognised antidote for α-adrenergic vasoconstrictor injury. Follow-up includes documentation, incident reporting, tissue viability or plastic surgery review where blistering, pain, pallor or delayed capillary refill occurs.

Long-term safety depends on governance: device training, preventive maintenance, electrical safety testing, drug libraries, standard concentrations, audit of overrides, and investigation of near misses. At clinical handover, infusion indication, concentration, rate, access route, remaining volume, pump battery status and contingency plan for interruption should be explicitly communicated.

Exam controversies and advanced synthesis

Standards, governance and what “accuracy” really means

In viva discussion, avoid quoting a single accuracy figure without context. Infusion devices are regulated as active medical devices and tested against standards such as IEC 60601-2-24, which assesses flow accuracy using trumpet curves: the percentage flow error is expressed over observation windows, commonly 2, 5, 11, 19 and 31 minutes, after an initial stabilisation period. A syringe pump may have a quoted steady-state accuracy of approximately ±2%, and a volumetric pump approximately ±5%, but short-term error at very low flow rates may be clinically important, especially for vasoactive drugs, neonates and TIVA.

UK practice is shaped by MHRA device alerts, local medical-device policies, NPSA principles for injectable medicines, and specialty recommendations emphasising standardised concentrations, line labelling, anti-free-flow mechanisms, dose-error reduction software and staff training. “Smart pumps” with drug libraries reduce some programming errors but do not abolish them: common bypasses include using a generic mode, selecting the wrong drug profile, accepting soft limits, and failure to update libraries after formulary changes.

Contentious areas: syringe pumps versus volumetric pumps

Issue	Syringe pump	Volumetric pump	Exam implication
Best use	Small-volume, high-potency infusions: noradrenaline, remifentanil, insulin, paediatric drugs	Larger-volume crystalloid, blood-compatible sets, maintenance fluids, enteral feeds	Match device to clinical risk, not merely availability
Accuracy determinant	Syringe internal diameter, plunger friction, drive mechanics, compliance	Giving-set compliance, drop sensor/linear peristaltic cassette, back-pressure	Wrong syringe brand selection can cause dose error
Hazards	Start-up delay, siphoning, bolus after occlusion release, vertical displacement effects	Free flow, air entrainment, set misloading, door/cassette errors	Anti-siphon valves and correct loading are safety-critical
Low-flow performance	Potentially poor below 0.5–1 ml h-1	Usually unsuitable for very low-rate potent drugs	Use higher concentration? Not always: balance dead-space delay versus pump mechanics

A classic viva trap is the assumption that occlusion alarms protect the patient immediately. Alarm delay depends on flow rate, compliance, syringe size, line elasticity, downstream pressure and alarm threshold. At 1 ml h^-1, an occlusion may take many minutes to generate sufficient pressure. Alarm thresholds are commonly selectable, for example 300–900 mmHg (40–120 kPa). High thresholds reduce nuisance alarms but increase stored elastic energy and post-occlusion bolus risk. Low thresholds improve sensitivity but may alarm during vasopressor infusion through narrow cannulae or long extension tubing.

Hydrostatic and compliance effects: high-yield viva physics

Vertical movement of a syringe pump relative to the patient changes outlet pressure by approximately 0.74 mmHg cm^-1 for water-like fluids. Raising the pump may transiently increase flow or cause bolus; lowering it may transiently reduce or reverse flow. This is clinically relevant for catecholamines, neonatal infusions and TCI at low flow. Siphoning occurs when the syringe is above the patient and the plunger is not mechanically restrained; prevention requires an anti-siphon system, secure syringe clamp, Luer-lock connections and avoidance of unsecured extension lines.

Dead space is another underestimated problem. A 150 cm low-bore extension set may contain approximately 1–2 ml; a standard giving set may contain 10–20 ml. At 2 ml h^-1, a drug change may therefore take 30–60 minutes to reach the patient unless the line is flushed or the infusion architecture is rationalised. Inotropes should enter as close to the patient as possible, ideally through a dedicated lumen with non-return valves to prevent retrograde flow when carrier rates change.

TCI, closed-loop delivery and trial-level synthesis

Target-controlled infusion is a pharmacokinetic control problem rather than a simple pump mode. Propofol models include Marsh and Schnider; remifentanil commonly uses the Minto model. Typical adult propofol effect-site targets are approximately 3–6 micrograms ml^-1 for induction/maintenance, with remifentanil effect-site targets often 2–8 ng ml^-1 depending on stimulus and co-administered hypnotic. Controversies include use at extremes of age, pregnancy, severe obesity, critical illness and major haemodynamic instability. Marsh uses total body weight scaling and may overdose the obese if used uncritically; Schnider incorporates age, height, weight and lean body mass but may behave unpredictably at extremes. No model “measures” plasma concentration; all generate predictions with inter-individual variability commonly 20–30% or more.

Closed-loop anaesthesia using processed EEG has shown improved time within target hypnotic range in multiple small to moderate trials, but outcome benefits such as mortality, awareness reduction or faster discharge remain inconsistent. For Primary FRCA, the defensible answer is that closed-loop systems integrate a sensor, controller and actuator; their limitations are sensor artefact, pharmacodynamic delay, model error, and failure modes requiring human override.

Common pitfalls and defensible safety statements

• Never rely on the pump alone: trace the line from device to patient, confirm route, concentration, units and lumen.
• Use standard concentrations for high-risk drugs where possible; prescribing in micrograms kg^-1 min^-1, units h^-1 or ml h^-1 must be internally consistent.
• Separate routes: epidural, regional and enteral pumps should be physically and connector-distinct; NRFit systems reduce misconnections.
• Account for failure states: battery depletion, mains disconnection, door open, clamp closed, air-in-line, upstream occlusion, downstream extravasation, and pump swap during transfer.
• During MRI or transfer: use MR-conditional devices or long extension tubing with awareness of increased compliance, resistance and alarm delay.

The mature exam answer integrates engineering limitations, pharmacological consequences and human factors: infusion pumps improve precision, but safety depends on the entire infusion system, including device, syringe or giving set, line configuration, drug library, user behaviour and clinical monitoring.