Cardiac Output Monitoring Using Indicator-Dilution Techniques: Basics, Limits, and Perspectives

Authors, Journal, Affiliations, Type, DOI

• Authors: Daniel A. Reuter, Cecil Huang, Thomas Edrich, Stanton K. Shernan, Holger K. Eltzschig
• Journal: Anesthesia & Analgesia, 2010;110(3):799–811
• Affiliations: Hamburg-Eppendorf University Hospital (Hamburg); Brigham and Women's Hospital / Harvard Medical School (Boston); Tuebingen University Hospital; University of Colorado Denver
• Type: Review article
• DOI: https://doi.org/10.1213/ANE.0b013e3181cc885a

Overview

This review covers the theoretical basis and clinical application of indicator-dilution techniques for cardiac output (CO) monitoring in critically ill and perioperative patients. It reviews three methods based on this principle: (1) intermittent bolus PA thermodilution (IB-PATD) via pulmonary artery catheter — the longstanding clinical standard; (2) continuous PA thermodilution (CPATD) using the Vigilance II or Q2plus systems; and (3) transcardiopulmonary thermodilution (TCPTD, PiCCO) and lithium dilution (LiDCO) as less-invasive alternatives. The technical principles, measurement error sources, and clinical advantages/limitations of each method are outlined. The review concludes that PAC-free techniques (TCPTD, LiDCO) provide accurate CO measurement while TCPTD uniquely adds global end-diastolic volume (GEDV) and extravascular lung water (EVLW) as volumetric preload and pulmonary oedema markers.

Keywords

Cardiac output, thermodilution, indicator dilution, pulmonary artery catheter, transcardiopulmonary, PiCCO, lithium dilution, hemodynamic monitoring, critically ill, extravascular lung water

Key Takeaways

Stewart-Hamilton Principle — Mathematical Basis

• All indicator-dilution methods use the same principle: inject a known amount of indicator → measure downstream concentration-time curve → CO = amount injected / area under dilution curve
• Stewart (1897): bolus injection; collection of diluted indicator; F = C₀V₀ / (C₁ × t)
• Hamilton modification (1928): introduced explicit time-concentration curve c(t) to capture laminar flow, variable transit times, and continuous dilution downstream
• Stewart-Hamilton equation: CO = C₀V₀ / ∫c(t)dt

Intermittent Bolus PA Thermodilution (IB-PATD)

• Cold indicator (iced or room-temperature saline) injected into the right atrium via PAC proximal port; thermistor at PA detects temperature change
• CO inversely proportional to the area under the thermodilution curve
• De facto clinical standard for >40 years; measures right heart output (pulmonary blood flow), not systemic CO — clinically equivalent in absence of shunts
• PAC additionally provides PA pressures and mixed venous O₂ saturation

Sources of Measurement Error:

1. Loss before injection — underfilling syringe, warming of iced injectate (each 1°C increase in 0–4°C range contributes ~3% CO overestimate); discard first measurement in a run
2. Loss during injection — catheter dead space (0.7–1 mL), conductive warming through intravascular catheter wall → 9–17% indicator loss → ~20% CO overestimate; corrective constant K₂ precomputed per catheter type
3. Loss after injection — conductive rewarming by surrounding tissue (worse in low-flow states and with TCPTD due to longer transit distance)
4. Injectate temperature/volume variation — 10 mL iced injectate provides highest reproducibility; room temperature acceptable but larger percent error in high/low-flow states
5. Recirculation and detainment — left-to-right shunt → CO underestimation; one-lung ventilation (catheter in collapsed lung branch) → underestimation due to prolonged thermodilution curve
6. Tricuspid regurgitation — conflicting data on direction; reverse regurgitant flow prolongs indicator transit; both over- and underestimates reported; one-flow-state study: overestimates in low-flow, underestimates in high-flow; severity-dependent effects unresolved
7. Baseline temperature fluctuations — exogenous (CPB cooling, concurrent IV infusions) or endogenous (respiratory oscillations); stable respiratory pattern required in pre-measurement period
8. Cyclic CO changes — stroke volume varies up to 50% across respiratory cycle (more prominent for RV); 3 injections at same respiratory phase recommended (clinically performed); averaging asynchronous measurements over full cycle is theoretically correct but optimal n unclear — 3 may be insufficient
9. Curve truncation/extrapolation — manufacturer-specific algorithms apply; many truncate when curve returns to 50% of peak then add empirically derived correction

Reproducibility limitations:

• 22% CO change needed for statistical significance (single measurement, Stetz 1982)
• 13% change needed for triplicate measurements
• No confirmatory data published since 1982 (limitation noted by authors)

Continuous PA Thermodilution (CPATD)

• Heating filament on PAC (15–25 cm from tip) intermittently heats blood in superior vena cava; thermistor at tip detects response
• Vigilance II (Edwards): flat filament, 1–4s pseudorandom pulses; stochastic analysis
• Q2plus (Hospira): coiled filament, 20s pulses in repetitive on-off cycle every 40s; averaging algorithm
• Correlates well with IB-PATD across wide CO range; less affected by TR (pig model data)
• In hypothermia (post-CPB, liver transplantation): IB-PATD exceeds CPATD until resolution — large cold bolus less susceptible to thermal noise than small heat signals
• Key limitation — time delay: 50% response at ~9 minutes, 80% response at ~12 minutes (Vigilance, Haller et al.); in vitro: 2.9 vs 3.3 min (20% response), 4.7 vs 11.2 min (80% response) → unsuitable for detecting rapid hemodynamic changes
• Eliminates need for fluid boluses, reduces contamination risk, provides continuous CO trend

Transcardiopulmonary Thermodilution (TCPTD) — PiCCO System

• Cold indicator injected into superior vena cava via central venous catheter; thermistor in femoral (or axillary/brachial) artery records thermodilution curve after transit through right heart, pulmonary circulation, and left heart
• No PAC required — less invasive; avoids PA rupture, pulmonary embolism
• Currently available system: PiCCO monitor (Pulsion Medical Systems, Munich)
• Performs monoexponential extrapolation on TCPTD curve before CO calculation
• Combines TCPTD with pulse contour analysis for continuous real-time LV CO assessment

Additional variables (unique to TCPTD):

• GEDV (global end-diastolic volume): sum of all four cardiac chamber volumes = ITTV − PTV
  • ITTV = CO × mean transit time (MTT); PTV = CO × exponential decay time (EDT)
  • GEDV superior to filling pressures (PAWP, CVP) as preload estimate (multiple studies)
• EVLW (extravascular lung water): ITTV − GEDV − intrathoracic blood volume; quantifies pulmonary oedema
  • EVLW-guided protocols: hasten oedema resolution, shorten time to extubation, decrease ICU stay vs filling pressure-guided management (Goepfert 2007)
• Right-to-left intracardiac shunt detection: double peaks in TCPTD curves indicate right-to-left shunting

Accuracy vs IB-PATD: Correlation coefficient >0.9 and bias <10% in most studies; 96–97% of indicator reaching PA recovered in aorta; TCPTD measures LV CO vs IB-PATD measuring RV CO → transient cold-induced sinus slowing affects RV more than LV → TCPTD slightly higher values in some studies

Limitations vs IB-PATD:

• Cannot measure PA pressures, PA occlusion pressure, or mixed venous O₂ saturation (central venous O₂ saturation can approximate SvO₂)
• Femoral arterial cannulation: infection risk (requires diligent nursing); contraindicated in severe peripheral vascular disease (thromboembolism risk)

Transcardiopulmonary Lithium Dilution (LiDCO)

• Isotonic lithium chloride (0.002–0.004 mmol/kg) injected via central or peripheral venous route; detected by ion-selective electrode in arterial line flow-through cell (peristaltic pump at 4 mL/min)
• Voltage related to plasma lithium concentration via Nernst equation; CO from lithium dilution curve area (50% peak cut-off to limit recirculation effect)
• Plasma flow converted to blood flow via haematocrit: CO = plasma flow / (1 − PCV)
• LiDCO-Plus: combines lithium dilution with pulse contour analysis for continuous CO
• Peripheral venous injection feasible: comparable accuracy to central injection (two studies); only arterial + peripheral venous access required
• EVLW measurement described (Maddison animal study) but could not be confirmed clinically

Limitations:

• Contraindicated in patients on lithium therapy (background lithium → CO overestimation)
• Manufacturer limits repeated measurements over short period (lithium accumulation); precise number not specified
• Contraindicated: weight <40 kg; first trimester pregnancy
• Electrode drift with certain muscle relaxant infusions — inaccurate measurements

Limitations of the Document

• 2010 publication date: does not cover newer minimally invasive/non-invasive CO technologies (transesophageal Doppler, thoracic bioimpedance/bioreactance, non-invasive pulse contour analysis, near-infrared spectroscopy)
• PAC outcome data limited: large trials questioning PAC effect on outcome in critically ill are discussed but not updated; ESCAPE trial referenced (decompensated HF); no definitive CS-specific RCT data available at time of publication
• Tricuspid regurgitation effects: authors acknowledge unresolved direction and severity-dependent impact on thermodilution accuracy
• Reproducibility data: Stetz 1982 findings on measurement reproducibility (22% threshold for significance) had not been confirmed by subsequent studies at time of publication
• Technical detail level: highly mathematical; practical implementation guidance (e.g., catheter-specific K₂ correction constants, injection technique) may not reflect contemporary practice in all centres

Key Concepts Mentioned

• concepts/Cardiac-Output-Measurement — comprehensive review of all indicator-dilution CO methods
• concepts/Invasive-Hemodynamic-Monitoring-CS — PAC-guided monitoring; CO measurement limitations in CS
• concepts/Right-Heart-Catheterization — thermodilution CO technique; sources of error in PAC-based CO

Key Entities Mentioned

• PiCCO Monitor (Pulsion Medical Systems, Munich) — transcardiopulmonary thermodilution + pulse contour
• LiDCO-Plus (LiDCO, Cambridge) — lithium dilution + pulse contour
• Vigilance II (Edwards Lifesciences) — continuous PA thermodilution; pseudorandom heating
• Q2plus (Hospira, Lake Forest) — continuous PA thermodilution; repetitive 20s pulses

Wiki Pages Updated

• Created: wiki/sources/co-indicator-anesthanalg-2010.md
• Created: wiki/concepts/Indicator-Dilution-Cardiac-Output.md
• Updated: wiki/concepts/Right-Heart-Catheterization.md — added thermodilution error sources and source cross-reference
• Updated: wiki/concepts/Invasive-Hemodynamic-Monitoring-CS.md — added cross-reference to indicator-dilution concept
• Updated: wiki/sourceindex.md
• Updated: wiki/wikiindex.md