Beyond the vital signs. Without the catheter.
The bedside monitor tells you about pressure and oxygen, often late. Invasive catheters measure flow, at a price paid by the patient. Here is where non-invasive flow monitoring fits between them, and what the evidence says about each.
Five numbers that tell you the end of the story
Heart rate, blood pressure, oxygen saturation, breathing rate and temperature have been on every monitor for decades. They are essential, but in early circulatory failure the body defends blood pressure by tightening vessels and speeding the heart, so pressure stays normal while flow is already falling. Hypotension is a late sign: the international shock consensus says it should not be required to diagnose shock.
Drag to bleed a healthy adult, step by step. Watch which monitor notices first.
Standard monitor
Flow monitoring
Teaching approximation built from: stroke volume falling about 2% for each 1% of blood lost (Hinojosa-Laborde 2014, doi); systolic pressure and pulse falling only after about a 55% fall in stroke volume in human volunteers (Ryan 2008, doi); stroke volume changing in 74% of donors after a 450 mL donation, heart rate in 65%, systolic pressure in 52% (Stewart 2014, doi); ATLS haemorrhage classes; ESICM shock consensus, Cecconi 2014 (doi). Alarm thresholds here are illustrative. Real patients vary widely: real-world trauma data show heart rate and pressure move even less than the textbook suggests (Mutschler 2013, doi).
| Parameter | What it tells you | When it changes as the patient deteriorates | Limitations |
|---|---|---|---|
| ECG and heart rate | Rhythm, ischaemia, rate | Rises from about 15–30% blood loss, often blunted | Affected by beta-blockers, pain, age, pacing. Irreplaceable for rhythm. |
| Blood pressure | Perfusion pressure | Late: after about 30% loss, or a 50–55% fall in stroke volume | Held up by vasoconstriction; a poor guide to flow |
| SpO₂ | Oxygen saturation | Late in shock; the signal can be lost when perfusion is poor | Says nothing about flow or oxygen delivery. Irreplaceable for oxygenation. |
| Breathing rate, temperature | Compensation, infection | Rises in moderate to severe loss | Often poorly recorded; non-specific |
| Stroke volume | Blood ejected per beat | Earliest: falls about 2% per 1% of volume lost | Electrical Cardiometry's single values are imprecise; the trend matters most |
| Cardiac output | Total flow | Intermediate: the rising heart rate protects it at first | Same caveat as stroke volume |
| Vascular resistance | Vessel tightening | Rises early, then collapses just before decompensation | Calculated from pressure and output, so it carries both errors |
| Thoracic fluid | Fluid in the chest | Rises with congestion and oedema | Affected by effusions and body habitus |
The physiology, stroke volume falling well before pressure, is strong. Evidence that a flow monitor changes outcomes by detecting deterioration earlier is still limited and mostly observational. Flow monitoring adds to the standard monitor; it never replaces the ECG or SpO₂.
Measuring flow: catheter or skin sensors?
Thermodilution through a pulmonary artery catheter, or through a central line and an arterial line (transpulmonary), has long been the reference for cardiac output. It comes with a cost to the patient that non-invasive monitoring avoids.
| Pulmonary artery catheter (thermodilution) | Transpulmonary thermodilution and pulse contour | Arterial pulse contour (uncalibrated) | Electrical Cardiometry | |
|---|---|---|---|---|
| What goes into the patient | Catheter through a central vein and the heart into the pulmonary artery | Central venous line plus an arterial line, usually femoral | Arterial line | Nothing: four skin sensors |
| Risks | Arrhythmia, infection, bleeding, pneumothorax, thrombosis, rare pulmonary artery rupture | Line infection, bleeding, thrombosis, limb ischaemia | Arterial line risks | Skin irritation from the sensors |
| Readings | Intermittent boluses; continuous only with special catheters | Calibrated bolus, then continuous estimate that drifts with vascular tone | Continuous; less reliable when vessel tone changes fast | Continuous, beat to beat |
| Time to first value | Insertion procedure, imaging, calibration | Two lines and calibration | Once the arterial line is in | Seconds after the sensors go on |
| Newborns and small children | Rarely feasible | Possible in larger children, with paediatric catheters | Possible, but less validated | Designed for all ages, including preterm babies |
| Where it can be used | Intensive care, cardiac theatre | Intensive care, theatre | Wherever an arterial line exists | ICU, theatre, emergency, ward, clinic, transport |
| Skill needed | Specialist insertion and interpretation | Central and arterial access | Arterial access | Sensor placement in minutes |
| Accuracy of a single value | Reference method, about ±20% itself | Close to the pulmonary catheter | Varies with vascular tone | Wider spread than thermodilution; strongest for trends |
The disadvantages of invasive monitoring, from the trials
No survival benefit
Randomised trials of the pulmonary artery catheter found no reduction in deaths: in high-risk surgery (Sandham 2003, NEJM), in intensive care (PAC-Man, Harvey 2005, Lancet) and in heart failure (ESCAPE 2005, JAMA). A Cochrane review confirmed no difference in mortality or length of stay (Rajaram 2013).
Real complications
In PAC-Man about one in ten catheterised patients had a complication, such as a haematoma at the insertion site or an arrhythmia. In ESCAPE, adverse events were roughly twice as common with the catheter. An earlier observational study linked right-heart catheterisation with higher mortality (Connors 1996, JAMA).
The reference has its own error
Thermodilution's own precision is about ±20%, which is why studies accept up to 30% disagreement between methods (Critchley 1999). It is also less reliable with tricuspid regurgitation (Hamilton 1989), intracardiac shunts and arrhythmias, and depends on bolus technique.
Too late, too narrow
A catheter is placed once a patient is already critically ill, in intensive care or theatre. It cannot screen the emergency department, follow a patient on the ward, or monitor a 900-gram baby. Non-invasive monitoring can start before the crisis.
So, is Electrical Cardiometry as accurate as thermodilution?
For a single absolute number, no: thermodilution remains the reference, and Electrical Cardiometry's individual readings spread more widely around it. Pooled studies put the difference at a 42–48% percentage error, similar to other non-invasive monitors (see the evidence). But part of that spread belongs to thermodilution itself, and bedside decisions mostly rest on change: did stroke volume rise after the leg raise, the fluid, the drug? That is where a continuous, risk-free monitor earns its place, and where no catheter can follow most patients.
The fair summary: use thermodilution when you need a reference value and the patient already justifies the lines; use Electrical Cardiometry to see trends early, in any patient, at any age, without a needle.