Hemodynamic Normals

In depth

The one-page summary above is what the PDF prints. These sections go further.

Cardiac output and its four determinants

Cardiac output is heart rate times stroke volume, and stroke volume is set by preload, afterload and contractility. Output carries oxygen: DO₂ = CO × CaO₂ × 10, so a CO of 5 L/min with a CaO₂ of 20 mL/dL delivers 1,000 mL/min against a resting VO₂ of about 200–250 mL/min, an extraction near 25%. Every monitor on this page measures one of these pieces or a pressure that stands in for one.

Frank-Starling: on the steep part of the curve more end-diastolic volume gives more stroke volume; on the flat part it only raises filling pressures and drives edema. A failing ventricle has a flatter curve, and inotropes lift it. CVP and PAOP stand in for volume only through compliance, so a stiff ventricle (hypertrophy, ischemia, tamponade, high PEEP) shows a high pressure with a small volume.

Afterload is wall stress during ejection (Laplace: pressure × radius ÷ 2 × wall thickness), so a dilated ventricle faces more of it at the same blood pressure; SVR is only a calculated surrogate. A failing LV loses stroke volume steeply as SVR rises, which is why afterload reduction and IABP raise output in cardiogenic shock. The thin RV is even more afterload-sensitive: a previously normal RV cannot generate a mean PA pressure much above 40 mmHg, so massive PE presents as RV failure rather than high PA pressure. Rate raises output until diastole is too short to fill, often from about 120–150 in adults and earlier with a stiff ventricle; the atrial kick supplies about 15–30% of LV filling, so atrial fibrillation or VVI pacing can drop output with no change in volume.

Transducer physics: leveling, zeroing, dynamic response

Zeroing opens the transducer to air so the monitor subtracts atmospheric pressure; do it at setup, after disconnection and whenever the numbers do not fit the patient. Leveling puts the air-fluid interface at the phlebostatic axis. Each 1.36 cm of height error is 1 mmHg, so a 10 cm error is about 7 mmHg: trivial against a MAP of 80, but it turns a CVP of 8 into 15 or 1. Re-level pole-mounted transducers after every change in bed height; HOB 0–60° supine is acceptable, and lateral positions need the angle-specific reference points in the AACN 2016 alert.

Dynamic response: catheter, tubing and diaphragm form a spring-mass system with a natural frequency (how fast it rings) and a damping coefficient (how fast the ringing dies). The arterial pulse contains harmonics up to about ten times the heart rate, so a system with too low a natural frequency resonates and overshoots systolic. Bedside kits sit around 10–20 Hz, so damping must be near the optimum of about 0.6–0.7 (Gardner 1981). On a printed square-wave test, natural frequency is paper speed divided by the distance between oscillation peaks (25 mm/s ÷ 1.5 mm ≈ 17 Hz); at the bedside, 1.5–2 oscillations is the practical proxy.

Arterial waveform analysis

The trace has an upstroke (its slope reflects contractility), a systolic peak, a dicrotic notch (aortic valve closure plus reflected waves) and diastolic runoff; the area under the systolic portion tracks stroke volume, which pulse-contour monitors exploit. From aorta to radial artery, reflected waves raise the systolic peak (often 10–20 mmHg) and lower the diastolic slightly, while MAP falls only 1–2 mmHg; hence vasopressors are titrated to MAP. In septic shock on high-dose vasopressors, after bypass and in hypothermia, radial MAP can read several mmHg below central pressure (Dorman 1998), so consider a femoral line when the radial number does not fit the patient.

On positive-pressure ventilation, inspiration first squeezes blood from the lungs into the LV (systolic rises), then cuts RV filling so LV output falls a few beats later. A large fall means preload dependence and is the basis of pulse pressure variation.

CVP: what it measures and where it fails

CVP is right atrial pressure and, without tricuspid stenosis, RV end-diastolic pressure. It is also the back pressure for venous return (Guyton: mean systemic filling pressure minus RAP, divided by venous resistance), so a high CVP obstructs flow and lowers organ perfusion pressure; kidney perfusion is roughly MAP − CVP.

As a guide to fluid, CVP is a coin toss in the normal range: an area under the curve of 0.56 in the Marik 2013 meta-analysis. Extremes help a little: under about 6–8 makes response more likely, above about 12–15 less likely (De Backer and Vincent 2018). Its best use is as a safety limit: if CVP climbs with a bolus while stroke volume does not, the RV is on the flat part of its curve. Besides volume, CVP rises with RV failure, pulmonary hypertension, PE, tricuspid regurgitation, tamponade, PEEP and auto-PEEP, abdominal hypertension and tension pneumothorax.

With no a wave (atrial fibrillation), read CVP at the z point, at the end of the QRS just before the c wave. With large v waves, the digital mean overestimates filling pressure.

PA catheter: insertion, waveforms and wedge validity

The adult catheter is 110 cm, 7–8 Fr, with marks every 10 cm. The distal lumen reads PA pressure and gives mixed venous samples; the proximal lumen, about 30 cm back, sits in the RA for CVP and injectate; the balloon takes 1.5 mL; a thermistor near the tip reads blood temperature. From the right internal jugular the RA is reached at about 15–20 cm and wedge usually at 45–55 cm; subclavian is about 5 cm shorter, femoral about 10–15 cm longer (Edwards Quick Guide).

The balloon is inflated in the RA and the catheter floats with the flow. RA shows a and v waves; RV has a diastolic near 0 that rises through diastole; PA has a diastolic step-up that falls through diastole and a dicrotic notch; wedge is damped with a and v waves. No RV trace within about 20 cm of the RA suggests coiling: deflate, withdraw to the RA and retry. Withdraw only with the balloon down. Watch for ventricular ectopy as the tip crosses the RV and for right bundle branch block, which with existing LBBB means complete heart block, so have pacing available. Final position: nearly the full 1.25–1.5 mL is needed to wedge and the PA trace returns at once on deflation.

PAOP reflects left atrial pressure only through a continuous column of blood to the LA, which requires West zone 3, where pulmonary arterial and venous pressures exceed alveolar pressure. In zones 1 and 2 the catheter reads alveolar pressure instead. Hypovolemia and high PEEP enlarge those zones; flow-directed catheters usually float to dependent zone 3. Signs the wedge is valid:

• Tip at or below the LA on a lateral film.
• Clear a and v waves; a smooth, flat trace suggests zone 1 or 2.
• PAOP lower than PAD.
• Respiratory swings in PAOP no larger than those in PAD.
• PAOP rises by less than half of any increase in PEEP.

Read at end-expiration, when pleural pressure is closest to atmospheric: just before the inspiratory dip when breathing spontaneously, just before the inspiratory rise on the ventilator. The digital display can be wrong by the full respiratory swing, so use a frozen trace beside the airway pressure or capnography waveform. Active exhalation raises end-expiratory values falsely and has no clean fix. PEEP above about 10 cmH₂O makes PAOP overestimate transmural LA pressure (Bootsma 2022), but never take the patient off PEEP to measure: derecruitment is harmful and the number no longer describes the patient as treated.

Measuring flow and oxygen: thermodilution, Fick, SvO₂

Bolus thermodilution: cold injectate enters the RA through the proximal port and the thermistor records the temperature curve in the PA; output is inversely proportional to the area under the curve (Stewart-Hamilton). Inject 10 mL at least 10 °C colder than blood smoothly in under 4 seconds at end-expiration; average three values within about 10% of each other. Continuous-output catheters heat blood with a filament and report a rolling average over several minutes, so they lag acute changes.

Fick: CO = VO₂ ÷ (CaO₂ − CvO₂) × 10. VO₂ is rarely measured outside the cath lab, and the assumed 125 mL/min/m² can be badly wrong in fever, sepsis or sedation, so indirect Fick is a cross-check. The normal arteriovenous content difference is about 4–6 mL/dL and widens as output falls.

Low SvO₂: delivery is not meeting demand (low output, hemoglobin or SaO₂, or fever, shivering, seizures, pain, work of breathing). High SvO₂ above 80%: tissues are not extracting (sepsis, cyanide from nitroprusside, hypothermia, left-to-right shunt) or a wedged catheter is sampling arterialized blood. A normal ScvO₂ does not exclude hypoperfusion; a venous-to-arterial PCO₂ gap above 6 mmHg suggests flow is still inadequate. ProCESS 2014, ARISE 2014 and ProMISe 2015 found no mortality benefit from ScvO₂-targeted early goal-directed therapy over usual care.

Derived values and worked profiles

CPO is the heart's hydraulic work. In the SHOCK trial registry it was the strongest hemodynamic predictor of in-hospital death, with a cutoff of about 0.53 W (Fincke 2004). PAPi asks how well the RV turns filling pressure into pulsatile PA pressure: under 0.9 predicted RV failure or death in inferior MI (Korabathina 2012), and under about 1.85 predicted RV failure after LVAD (Morine 2016). Thresholds vary by population; treat them as flags.

Worked example: BP 84/52 (MAP 63), HR 110, CVP 14, PA 48/28 (mean 35), PAOP 26, CO 3.2 L/min, BSA 2.0. CI = 1.6 and SV ≈ 29 mL. SVR = (63 − 14) × 80 ÷ 3.2 ≈ 1,225. PVR = (35 − 26) × 80 ÷ 3.2 = 225 (2.8 WU) with a diastolic gradient of 2, so the pulmonary hypertension is post-capillary. CPO = 63 × 3.2 ÷ 451 ≈ 0.45 W; PAPi = 20 ÷ 14 ≈ 1.4; RAP:PAOP ≈ 0.54. Reading: LV-predominant cardiogenic shock with inadequate power and an RV still coping, the profile in which inotropes or LV mechanical support are considered.

Minimally invasive monitors, echo and fluid tests

These devices follow trends better than absolute values; uncalibrated pulse contour often misses the 30% percentage-error standard for agreement with thermodilution (Critchley 1999) in vasoplegia and arrhythmia. PPV above 13% predicts response (Michard 2000), but 9–13% is a gray zone holding about a quarter of patients (Cannesson 2011). PPV and SVV also need a heart rate to respiratory rate ratio above about 3.6 and no RV failure, beyond the conditions on the one-pager, and point-prevalence studies (Mahjoub 2014, Delannoy 2016) found few ICU patients meeting all of them. Under 6 mL/kg ventilation, a tidal volume challenge restores some value.

Echo questions a nurse can follow: is the LV squeezing well, poorly or hyperdynamically; is the RV larger than the LV or the septum D-shaped; is there pericardial fluid with RA or RV diastolic collapse? IVC under 2.1 cm collapsing more than 50% on a sniff suggests RAP about 3, over 2.1 cm collapsing less suggests about 15 (ASE 2015); the IVC is unreliable on the ventilator. LVOT VTI is normally about 18–22 cm.

Only about half of ICU patients given fluid raise their output (Michard and Teboul 2002). The leg raise moves about 300 mL of venous blood into the chest and is reversible; the Monnet 2016 meta-analysis gave sensitivity 85% and specificity 91% when judged by output, and much worse by pulse pressure. Pain, coughing or waking during it give false positives. Each test's threshold must exceed its monitor's noise: a 5% rise shows on continuous pulse contour but not reliably on echo. Responsiveness is not an indication; also ask whether low flow is the problem and whether the patient tolerates more volume.

PA catheter complications and evidence

• Insertion: ectopy (common), sustained VT (rare), RBBB, pneumothorax, arterial puncture, air embolism.
• Indwelling: infection, thrombosis, valve injury, knotting; remove when the data no longer change management.
• Balloon: pulmonary infarction from persistent or spontaneous wedge; PA rupture (rare, often fatal); balloon rupture.

Routine PAC placement does not improve outcomes in general ICU, surgical, ARDS or heart failure patients, and ESCAPE recorded more in-hospital adverse events with it (about 22% vs 12%). It is a monitor, so outcome depends on reading it well; Iberti 1990 found nearly half of surveyed physicians could not identify a wedge on a clear tracing. Use has narrowed to cardiogenic shock (staging, choosing and weaning mechanical support, judging the RV), pulmonary hypertension, transplant and LVAD care, and complex cardiac surgery. Whether PAC-guided care improves survival in cardiogenic shock is unsettled: the supporting data are observational.

Exam pearls

• Transducer below the phlebostatic axis reads high, above reads low; zero to air, level to the 4th intercostal space at mid-chest.
• Square wave: 1.5–2 oscillations; overdamped underreads SBP, underdamped overreads it; MAP is the most trustworthy number.
• Read CVP and PAOP at end-expiration on a frozen trace, at the mean of the a wave.
• PAD exceeding PAOP by more than about 5 mmHg suggests high PVR; PAOP above PAD suggests a non-zone-3 tip.
• Cannon a waves: AV dissociation. Giant v waves on the wedge: mitral regurgitation. Equal CVP, PAD and PAOP: tamponade.
• Too little or warm injectate gives a falsely high thermodilution CO.
• High SvO₂: sepsis, cyanide, hypothermia, wedged catheter.
• CPO under 0.6 W is a poor sign in cardiogenic shock; PAPi flags RV failure.
• Routine PAC does not lower mortality (PAC-Man, ESCAPE, FACTT).