Bedside primer

Principles of RHC

Everything you need to finish the lab cases: how a Swan-Ganz is zeroed and wedged, what each PA-catheter waveform is supposed to look like, how the wedge lies, how to name a shock profile, and how CPO, PAPi, and the ESC/ERS PH rules keep you from treating the wrong ventricle.

Catheter path, balloon, and the zero

A balloon-tipped pulmonary-artery catheter is floated from a central vein (usually RIJ) through the RA → RV → PA. The distal lumen sits in a PA branch. Inflating the balloon occludes that branch so the distal port sees pulmonary-capillary / left-atrial pressure (PCWP). Proximal ports read RA (CVP) and, on some catheters, RV.

Every number is only as good as the transducer. Zero at the phlebostatic axis: 4th intercostal space, mid-axillary / mid-chest, patient supine. A 10 cm height error is about 7.5 mmHg — enough to misclassify pulmonary hypertension or invent hypovolemia. If every chamber is impossibly low and the MAP is fine, re-zero before you treat.

Read pressures at end-expiration (the moment pleural pressure is closest to zero). Digital means average the whole cycle and will lie on a swinging ventilator.

Normal PA-catheter waveforms

The console in this lab redraws four traces as the physiology changes. Learn the pictures here, then open a case and name the chamber before you believe the digital mean. A Swan is only as honest as the waveform it is sitting in.

What each port is for

Four lumens, four jobs. The balloon is only inflated to get to the wedge and to confirm it — not to live there.

PortSits inWhat you readWhat it estimates
Proximal / RARACVP: a, x, v, yRV preload. Not LV preload.
RV (if present)RVRV sys / EDPRVEDP should ≈ RA. Peak ≈ PA sys.
Distal, balloon downPA branchPA sys / dias / mean, notchRV afterload. PAD ≈ PCWP if PVR is low.
Distal, balloon upOccluded PAPCWP: delayed a / vLA pressure ≈ LVEDP if the MV is open.

How the four traces line up on one beat

Read them against the ECG in your head, not as four independent graphs.

  • a wave follows the P wave — atrial kick. On RA it fills the RV. On PCWP it fills the LV.
  • c wave (often tiny) is the AV valve bulging back into the atrium as the ventricle starts to contract.
  • x descent is atrial relaxation plus the floor of the atrium being pulled down. Preserved in tamponade.
  • v wave is the atrium filling while the AV valve is closed (ventricular systole). Giant v = the valve is not competent (TR on RA, MR on PCWP).
  • y descent is the atrium emptying when the valve opens. Blunted in tamponade. Brisk in constriction and after a giant v.

PCWP is an RA-like tracing, delayed by transit through the pulmonary veins: a after QRS, v after T. If your “wedge” still has a dicrotic notch, you are in the PA.

Numbers that have to agree

CheckShould beIf it is not
RV peak vs PA sysNearly equalPulmonic stenosis, or you are not in the PA
RVEDP vs RANearly equalCatheter whip, wrong port, or not end-expiratory
PCWP vs PADPCWP ≤ PADOverwedge, still in PA, or a giant v you averaged
PCWP vs LVEDPClose if no MSMitral stenosis (PCWP > LVEDP), stiff LV
  • RA / CVP
  • RV
  • PA
  • PCWP
Normal float — four chambers, one beatsinus · end-expiration
RA / CVP
7
axvy
RV
26/6
sysEDP
PA
26/12
sysnotchdias
PCWP
11
axvy
RA mean in the single digits, a a bit taller than v. RV shoots up from a near-zero diastolic and falls back to an EDP that matches RA. PA keeps the same systolic peak, adds a dicrotic notch, and its diastolic stays well above RVEDP. Wedge is an RA-like tracing at or just under PAD. This is the only stack you treat as a baseline.
Normal RA / CVPsinus · end-expiration
RA / CVP
7 · a 9 · v 7
axvy
a — atrial kick into the RV (follows P). x — atrial relaxation / descent. v — atrial filling while the tricuspid is closed. y — emptying into the RV when the TV opens. Normally a ≥ v on the right. Mean 3–8 mmHg in a euvolemic adult. A blunted y with a preserved x is tamponade language. Giant v waves here are TR, not MR.
Normal RVsinus · end-expiration
RV
26/6
sysEDP
Near-zero diastolic, a rapid systolic upstroke, then a fall almost to baseline. There is no dicrotic notch — that is how you know you have not reached the PA yet on a float. RVEDP should sit near RA. A high, square diastolic plateau that matches RA and PCWP is equalization (tamponade / constriction), not a “stiff RV.” Peak RV that never matches PA systolic means you are not where you think.
Normal PAsinus · end-expiration
PA
26/12 (17)
sysnotchdias
Systolic peak close to RV systolic (no pulmonic stenosis). A dicrotic notch as the pulmonic valve closes. Diastolic stays well above RV diastolic — the pulmonary vascular bed does not empty the way the RV cavity does. Normal is roughly 15–30 / 4–12, mean ≤20. PA diastolic is your first check on a claimed wedge: PCWP should be at or just below PAD, never above it for long. A tall PA with a low wedge is pre-capillary (PE, Group 1 / 3 / 4), not left heart.
Normal PCWPsinus · end-expiration
PCWP
11 · a 13 · v 12
axvy
An RA-like tracing, slightly delayed: a and v waves, x and y descents. Normally a and v are similar; v may be a few mmHg taller. Mean wedge 6–12 mmHg estimates LA pressure and, when the mitral valve is open and there is no MS, LVEDP. Read the waves, not only the digital mean — the monitor averages the giant v into a polite 18 and hides acute MR. Confirm every surprising wedge (morphology, distal sat, fluoro) before you diurese or start a pulmonary vasodilator.

Abnormal PCWP — the tracing that changes therapy

Most of the harm from a Swan comes from believing a bad wedge. The left-atrial tracing is where volume, MR, tamponade, occult HFpEF, and artifact all pretend to be each other. Name the shape before you name the mean.

A one-pass read of every wedge

  1. Is it even a wedge? a/v morphology, no dicrotic notch, distal sat ≥95%, balloon in a distal PA. A leftover PA tracing is the commonest fake.
  2. Is it overwedged? Damped, too high, often above PAD. Deflate, pull a centimeter, reinflate gently.
  3. Read a and v, not the mean. Giant v is MR (or a stiff LA). Giant a is MS, a stiff LV, or a cannon wave. No a is AF.
  4. Read the y descent. Blunted = tamponade (or MS). Brisk / dip-and-plateau = constriction or a giant v emptying.
  5. Compare to PAD and RA. PCWP stays at or below PAD; a wedge above PAD is almost never real. PCWP ≈ RA and a blunted y is equalization. PCWP low + RA high is the right heart (PE, RV infarct).
PCWP pictureMeanThe tellFirst move
Normal6–12a ≈ v, y present, ≤ PADBelieve it
High, preserved a/v18–30Both waves up, wet lungsUnload / diurese
Low, small a/v<6Dry, or the zero is on the floorFill — after you re-zero
Giant vMean lies lowv to 40–50, brisk yAfterload ↓. Echo. Surgery if papillary.
Giant a / cannon aMean middlinga ≫ vMS, stiff LV, or AV dissociation
No a waveVariableIrregular v onlyAF. Do not invent an a.
OverwedgeFalsely highDamped, often > PADDeflate, withdraw, retry
Still in PALooks like PADNotch still there, mixed-venous satYou never wedged
TamponadeEquals RAy gone, x keptDrain. Do not fill for long.
ConstrictionEquals RADeep y, square rootNot tamponade. Imaging / peel.
Occult HFpEF12 at restJumps on 500 mL or bikeDo not start Group-1 therapy
High PCWP — left heart is fullhigh
PCWP
26 · a 27 · v 29
axvy
Morphology is still honest — a and v are both there, y is present — the whole tracing has just been lifted. This is the wet, cold (or wet, warm) left heart: infarct, decompensated HFrEF, iatrogenic volume. Mean 18–30. PAD is also high; the gradient PAD − PCWP stays small. Diurese / unload / afterload-reduce. Do not give a liter for the kidneys. A high mean with a giant v on top is MR plus a full LA — treat both.
Low PCWP — empty, or the zero is wronglow
PCWP
4
axvy
Small a, small v, mean in the low single digits. If RA is also low and the patient looks dry, fill (and think blood). If every chamber is impossibly low and the cuff MAP is fine, the transducer is on the floor — re-zero at the phlebostatic axis before you invent hypovolemia. A low wedge with a high RA is not empty: that is PE, RV infarct, or pulmonary vascular disease. Look at the PA pulse and the PAPi.
Giant v waves — acute MR (or a stiff LA)largeV
PCWP
mean 22 · v 48
avy
The digital mean averages the story away. A v wave to 40–50 mmHg after an infarct is acute MR until echo disagrees — papillary muscle or ischemic leaflet. Less often a non-compliant LA (diastolic HF, MS). The y descent is brisk because the LA dumps into the LV when the mitral valve finally opens. Afterload reduction shrinks the regurgitant fraction; a pure pressor enlarges it. Papillary-muscle rupture is a surgical disease. The catheter only names it. Do not confuse this with a PA tracing: there is no dicrotic notch, and there is still an a wave if the patient is in sinus.
Large a wave — MS, stiff LV, or nowhere to kicklargeA
PCWP
mean 18 · a 28 · v 16
axvy
The atrium is kicking against a closed or stenotic mitral valve, or into a rock-hard LV. Classic mitral stenosis: large a, slow y (impaired LA emptying), and PCWP that overestimates LVEDP. HFpEF / LVH gives a large a with a faster y. Read the valve on echo before you treat the number as LVEDP. A pulmonary vasodilator on an MS wedge floods the lung.
Cannon a waves — atrium versus a closed valvecannon
PCWP
a 34 · v 14
axvy
Intermittent or beat-to-beat giant a waves when the atrium contracts against a closed mitral valve — complete heart block, VT, VVI pacing with VA conduction. The same physics on the RA tracing is a cannon a into a closed tricuspid. Treat the rhythm, not the “high filling pressure.” The mean will jump around with the next dissociated P wave; do not diurese a cannon.
Atrial fibrillation — no a waveafib
PCWP
16 · v only
vy
No P wave, no a wave. What remains is an irregular v and a y descent. Do not label a random bump “a.” The mean is still usable if the balloon is truly wedged; beat-to-beat variation is expected. A giant v in AF is still MR. Loss of atrial kick is why the same LV looks worse the night they flip.
Overwedge — the balloon sees itselfoverwedge
PCWP
34 · damped
Flat, overdamped, too high — often above PAD, which is physically hard to defend. The balloon is overinflated in a small branch and the distal port sees balloon pressure, not LA. Distal blood, if you can draw it, is not a clean arterial sat. Deflate fully, withdraw a centimeter, reinflate slowly until the tracing just becomes a/v. Treating a fake wedge of 34 with diuretics creates real hypovolemia.
Never wedged — leftover PA on the PCWP channelstillPa
PCWP
32/16 · notch
sysnotchdias
Dicrotic notch, systolic peak, no a/v. You floated to the PA and called it a wedge. Distal sat will be mixed-venous. Fluoro shows the balloon still in a large PA. This is how people invent “combined pre- and post-capillary PH” out of a PAD. Re-wedge, or believe the PAD as PAD and stop calling it PCWP.
Tamponade — y descent gone, filling equalizedtamponade
PCWP
18 · y∅
axvy∅
RA ≈ RVEDP ≈ PAD ≈ PCWP. On the atrial tracings the y descent is blunted (x is often preserved) — the ventricle cannot accept volume in early diastole because the pericardium is already full. Pulsus, a vanishing pulse pressure, and a climbing CVP after cardiac surgery do not wait on a formal lab. Fluids are a bridge of minutes. Drain the pericardium. Filling the tank will not restore a y.
Constriction — deep y, square-root signconstrict
PCWP
18 · deep y
avy
Equalized filling like tamponade, opposite atrial shape. The y descent is brisk — early diastolic filling is abrupt, then the rigid shell stops it (dip and plateau / square-root sign). Kussmaul (RA that rises with inspiration) leans constriction, not tamponade. This is the tracing that keeps people from taking a needle to a calcified pericardium. Imaging and a surgeon, not a drain-and-see.
Low PCWP, tall PA — the left side is not the problemstrain
PA
58/28
sysnotchdias
PCWP
9
axvy
Wedge is low or normal; PA systolic is not. The gradient lives in the pulmonary circuit: massive PE, Group 1 / 3 / 4 PH, hypoxic vasoconstriction. RA is high, PAPi is falling, the LV Impella (if there is one) will start to suck. Do not diurese a low wedge. Do not start a pulmonary vasodilator until you have confirmed the wedge is real and ≤15. A leftover PA called “wedge” will send someone home on the wrong drug.

Two more ways a wedge lies

  • Not end-expiration. Spontaneous breathing: inspiration pulls the number down — read the high point (end-expiration). Positive-pressure ventilation: inspiration pushes the number up — read the low point (end-expiration). Digital means average the swing and will misclassify PH or invent hypovolemia on a 15 cmH₂O PEEP.
  • Resting 12, stressed 28. Occult HFpEF. A pretty resting wedge does not clear the left atrium. Supine bike past ~25 mmHg, or a 500 mL fluid challenge that jumps PCWP above 18 with little CO gain, is the same disease. One bolus is a test. The second is a treatment you did not mean to write. Do not start Group-1 therapy on a resting tracing and a story.

Looks like X, is actually Y

You seeCould beWhat splits them
High mean wedgeLV failurePreserved a/v, wet, high PAD ≈ PCWP
High mean wedgeOverwedgeDamped, > PAD, sat not arterial
High mean wedgeGiant v averagedLook at the paper. v ≫ mean.
High mean wedgeStill in PANotch present, mixed-venous sat
Giant peak on wedgeMR v wavea still there, no notch, brisk y
Giant peak on wedgeCannon aTracks the P wave, not systole
Equalized diastolicTamponadey blunted, pulsus, post-op / bleed
Equalized diastolicConstrictionDeep y, Kussmaul, pericardial knock / Ca²⁺
Low wedgeHypovolemiaRA also low, dry exam
Low wedgePE / RV failureRA high, tall PA, PAPi < ~1
Low wedgeTransducer on the floorEvery chamber low, MAP fine

Confirm every surprising wedge

A real wedge has three votes. Need two of three if the number is going to change therapy:

  • Waveform — a/v morphology, not a damped flat line and not a leftover PA tracing.
  • Saturations — blood from the distal port should look arterial (typically 95% or higher). A mixed-venous sat means you are still in the PA.
  • Fluoro / CXR — balloon in a distal PA, not overinflated into a tiny branch (overwedge) and not still in the main PA.

Overwedge: the balloon sees itself. The number is too high, often above PAD, and the tracing is damped. Deflate, withdraw a centimeter, reinflate gently. Treating a fake wedge of 34 with diuretics creates real hypovolemia.

Cardiac output: thermodilution vs Fick

Thermodilution injects cold (or room-temp) indicator through the RA port and reads the PA thermistor. Average three decent curves. It fails when the indicator recirculates or never mixes — severe TR wrecks TD, as do shunts, very low flow, and an injection in the wrong port.

Fick is VO₂ / (CaO₂ − CvO₂). Use a measured VO₂ when you can; assumed VO₂ is a guess. In a left-to-right shunt, TD over-reads pulmonary flow (Qp). Know which cardiac output you are quoting.

This lab recomputes the derived set on every tick:

  • SVR = (MAP − RA) / CO × 80
  • CI = CO / 2.0 (BSA assumed 2.0 m²)
  • CPO = MAP × CO / 451
  • PAPi = (PAsys − PAdias) / RA
  • PVR = (mPAP − PCWP) / CO  ← Wood units

Four shock profiles

Name the column before you pick a drip. Filling pressures, flow, and SVR do most of the work; the waveform tells you when the column is a lie.

RA / PCWPCOSVRFirst lever
HypovolemicBoth lowLowHighFill (and blood)
CardiogenicPCWP highLowHighInotrope / unload
DistributiveNormal-lowHighLowPressor + source
ObstructiveRA high, PCWP low*LowHighRelieve the obstruction

*Tamponade equalizes RA and PCWP — it is obstructive with a different tracing. Massive PE keeps the wedge low.

Forrester: cold-wet vs warm-dry

Forrester plotted CI against PCWP after infarct. The bedside version:

  • Warm-dry — CI adequate, wedge low. Leave them alone, or they are empty.
  • Warm-wet — CI adequate, wedge high. Diurese / vasodilate.
  • Cold-dry — CI low, wedge low. Fill carefully, then inotrope. RV infarct lives here.
  • Cold-wet — CI low, wedge high. The classic cardiogenic box. Unload. Do not give a liter for the kidneys.

CPO and PAPi

Two dimensionless-feeling numbers that are not dimensionless, and that decide devices.

Cardiac power output (Fincke, SHOCK trial): CPO = MAP × CO / 451, in watts. A left ventricle that cannot deliver 0.6 W is in trouble. CPO is also low in empty hypovolemia — it is a product, not a diagnosis. Read it with the wedge.

PA pulse-pressure index (Korabathina): PAPi = (PAsys − PAdias) / RA. The RV that cannot generate a pulse pressure over a high RA has a PAPi below 1.0. That is the number people use when they talk about RV MCS.

Both down at once is biventricular crash. Stacking drips is not a plan.

PH classification (ESC/ERS 2022)

Pulmonary hypertension is mPAP above 20 mmHg. The wedge and the PVR decide the compartment:

PCWPPVRWhat it is
Pre-capillary≤15>2 WUGroup 1 / 3 / 4 pattern
Isolated post-cap (Ipc)>15≤2 WULeft heart — diurese
Combined (Cpc-PH)>15>2 WULeft heart plus a pre-cap component

Confirm the wedge before you classify. A pulmonary vasodilator on a high LA pressure can flood the lung. Diurese Cpc-PH first; the residual PVR is the pre-capillary piece.

Giant v waves and acute MR

The digital mean PCWP averages away the story. A v wave to 40–50 mmHg after an infarct is acute MR (or, less often, a stiff LA) until echo disagrees. Afterload reduction shrinks the regurgitant fraction; a pure pressor enlarges it. Papillary-muscle rupture is a surgical disease — the catheter only names it.

Tamponade equalization

RA = RVEDP = PAD = PCWP is an equalized diastolic plateau. On the RA tracing the y descent is blunted (x is often preserved). Pulsus, a vanishing pulse pressure, and a climbing CVP after cardiac surgery do not wait on a formal lab. Fluids are a bridge of minutes. Drain the pericardium.

Exercise RHC and the fluid challenge

A resting PCWP of 12 does not clear the left atrium. Occult HFpEF is diagnosed when the wedge rises on stress:

  • Supine bike — PCWP rising past ~25 mmHg is the usual teaching cutoff.
  • 500 mL fluid challenge — when the patient cannot pedal. PCWP jumping above 18 with little CO gain is the same disease. One bolus is a test; the second is a treatment you did not mean to write.

Do not start Group-1 therapy on a pretty resting tracing and a story.

O₂ sat run

Sample SVC, IVC, RA, RV, and PA. A step-up of about ≥7% from the cavae into the RA is the classic atrial-level left-to-right screen (ASD, anomalous pulmonary return). A PA sat in the mid-80s without fever is not “great cardiac output” — it is arterial blood arriving early. Thermodilution then quotes Qp, not systemic CO.

High-output failure

High CO, low SVR, high filling pressures: the heart is overworking into a low-resistance circuit. Drivers to name out loud — ESLD, large AV fistula, thyrotoxicosis, beriberi, severe anemia, sepsis-overlap. Stop the fluids. Echo is hyperdynamic, not empty. An inotrope on a 9 L heart is not a plan.

When to escalate MCS

Think about a device when:

  • CPO stays below 0.6 W after you have unloaded and inotroped.
  • PAPi stays below 1.0 and the RA is climbing (RV MCS / biventricular plan).
  • Lactate and urine are still going the wrong way on two drips.
  • You are about to add a third pressor to a wet, cold patient.

The catheter does not pick Impella versus VA-ECMO versus a surgical pump. It tells you which ventricle has left the building.

How to work a case in this lab

Same order every time — it is the order that keeps people alive at 03:00.

  1. Look at the patient: warm or cold, wet or dry, urine, lactate.
  2. Look at the tracing: RA a/x/v/y, PA notch, wedge morphology. Confirm any number that would change the drug.
  3. Name the profile in one sentence (empty, cold-wet, vasoplegic, obstructive, pre-cap, artifact…).
  4. Do the matching move. Harmful clicks treat the wrong chamber.
  5. Untreated tamponade, massive PE, or a crashed CPO/PAPi will keep dropping MAP and climbing lactate until you fix the cause.

Guided mode forces three decisions. Free play is the same physiology without the multiple choice. The debrief is the attending in the doorway.