Pericardial Drainage (Window / Pericardiocentesis)
Updated On: July 23, 2026
Anesthetic Approaches
Drainage before induction - If tamponade physiology is still present at induction, you own the arrest that follows. Push for pericardiocentesis or a subxiphoid window under local, awake and sitting up; the sac is non-compliant, so even a small volume off transforms the hemodynamics.
Hemodynamic goals - Full, fast and tight. Stroke volume is fixed and small because the ventricle cannot fill, so output is rate times a number you cannot change. Volume load, welcome a sinus tachycardia, and hold afterload with a vasoconstrictor rather than with anesthetic depth.
Head-up until drainage - A tamponade patient propped upright often cannot lie flat, and laying them supine for your convenience can drop the pressure before any drug is given. Keep the head up at 30 to 45 degrees and lay them flat only after decompression.
Induction as arrest trigger - Propofol and volatile agents venodilate and cut preload, drop systemic vascular resistance, and blunt the sympathetic tone holding the tachycardia up. Remove all three together and a compensated tamponade becomes pulseless electrical activity inside a minute.
Positive pressure ventilation - Raised mean airway pressure cuts the gradient for venous return when nothing else is filling the ventricle, so spontaneous ventilation is protective. Intubate without a relaxant and let them breathe; if ventilation is unavoidable, low tidal volumes, low rate, minimal PEEP.
Room setup before induction - If general anesthesia is unavoidable, sequence the room before the drugs: surgeon scrubbed, patient prepped and draped for a possible sternotomy, pads on, vasopressor running. Induction to decompression is measured in seconds.
Drug choice - Ketamine supports rate, contractility and vascular tone while preserving spontaneous ventilation, which makes it first choice; etomidate is the alternative when minimal myocardial depression is the priority. Give either in small increments.
Drugs to avoid - Avoid anything that slows the rate: no beta blocker, no large opioid bolus with its vagal bradycardia. Output is rate times a fixed stroke volume, so treat a falling heart rate as an emergency to correct rather than an observation to note.
Access and monitoring - Two large-bore IVs and an arterial line sited awake under local, so you watch the pressure beat to beat through induction. A central line is not worth the wait, and if the patient is crashing the drainage goes first and the lines follow.
Falsely high filling pressures - Do not be reassured by a high central venous pressure. Intrapericardial pressure is transmitted to every chamber, so right and left filling pressures equalize and read high while true transmural preload is very low.
Full stomach conflict - These arrive unfasted, and the standard answer of paralyze, ventilate and secure is exactly the combination that arrests them. Decide which risk you accept: draining under local first dissolves it, otherwise keep the dose small, keep them breathing, and have suction and a second pair of hands at the head.
The swing after drainage - Output can jump within seconds, so a patient you were propping up on a vasopressor turns hypertensive: deepen the anesthetic and wind support down rather than chase the old number. Decompression also triggers vasovagal bradycardia, so have an antimuscarinic ready.
Pericardial decompression syndrome - Watch for paradoxical collapse or pulmonary edema after an uncomplicated drainage, most likely when a large effusion is emptied quickly. Take off only enough to break tamponade and let the rest drain slowly through the catheter, patient in a monitored bed.
Fluoroscopy / Xray (general considerations): Have lead aprons and thyroid shields available. Alternatively, distancing yourself 3 to 6 feet will reduce scatter radiation to 0.1% to 0.025% respectively. Occupational maximum exposure to radiation should be limited to a maximum average of 20 Sv (joules per kilogram - otherwise known as the Sievert/Sv) per year over a 5 year period. Limits should never exceed 50 Sv in a single year.
Arterial line (general considerations): Preoperatively check pulses to gauge the best side to attempt the A-line. Perform an Allen test to ensure adequate blood flow. Have the A-line equipment set up and ready in the room.
Off-site (general considerations): Extra precaution should be taken preoperatively if the surgery is planned off-site (away from the OR or your normal work spaces). Depending on your facility setup, it's wise to take emergency airway equipment and drugs with you that would be necessary to treat the gamut of anesthesia emergencies.
The pericardium is a stiff fibroelastic sac normally holding 15 to 50 mL of serous fluid. Its compliance, not the volume alone, decides whether an effusion matters. Fluid that collects over weeks stretches the sac and 1 to 2 liters can be tolerated, while 100 to 150 mL arriving acutely from a stab wound, a ventricular rupture or a post-cardiac-surgery bleed produces tamponade. Once intrapericardial pressure exceeds chamber diastolic pressure, diastolic filling stops, right and left filling pressures equalize, and stroke volume becomes small and fixed. Cardiac output then rests on the only two variables left: heart rate, and the venous pressure gradient still pushing blood into the chest. Sinus tachycardia and raised sympathetic tone are the compensations. Etiology spans malignancy, uremia, infection, autoimmune disease, post-cardiac-surgery bleeding and penetrating trauma.
Diagnosis and urgency - Tamponade is a clinical diagnosis confirmed by echocardiography. The findings are a swinging heart in a fluid-filled sac, right atrial systolic collapse, right ventricular diastolic collapse, septal shift toward the left ventricle, respiratory variation across the mitral and tricuspid valves, and a plethoric inferior vena cava with no respiratory variation. Beck's triad of distended neck veins, muffled heart sounds and low blood pressure is present in only about 30 percent of tamponade from trauma, so its absence proves nothing.
Choosing the route - Needle pericardiocentesis is the fastest option and can be done at the bedside, in the emergency department, in the ICU or in the cath lab. A surgical window is chosen when the effusion is loculated, when tissue is needed for diagnosis, when trauma is suspected, or when recurrence is expected and a permanent drainage channel is wanted. In penetrating trauma the modern preference has moved away from needle drainage toward prompt operative decompression, with pericardiocentesis used only to buy time.
Pericardiocentesis, subxiphoid approach - The head of the bed is elevated to 30 to 45 degrees so gravity pools fluid at the access point, unless hemodynamics forbid it. The skin is infiltrated with lidocaine 1 to 2 percent, 5 to 10 mL, if time allows. An 18 gauge spinal needle enters 2.5 cm below the xiphoid and 2.5 cm to the patient's left, angled at 45 degrees or less to the skin and directed toward the left shoulder, and is advanced with continuous aspiration under ultrasound vision. Agitated saline can be injected to confirm tip position if the operator loses the needle.
Pericardiocentesis, parasternal approach - The patient is placed supine or head down to pool fluid into the needle path. The effusion is identified in the 4th or 5th intercostal space between the sternum and the midclavicular line, with the operator confirming that lung is not in the ultrasound window. An 18 gauge spinal needle enters at 90 degrees to the skin and is advanced toward the patient's back under direct vision.
Drain placement - Even when the initial aspiration produces immediate improvement, a catheter is usually left in the pericardial space over a guidewire. It allows gradual removal of the remaining fluid, serial drainage of a re-accumulating effusion, and early detection of fresh bleeding. It is typically removed once daily return falls below roughly 30 to 50 mL.
Subxiphoid pericardial window - The entire chest is prepped for a possible sternotomy before the incision is made. A vertical midline incision is made over the xiphoid and upper epigastrium, the xiphoid is elevated or excised, and the anterior mediastinum and pericardiophrenic membrane are exposed. The pericardium is opened between two stay sutures and inspected. Fluid is evacuated and sent for cytology, culture and chemistry, and a pericardial biopsy is taken when the etiology is unknown.
Thoracoscopic window - A video-assisted thoracoscopic or anterior mini-thoracotomy window creates a larger and more durable communication draining the pericardium into the pleural space. It is favored for recurrent malignant effusions because it recurs less often than needle drainage, at the cost of single-lung ventilation and a chest tube.
Escalation in trauma - If blood is found on opening the pericardium in a trauma patient, the operation converts immediately to a median sternotomy, which gives the best exposure of the heart, great vessels and pulmonary hila. A left anterior or anterolateral thoracotomy in the 5th intercostal space is the alternative and the route used for emergency department thoracotomy. The bleeding heart is controlled with digital occlusion and the laceration closed with mattress sutures, taking care not to occlude a coronary.
Penetrating chest trauma changes the resourcing, not the physiology - A stab wound with tamponade is going to the operating room, and the subxiphoid window is often only the diagnostic prelude to a sternotomy or thoracotomy for a cardiac laceration. Plan for that escalation from the start: activate the massive transfusion protocol pathway, get blood in the room rather than in the bank, have a rapid infusion device and a forced-air or fluid warmer set up, and expect coagulopathy and hypothermia to be what actually kills rather than the laceration itself. Patients who reach the operating room stable do well, so the anesthetic job is to keep them stable long enough to get the pericardium open.
Malignant effusion is a different patient with the same physiology - These are recurrent, palliative, often frail patients on cytotoxic or immunotherapy regimens with poor functional reserve, and the effusion has usually built slowly to a large volume. Two things follow for you. First, a large chronic effusion is the highest-risk group for pericardial decompression syndrome, so drainage should be staged rather than emptied in one sitting. Second, the goals of care conversation belongs before the procedure and not after an arrest - confirm resuscitation status and how far escalation is intended to go while the patient can still tell you, because a drainage under local is a very different commitment from an induction with a surgeon standing by.
Much of this happens outside your operating room - Pericardiocentesis is frequently done in the cath lab, the ICU or the emergency department under fluoroscopy or bedside echocardiography, sometimes with you called in only when sedation is requested. Treat it as a full off-site anesthetic: bring your own monitoring, confirm suction and oxygen, know where the defibrillator is, and establish who is scrubbing if the needle lacerates a ventricle. Complication rates for pericardiocentesis run around 5 percent even in experienced hands, and the serious ones - myocardial or coronary laceration, iatrogenic hemopericardium, pneumothorax, arrhythmia - all present immediately and all land on you.
What you hand over - The effusion recurs, and the drain is the only early warning. These patients belong in a monitored bed with follow-up echocardiography, and the handover should say explicitly what the hemodynamics were before drainage, how much came off and how fast, what vasoactive support was running and whether it has been weaned, and whether decompression syndrome remains a live concern. If they came in coagulopathic or thrombocytopenic - common in uremia and in malignancy - flag it, because re-accumulation in that group is bleeding until proven otherwise. Verify anticoagulant and antiplatelet management against your institution's current guidance rather than from memory.