Bedside Paracentesis

Paracentesis is one of the safest and highest-yield procedures in clinical medicine. Every patient admitted to the hospital with cirrhosis and ascites should be considered for diagnostic paracentesis at the time of admission — regardless of whether spontaneous bacterial peritonitis (SBP) symptoms are present, because up to one-third of SBP cases are clinically silent.

Indications

Diagnostic

New-Onset Ascites

First presentation of ascites — determine etiology (cirrhotic vs. malignant vs. cardiac)
Serum-ascites albumin gradient (SAAG) requires simultaneous serum and ascites albumin
Identify unexpected infection before symptoms develop

Diagnostic

Clinical Deterioration

Any hospitalized patient with cirrhosis and ascites — rule out SBP at admission
Fever, abdominal pain, encephalopathy, or leukocytosis in a patient with known ascites
Unexplained acute kidney injury (AKI) or hemodynamic instability in a patient with cirrhosis

Therapeutic

Large-Volume Paracentesis (LVP)

Tense ascites causing respiratory compromise, discomfort, or early satiety
Refractory ascites not responding to maximal diuresis
Typically 4–6 L removed per session; albumin replacement required when >5 L is removed (6–8 g per litre of total volume removed)

Contraindications

There are no absolute contraindications to paracentesis in a patient with clinically apparent ascites. The following are relative, and the diagnostic or therapeutic benefit almost always outweighs the risk.

Relative contraindications — discuss risk vs. benefit, do not automatically defer:

Coagulopathy (international normalized ratio [INR] > 2.0 or platelet count < 50,000/mm³) — American Association for the Study of Liver Diseases (AASLD) does not recommend routine correction of coagulopathy with fresh frozen plasma (FFP) or platelets prior to paracentesis. Complication rates are not meaningfully increased at these thresholds. Correction is only warranted if disseminated intravascular coagulation (DIC) is suspected.
Prior abdominal surgeries — bowel adhesions alter normal anatomy; use ultrasound (US) guidance and select a site away from scars
Active skin infection at intended site — select an alternate site; do not puncture through cellulitis or active wound
Pregnancy — perform with real-time US guidance; use sites away from fundal height
Ileus or significantly dilated bowel — US-guided confirmation of fluid pocket is essential
Uncooperative patient — discuss goals of care; procedural sedation may be appropriate in select cases

⚠ Coagulopathy in cirrhosis: Standard laboratory tests (INR, prothrombin time [PT]) overestimate bleeding risk in patients with cirrhosis. The liver simultaneously loses procoagulant and anticoagulant synthetic function, producing a rebalanced hemostasis that does not reliably predict procedure-related bleeding. Prophylactic transfusion in this setting exposes the patient to transfusion risk without clear benefit and may worsen portal hypertension.

The consent discussion for paracentesis should be brief, accurate, and patient-centered. Most patients with cirrhosis and recurrent ascites will have had this procedure before — acknowledge that and focus on what is new or different about this encounter (e.g., first time, large volume, infected fluid concern).

What to Explain to the Patient

Purpose

Why We Are Doing This

For a diagnostic tap: "We're going to take a small sample of the fluid in your abdomen to check for infection and to help us understand what's causing it." For a therapeutic tap: "We're going to drain a larger amount of the fluid to relieve your discomfort and make it easier to breathe."

Process

What Will Happen

We will use ultrasound to find the best spot, numb the skin with local anesthesia, and then insert a thin needle or catheter to collect or drain the fluid. The procedure usually takes 10–20 minutes. Some pressure is normal; sharp pain is not — tell us immediately if that happens.

Alternatives

If You Decline

For diagnostic indications: we cannot assess for infection without a fluid sample — watchful waiting is not appropriate when SBP is in the differential. For therapeutic indications: continued diuresis may partially manage the ascites but will not produce the same relief. Explain risks of untreated infected ascites.

Risks to Disclose

Standard risks for the consent discussion:

Pain or discomfort — local anesthetic minimizes this; pressure is expected; sharp pain should be reported immediately
Bleeding or hematoma at the needle site — most are self-limited; significant bleeding is rare (<1%)
Infection — iatrogenic peritonitis from the procedure itself is very rare (<0.1%) with sterile technique
Bowel perforation — very rare (<0.1%) with ultrasound guidance; usually self-healing; CT scan and surgical consultation required if suspected
Persistent ascites leak from the needle site — typically resolves with a pressure dressing; rarely requires a mattress suture
Post-paracentesis circulatory dysfunction (PPCD) — applies to large-volume paracentesis (>5 L); compensatory neurohormonal activation can cause AKI and hyponatremia; mitigated with albumin infusion (6–8 g per litre of total volume removed)
Failure to obtain adequate fluid — may require repositioning, alternative site, or repeat attempt
Death — extremely rare; mention only when procedural risk is genuinely elevated

Documentation tip: Note the indication, risks and benefits discussed, patient's questions and responses, and confirmation that the patient verbalized understanding and agreed to proceed. For patients who cannot consent (encephalopathy), document the surrogate decision-maker and the medical necessity.

Many institutions use pre-packaged commercial paracentesis kits (e.g., CareFusion, Merit Medical). Verify your kit contents against the list below and supplement as needed. Have specimen collection containers labeled and ready before the procedure begins.

Imaging

Ultrasound

Point-of-care US (POCUS) is the standard of care — do not perform paracentesis without first confirming the fluid pocket location and depth
Curvilinear probe (3–5 MHz) for abdominal imaging
Identify the largest accessible pocket, assess depth to peritoneum, and use Doppler to avoid abdominal wall vessels (inferior epigastric arteries)
Mark the site with a skin marker while the patient is in the position they will hold during the procedure

Sterile Field

Prep and Drape

Sterile gloves (two pairs recommended — outer pair can be changed after prep)
Chlorhexidine-alcohol or povidone-iodine prep solution
Sterile fenestrated drape
Sterile gauze (4×4)

Anesthesia

Local Anesthetic

1% or 2% lidocaine, 5–10 mL
25-gauge needle for skin wheal
22-gauge needle for deep infiltration down to peritoneum
5–10 mL syringe for injection

Access

Needles / Catheter

Diagnostic tap: 18-gauge needle on a 20–60 mL syringe; alternatively a small over-the-needle catheter (18–20 gauge)
Large-volume: 14–16 gauge over-the-needle catheter (e.g., Caldwell needle or kit catheter) connected to extension tubing and drainage bag or vacuum bottles

Specimen Collection

Tubes and Bottles

Blood culture bottles (aerobic + anaerobic) — inoculate 10 mL each at the bedside before any other tubes; this is the most important step for culture yield
EDTA (purple-top) tube — cell count with differential
Red-top or serum separator tube — albumin, total protein, lactate dehydrogenase (LDH), glucose
Additional tubes as indicated (see Fluid Studies section)

Finishing

Dressing

Gauze and pressure dressing for site
Adhesive bandage or transparent dressing
Suture (2-0 nylon) and holder — available but rarely needed for persistent leak

Site Selection

The left lower quadrant (LLQ) is the preferred default site — the spleen lies more superiorly and medially, bowel is generally less dense in the LLQ, and the left-sided approach is ergonomic for right-handed operators standing at the patient's left. The right lower quadrant (RLQ) is an acceptable alternative but carries greater risk of encountering cecum. The infraumbilical midline (linea alba) can be used when bilateral flanks are suboptimal, but avoid in patients with portal hypertension-related dilated midline veins (caput medusae).

Standard LLQ site: 3–4 cm medial and 3–4 cm superior to the left anterior superior iliac spine (ASIS). This site sits lateral to the rectus sheath, which is what keeps the needle away from the inferior epigastric vessels — those ascend medially, within the rectus sheath, so the danger zone is the rectus muscle itself, not the flank. The site also has a thinner abdominal wall and a deeper fluid pocket. Always confirm with US Doppler before committing to a site.

At minimum, every paracentesis should yield a cell count, albumin, total protein, LDH, glucose, and bedside-inoculated blood culture bottles. Supplemental studies are ordered based on clinical suspicion and the SAAG result.

Study	Tube / Method	Key Threshold / Interpretation
Routine — Send on Every Tap
Routine Cell count with differential	EDTA (purple-top)	Polymorphonuclear neutrophil (PMN) count ≥ 250/mm³ = SBP; treat immediately without waiting for culture. Also assess for lymphocyte predominance (tuberculous peritonitis, malignancy).
Routine Albumin	Red-top or SST	Required to calculate SAAG (serum albumin − ascites albumin). Both serum and ascites samples must be drawn at the same time. SAAG ≥ 1.1 g/dL = portal hypertension (97% accuracy).
Routine Total protein	Red-top or SST	Ascites protein < 1.5 g/dL = low opsonic activity → SBP risk; key criterion for primary SBP prophylaxis eligibility. Protein > 2.5 g/dL in high-SAAG ascites → cardiac/vascular etiology (vs. cirrhotic).
Routine LDH + Glucose	Red-top or SST	Required for Runyon's criteria for secondary peritonitis. All three: total protein >1 g/dL, glucose <50 mg/dL, and LDH > the upper limit of normal for serum. Meeting two or more of the three should raise suspicion for secondary (surgical) peritonitis rather than SBP.
Routine Culture (bedside inoculation)	10 mL aerobic + 10 mL anaerobic blood culture bottles, inoculated at bedside	Bedside inoculation improves yield from ~50% to ~80%. Positive monomicrobial culture + PMN ≥ 250 = classic SBP. Polymicrobial growth = strongly suspect secondary peritonitis or needle bowel contamination.
Selective — Order When Clinically Indicated
Selective Amylase	Red-top or SST	Markedly elevated (>1,000 U/L, often >3× serum) = pancreatic ascites from ductal disruption. Useful in pancreatitis with new ascites or failed response to SBP treatment.
Selective Triglycerides	Red-top or SST	> 200 mg/dL = chylous ascites (lymphatic disruption). Fluid will appear milky or opalescent. Etiologies: lymphoma, retroperitoneal surgery, trauma.
Selective Bilirubin	Red-top or SST	Ascites bilirubin > serum bilirubin (typically ≥ 2×) = biliary ascites from bile duct leak. Fluid appears bile-tinged or golden. Requires ERCP (endoscopic retrograde cholangiopancreatography) or surgical repair.
Selective Cytology	Large-volume collection (50–100 mL) in a plain tube or cytology container	Positive in ~65–80% of peritoneal carcinomatosis. Sensitivity improves with volume. Send when malignant ascites is suspected and SAAG < 1.1 g/dL. False negatives are common — peritoneal biopsy may be required.
Selective Adenosine deaminase (ADA)	Red-top or SST	> 30–40 U/L = high sensitivity for tuberculous peritonitis in endemic populations. Low specificity in cirrhotic ascites (many false positives). Best used in low-SAAG, lymphocyte-predominant ascites in a patient from a TB-endemic region.
Selective Gram stain	Submitted with culture bottles	Rarely positive in SBP (sensitivity < 10%) — bacteria are present in very low concentrations. A positive Gram stain with multiple organisms strongly suggests secondary peritonitis or procedural bowel contamination.

⟶

Related topic

SAAG and Ascites Interpretation

How to calculate SAAG, interpret the result, diagnose SBP and its variants, and apply antibiotic prophylaxis guidelines.

Paracentesis performed with US guidance and sterile technique has an excellent safety profile. The most common complication is minor abdominal wall bleeding — almost always self-limited. Life-threatening complications are rare and most are preventable.

Complication	Frequency	Management
Abdominal wall hematoma	~1%	Apply firm, sustained pressure (5–10 min). If expanding or hemodynamically significant: US or CT to characterize; interventional radiology (IR) embolization for ongoing arterial bleeding. Pre-procedure Doppler reduces incidence.
Persistent ascites leak	<1%	Z-track technique is the primary prevention. If leak persists > 30 min: close the skin puncture with tissue adhesive (dermabond) or a figure-of-eight suture (2-0 nylon).
Post-paracentesis circulatory dysfunction (PPCD)	~70% without colloid replacement (Ginès et al., J Hepatol 1990)	Prevented by albumin infusion: 6–8 g of 25% albumin per litre of total volume removed when >5 L is drained; albumin is the evidence-based choice (PMID 33942342). If PPCD occurs (AKI, hyponatremia, hemodynamic instability): volume expansion, hold diuretics and nephrotoxins, reassess renal function daily. Consider hepatorenal syndrome (HRS) workup if creatinine rises.
Bowel perforation	<0.1% with US guidance	Suspect if: feculent fluid, frank peritonitis signs post-procedure, or Gram stain with multiple organisms. Order CT abdomen/pelvis immediately. Surgical consultation is required. Most small-needle perforations self-seal in the bowel wall; larger punctures or thin-walled bowel may require surgical repair.
Iatrogenic infection	<0.1%	Extremely rare with proper sterile technique. If post-procedural peritonitis is suspected: repeat paracentesis for cell count and culture. Distinguish from SBP discovered on the original tap versus new infection introduced by the procedure.
Significant hemorrhage	<0.1%	True large-vessel injury is exceedingly rare. Risk is not meaningfully elevated by abnormal coagulation studies in cirrhosis — do not transfuse prophylactically. If hemorrhage occurs: type and cross, IR evaluation, surgical backup. Resuscitate while characterizing source.

A note on coagulopathy and bleeding risk: Multiple large prospective series have confirmed that INR and platelet count do not predict paracentesis-related bleeding in patients with cirrhosis. The rebalanced hemostasis of liver disease means standard coagulation tests systematically overestimate bleeding risk. Withholding a necessary diagnostic paracentesis due to an elevated INR is a clinical error — the risk of untreated SBP far exceeds the risk of procedural bleeding.

Case Resolution

Returning to the case from the top of the page: 61-year-old woman with alcohol-associated cirrhosis (Child-Pugh B, MELD 3.0 = 18), admitted with 4 days of worsening abdominal distension and low-grade fever. INR 1.8, platelets 68,000/mm³.

Indication: Hospitalized patient with cirrhosis and ascites + fever → diagnostic paracentesis is mandatory at admission, regardless of whether she has abdominal pain. Clinical silence does not exclude SBP.
Consent: Acknowledge prior paracentesis, briefly review risks (bleeding, infection, leak), explain that the culture bottles are placed at the bedside this time (highlight improvement in practice). Confirm understanding.
Lab review: INR 1.8, platelet count 68,000/mm³. Do not transfuse FFP or platelets — not indicated. Proceed.
Setup: US confirms a 6 cm LLQ fluid pocket at 3.5 cm depth. Doppler shows no vessels at the intended entry point. Mark and prepare. Assemble aerobic + anaerobic blood culture bottles at bedside before starting.
Result: 60 mL of straw-colored fluid aspirated. PMN 310/mm³ (no RBC, no traumatic tap correction needed). Culture bottles inoculated immediately. Chemistry tubes sent.
Interpretation: PMN ≥ 250 → SBP. Start ceftriaxone 2 g IV now. Do not wait for culture. Give albumin 1.5 g/kg on day 1 and 1 g/kg on day 3 if the patient meets any high-risk criterion — creatinine >1 mg/dL, BUN >30 mg/dL, or total bilirubin >4 mg/dL (Sort 1999) — as albumin reduces the risk of hepatorenal syndrome and death in this group. Hold diuretics.
Teaching point: The patient had a diagnostic paracentesis three months ago. The chart shows her ascites protein was 1.3 g/dL at that time and her creatinine was 1.4 mg/dL — she met criteria for primary SBP prophylaxis (protein <1.5 g/dL + creatinine ≥ 1.2 mg/dL) and was not started on it. This hospitalization was potentially preventable.