Paracentesis
Lesson Outline
Intro
Important role in many settings, particularly medical patients (including but not limited to critical care), but could be needed anywhere.
Diagnostic and therapeutic roles, more often former, and can be needed for time-sensitive diagnoses, so very useful to have the skillset to do your own
Generally safe and straightforward, and a good exemplar of how many bedside taps are done, so can be a good "gateway" procedure
Indications
Diagnostic:
Goal of diagnosis; can be as little as a few cc's (25-30 is a good idea)
Evaluate cause of new/unexplained ascites
All new ascites should probably be considered unexplained, even if a cause is suspected; "every fluid collection should be tapped at least once."
Won't go into all causes of ascites here, but:
As in other cases of fluid collections, a majority of causes will be "benign" (i.e. secondary to known comorbidities and not indicative of another sinister problem), such as known cirrhosis
But some will be essential to diagnose, most importantly malignancy
Evaluate for infection, mainly SBP
Any pre-existing ascites can become secondarily infected, usually in cirrhotics
Clinical exam not sensitive to this
Probably any cirrhotic with tappable ascites admitted with any meaningful systemic illness, especially signs of infection (probably include encephalopathy and other non-specific systemic findings) should potentially have a diagnostic para
Mortality increases by 3.3 percent/hour of delay in performing a paracentesis
Leaks - lymph, pancreatic duct, etc (post-op patients often have drains though)
Blood - diagnostic peritoneal lavage essentially a paracentesis in trauma patients to evaluate for hemoperitoneum, although very rarely done nowadays
Therapeutic
Goal of symptom management; usually >500 ml
"Large volume paracentesis" usually defined as >5L
Relieve symptoms from tense ascites
Comfort
In ICU, most often respiratory compromise
Rarely the primary reason for respiratory failure but might contribute to dyspnea
May relieve abdominal pressure contributing to abdominal compartment syndrome
Again, rarely a monofactorial cause of this, but if ACS and ascites are both present, drainage will reduce pressure
Clears some total body fluid
Third spaced, not readily interchanging with circulating volume, but eventually it all equilibrates, so ultimately acts like a diuretic
Risks and contraindications
Generally a safe procedure
Real but extremely low risk (with ultrasound guidance and competent proceduralist) of injuring solid organs, bowel, other structures
Negligible risk of infection
Most important risk: bleeding
Significant bleeding is extremely rare - concern about bleeding probably more likely to be a "false positive" than helpful (ie. prevent or delay a needed procedure than actually avoid a complication)
Common cause for concern is lab abnormalities (elevated INR, thrombocytopenia, etc), which are common in cirrhotics
However, most significant bleeding is not due to coagulopathy - the procedural trauma is negligible - but due to mechanical injury
Most often, probably lacerating an abdominal wall vessel (veins may be prominent in patients with portal hypertension)
Avoided with ultrasound and good technique
In a critically ill patient needing a para, probably no lab cutoffs should preclude this procedure
Can try to stick with diagnostic paras, which often use smaller needles to further limit risk
Frank clinical signs of coagulopathy, such as bleeding from IV sites, might give you more pause (and/or data from functional testing like TEG/ROTEM
Bleeding rates as low as 0-0.19%
Performance standards for therapeutic abdominal paracentesis
Severe haemorrhage following abdominal paracentesis for ascites in patients with liver disease
AASLD: routine transfusion to reverse coagulopathies before paracentesis not recommended, outside frank clinical DIC or hyperfibrinolysis (role for TEG/ROTEM?)
SIR: generally safe even in thrombocytopenia and elevated INR. Do not recommend routinely checking INR or platelet count.
AGA: in routine cirrhosis, don't check coags/platelets and don't routinely transfuse to correct them
INR in particularly very poorly predictive of bleeding risk in cirrhotics - high INR in cirrhosis reflects dysfunction of one portion of clotting cascade, but other portions may be normal or actually accelerated (vWF/VIII often elevated). Overall clotting function as likely to be normal/hypercoagulable than reduced.
General recs from AGA, AASLD, SIR, etc: do the para to rule out SBP, don't be deterred by lab values. The procedure is very safe, the diagnosis is important (and time sensitive), and the numbers are probably meaningless.
Final important risk: hemodynamic instability after large-volume drainage
Caution performing a big para in shock, can precipitate instability
Sudden change in abdominal pressure = change in preload/afterload
Reduction in abdominal pressure -> Increased filtration (reaccumulation of ascites) -> hypovolemia
When in doubt, stick to diagnostic tap
Technique overview
Sterility
Most often "radial arterial line sterile": gloves, hat, mask, local drapes
Very low risk of infection; more likely to contaminate a sample, but that's not very likely either
Target site
Ultrasound guidance the STANDARD OF CARE at this point in the US
Can absolutely do this blind, with large pockets, but ultrasound is very easy, ubiquitous, identifies vessels and adherent structures, and makes tapping smaller pockets far more plausible
People routinely doing landmark guidance are probably periodically hitting other anatomy. "Dry taps" are not uncommon (almost never with ultrasound)
Most common: mark-and-tap approach
Realtime guidance is possible, and probably adds to safety; may be essential for small pockets
Easy to do for a diagnostic tap with small needle (same technique as vascular access)
Can be difficult with a large therapeutic needle which is unwieldy with one hand
Micropuncture or other seldinger approach can be a nice method
Curvilinear probe ideal; echo/phased array probe okay
Survey abdomen to identify best pocket
In general, landmark anatomy suggests avoiding the midline and umbilical area as highest risk of abdominal vasculature (the linea alba itself is avascular, but a small target); ultrasound helps show vessels but still a good idea
Most common target in bedbound patients: lower flanks
Some theoretical argument that left flank tends to be better
Experiment with position, but usually a small HOB elevation helps
Very trace ascites usually collects first around liver, but if only seen here, probably not a safe target; you do not want to hit a liver
Ascites visible as echo-free fluid flowing into recesses, bowel loops usually floating deep, abdominal wall in near field
Best target: largest pocket in all directions, including depth
Identify depth of pocket
Loosen probe pressure to relax tissue
Note how far needle will need to enter (approximately)
Gives a sense for needle selection
Ideally, path to pocket should be perpendicular to skin (peritoneal wall = straight); easiest to replicate this blindly
Mark site
Best method: apply a syringe (any kind), firm pressure (no twisting), apply suction and wait
Creates a "bullseye" that will last a while
Can mark skin as well with a permanent marker, but tends to come off with prep
Reapply probe after marking to confirm
Switch to linear probe
Apply to marked site
Apply color doppler over abdominal wall to help identify subcutaneous vessels
Prep skin with chlorhexadine
Local draping
Anesthetize
Skin first, then deeper tract
Inject every few mm
Aspirating pressure
Some suggest intermittent, not continuous aspirating pressure
Once ascites is aspirated, you are in
Withdraw and inject more? Most effective numbing, but theoretical risk for seeding malignant cells
Z-track limits subsequent leakage
Probably low yield for lidocaine, more useful for larger needles
Can confuse your landmark
Needle choice
Diagnostic: almost anything
Large therapeutic needle can be used, but may be overkill
IV angiocatheters can be used
Longer versions helpful in all but the slimmest patients
Ideally want a version with open proximal end, allowing free flow during insertion and/or aspiration with a syringe (ie not a protected flash chamber)
Spinal needles can be used
Usually very long, without a catheter
No catheter = needle must be left in place during aspiration. Requires some care.
However, allows the smallest possible hole.
Micropuncture sheaths can be used
Microneedle itself can be used (usually 21 g), or wire inserted and sheath railroaded (usually 4-5 Fr), allowing easy aspiration of small to large volumes
Still not quite the same as therapeutic kits, since it only has one hole
Therapeutic:
Micropuncture sheath
Various apparatuses: "Safe-T-Centesis," Caldwell needle, etc
Most often a fenestrated catheter-over-needle
May be a stiff needle without catheter that remains in place
General principles
Larger needles usually require a skin nick
As small as possible, and no deeper than skin; this is probably the highest risk step for bleeding
Needle insertion
Aspirate as you go (continuous vs intermittent)
If skin nick, insert needle into nicked hole
Therapeutic needle usually requires a two-handed technique
One hand with aspirating syringe pressure
Other hand very close to skin to stabilize needle
If aspirating directly from needle, simply hold in place once pocket entered and pull fluid
Usually requires switching syringe to a larger one
If placing catheter, once fluid aspirated, insert a couple mm further and advance off catheter
Aspirate samples and prepare appropriately
Varies by lab. Can just put a bunch of fluid in sterile containers, but many labs now want some samples in blood tubes. Order your labs ahead of time and see what it asks for.
Cultures may go into sterile bottles, or directly into blood culture bottles - increases culture yield, but some labs don't like it
If using culture bottles, remember to get them to the bedside and swab the tops
Wise to draw extra fluid for possible add-on tests
For therapeutic taps, after fluid samples:
Generally attach tubing, then attach to either glass vacuum bottles (evacuated, provide -300 mmHg suction pressure), or wall suction
Both can provide a lot of suction - bottles have several hundred mmHg of negative pressure
As you fill either, an assistant can switch your tubing to the next
Best time to assess fluid quality is within tubing (bottles/syringes tend to look dark)
Discard in biohazard waste, or can pour out fluid into a toilet or similar (check local regs)
How much to drain?
If tolerating, can drain dry
If tenous hemodynamics, maybe be judicious
Pressure can actually be transduced from the catheter, an invasive method of abdominal pressure measurement - may be useful if draining for abdominal compartment syndrome
Flow usually starts to become tenuous eventually; adjust catheter position, maybe adjust patient position, until no further flow
Remove catheter and apply some kind of dressing
If ascites left, may leak from hole; Z-track helps with this
Usually not a big deal; if very annoying, can apply some skin glue, or cover with an ostomy bag
Albumin replacement
Limits fluid shifts after large-volume drainage
Somewhat more evidence based than most albumin applications
Most experts suggest necessary only with >5L drainage
Give about 8 g per liter, within about 60 minutes of procedure
Most use concentrated (25% in US) albumin
Can consider in lower volume drainage for hemodynamically tenuous patients, but not very clear
Labs and interpretation
Three typical questions in descending order of frequency in ICU
Is there infection?
Is there portal hypertension?
Is there malignancy?
Always send:
Cell count and differential
Gram stain and culture
Albumin
Send cytology (and perhaps flow cytometry) if there is concern for malignancy, and probably always for a first tap
Other labs as needed
Protein
Glucose, LDH, amylase, bilirubin
TB testing as appropriate
Interpretation
Is there infection, and if so, is it spontaneous bacterial peritonitis (seeding of ascites via bacterial translocation) or secondary peritonitis ("regular" peritonitis from an abdominal focus, such as bowel perforation, abscess, biliary disease, etc)?
PMN (aka neutrophil) count >250 cells/mm^3, plus positive culture
Usually need to calculate by multiplying total nucleated cell count by % of neutrophils
Bloody taps may skew results - subtract one PMN from the result for every 250 red cells/mm3
Gram stain/culture
Positive ascites cultures are somewhat rare (especially after antibiotic administration), so PMNs are often the only signal of SBP
Supportive tests
Very high PMN count, glucose <50, protein count >1 g/dL, elevated LDH (above reference range of serum) support secondary bacterial peritonitis
Polymicrobial infection supports secondary
Bilirubin or amylase occasionally support biliary or pancreatic origin of ascites
If concerned for secondary peritonitis, pursue imaging (eg CT with contrast)
Calculate SAAG
Serum ascites-albumin gradient - similar to Light's criteria
Serum albumin always higher
Difference (not ratio) between albumin in serum and fluid
Ideally sampled at same time
>1.1 g/dL, portal hypertension present
<1.1 g/dL, probably no portal hypertension (hence consider other causes, such as exudative processes - infection, malignancy)
Protein count may help differentiate cirrhotic ascites from heart failure
<2.5 g/dL supports cirrhosis
>2.5 supports heart failure
If SBP present
Treat medically
If doing well, no problem
If not doing well, or any question of diagnosis (e.g. considering secondary peritonitis), consider repeat diagnostic paracentesis after 48 hours of treatment
PMN count should be DOWNTRENDING; otherwise, consider antibiotic resistance, or wrong diagnosis
Closing thoughts
Para frequency really depends on patient population (cirrhosis frequency)
Good care of cirrhotics usually means tapping a lot
Doing this early (for timely SBP diagnosis, especially before giving antibiotics) usually means ICU does it, not farming out to another specialty
Generally, a fairly satisfying, fairy straightforward, fairly safe procedure