Ascites Tapping

Also known as paracentesis

A more general discussion of Ascites is found elsewhere, including the medical management. This article will discuss only issues related to the tapping of ascites.

Tapping of ascites is a simple technique that can be used for diagnostic or therapeutic purposes. For the patient with cancer this tends to be palliative but in cirrhosis it may be an important contribution to tiding the patient over to a liver transplant.

Diagnostic purposes include where peritonitis is suspected.¹ This can occur spontaneously with cirrhosis and should be undertaken early if there is any suspicion. Diagnostic paracentesis is less often needed because of modern imaging techniques but it may still be valuable in ovarian carcinoma.² Analysis of the fluid will reveal if it is a transudate or exudate (low or high in protein respectively) and centrifugation and cytology may show malignant cells. The difference between transudate and exudate is usually taken as a protein content of the fluid of less than or more than 25g/L respectively. However, the assumption that transudate is cirrhosis and exudate is malignancy is false. It is often presumed that cardiac ascites is a transudate when this is rarely the case. Ascitic fluid in myxoedema is an exudate.³ Ascitic protein is >25 g/L in up to 30% of patients with otherwise uncomplicated cirrhosis, and patients with cirrhosis and tuberculous ascites may have a low ascitic protein. The serum ascites-albumin gradient (SA-AG) is far superior to simple protein content in categorizing ascites with 97% accuracy.³ The SA-AG involves ascitic-to-serum total protein ratio, lactic dehydrogenase, and ascitic-to-serum lactic dehydrogenase ratio in differentiating between ascites from liver disease and malignant ascites. If the ratio is greater than 1.1, this reflects portal hypertension.

A large amount of fluid is very uncomfortable and removal may also improve pulmonary function and oxygenation.⁴ Tense ascites may also have an adverse effect on the circulation, especially venous return from the lower limbs. Pressure on the stomach will impair appetite and cause nausea.

Where the ascites is due to congestive heart failure or cirrhosis, medical treatment should be energetic before resorting to tapping but if all else fails it offers some relief. It is also valuable in preparation of patients for liver transplantation. Bed rest does not reduce ascites.

Before tapping, there are certain investigations that should be undertaken. A more lenient approach is permissible if it is part of palliative care for advanced cancer.

• FBC, U&E, creatinine, LFTs including plasma proteins
• Clotting screen
• Ultrasound examination of the abdomen should include liver, pancreas, spleen and lymph nodes.
• The last is a very sensitive way of assessing ascites and may also show the causative pathology such as carcinoma of ovary or metastatic liver disease.

There are no data to support the use of fresh frozen plasma before paracentesis although if thrombocytopenia is severe (<40,000/mm^-3) most clinicians would give pooled platelets to reduce the risk of bleeding.

In patients receiving palliative care, it is important not to be excessive in investigations. A protocol for such care from Plymouth⁵ allows ultrasound to be reserved for cases of diagnostic uncertainty. It also permits up to 5 litres to be removed at a single time, leaving drains for no more than 6 hours and giving IV fluids only when indicated.

After a diagnostic tap the following investigations may be requested:

• All patients should be screened for spontaneous bacterial peritonitis (SBP), which occurs in approximately 15% of patients with cirrhosis and ascites admitted to hospital.⁶ An ascitic neutrophil count of >250 cells/mm^-3 is diagnostic of SBP in the absence of a known perforated viscus or inflammation of intra-abdominal organs.
• The concentration of red blood cells in cirrhotic ascites is usually <1000 cells/mm^-3 and bloody ascitic fluid (>50,000 cells/mm^-3) occurs in about 2% of cirrhotics. In approximately 30% of cirrhotics with bloody ascites, there is an underlying hepatocellular carcinoma.
• In 50% of patients with bloody ascites, no cause can be found.
• Gram stain of ascitic fluid is rarely helpful. The sensitivity of a smear for mycobacteria is very poor while fluid culture for mycobacteria has a sensitivity of 50%.
• Several studies have shown that inoculation of ascitic fluid into blood culture bottles will identify an organism in approximately 70% of cases whereas sending ascitic fluid in a sterile container to the laboratory will only identify an organism in about 40% of cases of SBP.
• In pancreatic ascites the amylase in the fluid will be markedly raised.
• Only 7% of ascitic fluid cytologies are positive yet cytological examination is 60 to 90% accurate in the diagnosis of malignant ascites, especially when several hundred millilitres of fluid is tested and concentration techniques are used. It is not so valuable for primary hepatocellular carcinoma.

• As the tapping of ascites is usually to relieve symptoms, it is not normally required unless the ascites is tense.
• Diagnostic tapping is a different matter and does not require the removal of a large volume.
• Therapeutic or palliative paracentesis should not be undertaken more often than is required to keep the patient reasonably comfortable.
• A major risk is the introduction of infection into the peritoneal cavity. Appropriate aseptic technique must be employed and the drainage catheter should not be in place for more than 6 to 8 hours.
• If repeated tapping is part of palliative care, it can be performed in a hospice or the patient's home provided that sterile precautions are taken. It is not imperative to admit the patient to hospital.
• Serious complications such as haemoperitoneum and bowel perforation occur in less than 1 in 1,000 procedures.⁷

• Check the equipment. There should be needles, a 10 ml syringe, local anaesthetic, antiseptic skin preparation may be employed but its value is dubious. There should be a very wide bore IV cannula, and IV giving set and a urine bag of the type attached to a catheter. Adhesive tape is also required. Surgical gloves are worn.
• Examine the abdomen to decide where (and where not) to make the tap. It is usually in the lower flank. Although the midline is a fairly avascular area, there is danger that the bladder may be more full than expected and it could be damaged. The inferior epigastric vessels run up the side of the rectus abdominis to anastomose with the superior epigastric vessels that come down. The typical patient with ascites does not have well demarcated abdominal muscles and so keep well laterally without going as far as the pelvis. This usually means about 15cm from the mid-line. Avoid solid tumour masses. Avoid scars. There may also be adhesions beneath them.
• Having decided on the site and possibly applied antiseptic cleanser, infiltrate the skin and then the deeper layers with local anaesthetic. Finally, move the needle down into the ascitic fluid and draw back. If typical straw-coloured fluid is not aspirated than another site must be chosen. If the tap is purely diagnostic, attach a large, new syringe and 10 to 20 ml of fluid can be aspirated into a syringe but if a therapeutic tap is required to withdraw a much larger volume, continue as below.
• The needle of an IV cannula is usually inserted into the left (preferably) or right lower abdominal quadrant using the Z track technique in which the skin is penetrated perpendicularly, the needle is advanced obliquely in subcutaneous tissue and then the peritoneal cavity is punctured with the needle pointing perpendicular to the abdominal wall. This will ensure that the puncture site on the skin and the peritoneum do not overlie each other.
• When the needle enters the peritoneal cavity, there is a reduction in resistance. Partially withdraw the inner needle and push the cannula in a little further. Withdraw the inner needle and there should be a free flow of fluid.
• Attach the cannula to the giving set and the giving set to the urinary bag.
• Ascertain that there is free flow of fluid and tape the cannula to the skin. Try to position it so that the cannula is at right angles to the skin to permit free drainage.

• The patient needs to be checked regularly to ascertain that drainage is free and the bag is emptied and the volume recorded. There may be 2 or 3 litres drained in the first hour with complete drainage in 4 to 6 hours.
• All ascitic fluid should be drained to dryness in a single session as rapidly as possible over 1 to 4 hours, assisted by gentle mobilization of the cannula or turning the patient on to the side if necessary.
• Expert opinion is that swift drainage is safest as it reduces the risk of introducing infection but if the patient develops symptoms of hypotension then the blood pressure should be checked and, if necessary, the drainage stopped.
• When drainage ceases and clinically the ascites is much less or after 6 to 8 hours at the most, the catheter is gently withdrawn and the hole closed. This may require a single silk suture. Many authorities would recommend a maximum drainage time of 4 rather than 8 hours, at least with cirrhosis.

Removal of large volumes

• Large volumes, even in excess 10 litres over 2 to 4 hours may be removed.
• This causes a marked reduction in intra-abdominal and inferior vena cava pressure, leading to an increase in cardiac output. These haemodynamic changes are maximal at 3 hours. Pulmonary capillary wedge pressure decreases at 6 hours and continues to fall further in the absence of colloid replacement. On average, blood pressure decreases by about 8 mm Hg.⁸
• It may be safe to omit plasma expansion if less than 5 litres is withdrawn but for larger volumes, it is recommended. There is debate about the relative value of albumin or artificial plasma expanders but trials to date have been inadequately powered to answer the question.
• The use of albumin infusion is not normally recommended during palliative paracentesis of a patient with advanced cancer but cirrhosis is a different matter. After withdrawal of 5 litres or more these patients are at risk of post-paracentesis circulatory dysfunction (PPCD). This is characterised by hyponatraemia, azotaemia, and an increase in plasma renin activity. There is an increased mortality that may be prevented by administration of intravenous albumin at 6 to 8 grams per litre of ascites removed.⁹ There is also evidence that the use of albumin reduces the risk of hepato-renal syndrome and spontaneous bacterial peritonitis¹⁰ although the latter is disputed.

The introduction of spironolactone after paracentesis may reduce the need to repeat the procedure from over 90% to less than 20% in cirrhosis.¹¹ Such benefit it not to be expected in malignant ascites.

• The following day, or within 48 hours, the blood tests should be repeated although, as always with terminal care, such rigidity is not always necessary. The ascites will re-form and can be tapped again.
• A major concern is the introduction of infection. This may not be associated with signs of peritoneal irritation and so if the temperature rises over the next few days, then antibiotics should be given to cover such infection.
• If a large amount of protein has been removed, a high protein diet is in order. Drainage of ascites may have reduced pressure on the stomach but the medical condition is usually such that appetite remains poor.

• In the 1950s the introduction of potent diuretics made the tapping of ascites seem crude and dangerous but a review in 1990 suggested that it is both safe and effective.¹²
• Controversies have existed about the volume that may safely be tapped, the rate of removal and the simultaneous use of intravenous albumin. Confounding factors may include the underlying disease. In the terminal patient the prime concern is comfort.
• Most studies seem to have been on patients with cirrhosis where volumes as large as 4 to 6 litres a day can be safely withdrawn as long as intravenous albumin is given. In one study¹³ it was common to remove 10 litres of fluid in about an hour.
• In malignant disease tapping of ascites can bring some relief to 90% of patients.¹⁴ Where frequent drainage is required a permanent drain can be left in place but this increases the risk of infection.¹⁵
• An alternative is a peritoneovenous shunt but blockage occurs in about a quarter and in cirrhosis it increases mortality.⁹
• In conditions such as carcinoma of ovary where the exudate is high in protein, repeated tapping of large volumes may cause protein depletion but the relief of symptoms in a terminal patient may justify this.