Percutaneous Abscess Drainage Periprocedural Care

Updated: Jan 24, 2022
  • Author: Evan J Samett, MD; Chief Editor: Kyung J Cho, MD, FACR, FSIR  more...
  • Print
Periprocedural Care

Preprocedural Evaluation

Diagnostic imaging

Occasionally, an abscess is noted as an incidental finding on computed tomography (CT) or ultrasonography (US) in a patient without clear stigmata of sepsis or systemic infection (eg, an older patient with failure to thrive). More often, a diagnostic study is performed for a suspected abscess in a patient with clinical sepsis or infection syndrome (fever and leukocytosis). The patient usually has risk factors for abscess, such as recent surgery, infection, perforated viscus, or immunosuppression.

Serious local infection can be categorized on the basis of location, viscosity, complexity, contents, etiology, or surrounding structures. It can occur in solid organs (eg, liver or muscle), [16, 17]  in potential spaces (eg, pleural or peritoneal cavity), or in preexisting or physiologic fluid collections or organs (eg, gallbladder or urinary tracts).

Computed tomography and ultrasonography

CT and US are excellent at identifying potential abscess areas. In the proper clinical setting, strong signs are a well-circumscribed fluid collection, thickened membranes or septations, gas bubbles, contrast enhancement, and debris. Weaker signs are nonloculated or thin-walled nonenhancing collections that have alternative pathophysiologic explanations, such as peritoneal or pleural fluid.

The choice between CT and US depends on physician preference and/or issues of logistics or convenience. The availability of multiplanar reformatted (MPR) imaging and CT fluoroscopy, as well as advances in US image quality, make either option excellent in the right hands. CT may be preferred for abdominopelvic collections that are small or located in deep recesses. [18]

Magnetic resonance imaging

Magnetic resonance imaging (MRI) is most useful in identifying abscesses in the musculoskeletal system. Subtle infection from streptococci may cause erysipelas, which usually manifests as a spreading cellulitis. However, deeper infection in the fascia between the subcutaneous tissue and muscle or between muscle layers or at the periosteum also can be associated with streptococci and imaged well with short-axis contrast-enhanced MRI.

MRI is the best way to find intraosseous and subperiosteal abscesses associated with osteomyelitis. CT or US may be needed to determine if a process contains sufficient fluid elements to benefit from percutaneous abscess drainage (PAD). [17]

Categorization of local infection

The spectrum of infected tissue ranges from infected viable organs with no significant degradation of physical integrity to total liquefactive necrosis. Main determinants of ease of drainage are viscosity and complexity (ie, septation, loculation, and debris). Viscosity ranges from near-water to near-solid. Abscess complexity ranges from unilocular to innumerable septations and loculations. Typical examples of complex viscous abscesses include infected pancreatic abscesses or infected retroperitoneal hematomas. These processes are challenging to treat percutaneously and may require multiple interventions over a prolonged recovery period.

Abscesses associated with a perforated bowel (high- and low-flow fistulas) are also challenging to treat. The presence of a low- or high-flow enteric fistula dramatically increases the expected duration of therapy. Successful treatment requires the understanding and cooperation of the patient and referring physician. Knowledge of abscess etiology occasionally may guide therapy. A patient with a history of ruptured viscus is at high risk for enteric organisms. Infection with Echinococcus species must be treated cautiously because of the risk of anaphylaxis.

Viscosity and complexity are difficult to assess with imaging alone. Viscosity cannot be readily predicted unless debris or suspended bubbles are seen. Fine septations are easily missed on CT because of partial volume averaging on routine scans. As an example, a patient had a large intrahepatic pseudocyst infected with fungus. CT demonstrated a simple cyst appearance, a sharply demarcated isodense unilocular collection. US demonstrated innumerable fine septations. The complex nonviscous abscess was successfully treated with PAD.

As another example, a patient with sepsis syndrome had a minimally irregular unilocular abscess appearance on CT. US revealed a well-demarcated hyperechoic mass with central lucency, possibly a hemangioma or necrotic malignancy. The patient was later found to have a perirectal abscess as the source for infection.

Differentiating bowel from pathologic intra-abdominal fluid collections is important. This may be difficult in patients with scant body fat, postsurgical abdomen, or unopacified bowel loops. Multislice spiral CT with gastrointestinal (GI) preparation and intravenous (IV) contrast media is a helpful tool. Thin-slice, coronal, or sagittal reformatted images may help the practitioner understand the anatomy. Delayed or prone imaging with additional oral contrast preparation may be considered in challenging cases. This is particularly true in the frail patient with poor fat-plane differentiation.

Patient selection and preprocedural workup

PAD is generally a safe and effective treatment. The preprocedural evaluation assists procedure planning and risk assessment. Clinical factors increase or decrease complication risks. (See the image below.)

Radiology procedure preassessment form. Radiology procedure preassessment form.

Safe PAD depends on an intact clotting system. In some practices, coagulation abnormalities are sufficiently common that the international normalized ratio (INR), the prothrombin time (PT), and platelet counts are routinely obtained before the procedure.

Bleeding risk depends on the etiology, the degree of coagulopathy, and the reversibility of coagulopathy. The risk is nominal as long as the PT/INR and platelet counts are stable after correction. The risk is increased if coagulopathy recurs quickly. Recurrent or irreversible coagulopathy is caused by synthetic failure (eg, liver or bone marrow) or consumption (eg, hypersplenism or disseminated intravascular coagulation [DIC]).

A common practice is to correct recurrent coagulopathy for the duration of a procedure. A risk of delayed hemorrhage may be present if the coagulopathic state recurs in the early postprocedural period. Anecdotal reports describe of fatal bleeding more than 24 hours after the procedure in this situation. The extensive amount of transfusion required to achieve this level of protection is expensive and burdens blood bank resources. If a patient cannot be treated noninvasively, coagulopathy should be managed as follows:

  • INR < 1.8 - Perform examination with radiology attending approval, or correct prior to procedure
  • INR ≥1.8 - Correct coagulopathy prior to examination, and verify repeat INR at less than 1.8
  • INR >2.5 - Patients on warfarin or severe liver disease may require 24 hours or more for correction of the INR, depending on the clinical condition and treatment given; adequate correction may be difficult in patients with synthetic failure; patients with severe coagulopathy requiring procedures may need coagulation support after the procedure

Concurrent anticoagulant and/or antithrombotic medication use can also increase risk of bleeding complications. Aspirin and clopidogrel can have clinical effects lasting up to 5 days or longer. The traditional anticoagulants heparin and warfarin have the benefit of available blood tests to monitor therapeutic effects—namely, partial thromboplastin time (PTT) and PT/INR. The newer medications do not have similar clinical assays available, and appropriate clinical correlation is required.


Preprocedural Planning

CT is an excellent modality for planning the PAD access path. Position, angle, and depth can be assessed along with the presence of intervening or adjacent vital structures (eg, bowel, lung, spleen, and liver), which increase the overall risk of PAD. [19] If abscess suspicion is high, and it is necessary to traverse intervening structures, then a Seldinger technique may be planned to confirm the presence of an abscess with a needle aspiration before dilating the tract for catheter placement.

US may be used for large or superficial collections or for bedside procedures. Advantages of US include real-time imaging of the needle and catheter path, reduced radiation, and the ability to perform the procedure in the interventional radiology suite in concert with conventional fluoroscopy.

Risk assessment helps determine whether the treatment plan is appropriate and reasonable. Decisions are based on the risk of complication, the expectation of clinical benefit, the overall prognosis, and the availability of treatment alternatives. The more likely it is that a fluid collection is infected, the more likely it is that drainage will be clinically beneficial. It is acceptable to be less aggressive with nonloculated fluid, such as free ascites or pleural effusion. For these collections, a diagnostic aspiration may be warranted before attempted catheter drainage.

A white blood cell (WBC) count with differential and vital signs helps determine the infection's seriousness. Treatment alternatives should be considered when fluid collections are in proximity to vital structures, have a difficult access route, or are found in a severely coagulopathic patient. Neurologic and psychological factors impact a patient's ability to tolerate the procedure without sedation and the risk of postprocedure catheter dislodgment.

Both medical and surgical treatment alternatives should be considered. For example, if an abscess is small and close to or in the spine, the drainage procedure could spread the infection to the cerebrospinal fluid (CSF) and spinal cord. These infections may benefit from an initial trial of IV antibiotics.

A risk-benefit categorization can be adapted from angioplasty guidelines for all invasive procedures. This uses an ordinal scale from 1 through 4 (the higher the number, the less ideal the procedure), as follows:

  • Category 1 - The proposed treatment or procedure is the procedure of choice; a high technical success rate is expected and generally results in complete relief of symptoms or normalization of abnormal clinical conditions or provides diagnostic clinical information
  • Category 2 - Lesions or processes are well suited for the proposed treatment or procedure; a high technical success rate is expected and generally results in complete relief or significant improvement or provides relevant clinical information; this includes lesions and processes that are treated with adjunctive surgery or other modalities
  • Category 3 - Lesions or processes are amenable to the proposed treatment or procedure; a moderate technical success rate is expected, and because of disease severity, type or location, or procedure risk, there is a moderate expectation of obtaining long-term benefit or relevant clinical information; the procedure may be indicated because of other patient risk factors or because of the lack of alternative treatment options
  • Category 4 - The procedure has limited utility because of a low technical success rate, high risk, or poor long-term benefit; the patient has extensive, possibly comorbid disease; the procedure may be indicated in high-risk situations when no treatment alternative is available

Marginal procedures (categories 3 and 4) require close interdisciplinary consultation. Serious clinical conditions warrant an aggressive approach with an attendant increase in risk. Potentially helpful procedures must be differentiated from those with a low expectation of durable clinical benefit. Difficult PAD procedures generally are well tolerated if performed with a minimum of needle passes. Solid-organ, even intrasplenic, abscess may be drained. Drainage of a high-probability abscess should be considered in a salvageable patient with coagulopathy.




Typical abscess fluid is readily aspirated through an 18-gauge needle, which has approximately one 20th the resistance to fluid flow that a 21- to 22-gauge needle does. Viscous or debris-laden fluid is more likely to cause a false-negative aspirate with a 21-gauge needle, in that only clear supernatant may return through the needle. An 18-gauge needle is easier to control and image and accepts a 0.038-in. guide wire. Neff and Accustick sets first require a 0.018-in. mandrel wire followed by a coaxial dilator. The 18-gauge needle costs less than the Neff access set.

There is no clinically significant difference in needle trauma between the 18- and 22-gauge needles when a catheter is placed through the same tract. (This has been demonstrated for arteriotomy.) The 21-gauge needle may be used to minimize trauma for a challenging localization, low-probability fluid collection, or personal preference.

Other supplies

For the trocar technique, all that is required is a trocar-type drainage catheter, a drainage bag, and a minor procedure tray containing preparations, local anesthesia, a scalpel, and draping material. For the Seldinger technique, additional supplies include a guide wire and fascial dilators. Heavy-duty guide wire and hydrophilic dilators with Coons taper (Cook) may be helpful. Careful attention must be paid to securing the PAD catheter.

Drainage catheters and securing devices

The drainage catheter's effectiveness is determined chiefly by its inner diameter and kink resistance. Determine the catheter choice with the needle aspiration test (NAT). If the fluid can be easily aspirated (1 mL in 1 s) by a 10-mL syringe through an 18-gauge needle, then the abscess is drained with the 8.5-French catheter. Gobien and Park's published data support this approach. The catheter may be used for 3-6 months.

Most abscess fluid should pass the NAT. Viscous or debris-laden fluid necessitates the use of larger-diameter catheters: 10-14 French for complex abscesses and 24-30 French for severe pancreatic necrosis. If no significant fluid can be aspirated, biopsy should be considered.

Viscous collections require larger-diameter catheters and more aggressive flushing. Septations may be disrupted with a guide-wire maceration technique once initial drainage subsides. Multiple catheters may be required. Some authors have had success using thrombolytic agents (urokinase, tissue plasminogen activator [t-PA]) for complex abscesses. Thrombolytics may disrupt fibrin and bacterial cell structures.

Commercial drainage catheters come in a wide variety of sizes and materials. Factors to evaluate are cost, comfort, flexibility, ease of tracking (ability to follow the guide wire), kink resistance, inner diameter, side-hole diameter, durability, tip configuration, securing mechanism, and catheter coating.

Most drainage catheters are placed with an inner stiffening cannula using an over-the-wire technique. There is usually an option to use a traditional metal cannula or a newer flexible plastic cannula. When the metal cannula is used, the catheter-cannula combination is advanced as a unit until the abscess is entered. The cannula is unlocked from the catheter. The catheter is deployed much as an IV catheter would be: the cannula is held steady while the catheter is advanced into the abscess.

The disadvantage of the metal cannula is that the catheter is more likely to kink once it is disconnected from the cannula. The author prefers to use a flexible plastic cannula that remains connected to the drainage catheter until the final position is achieved. Once the catheter is in place, the cannula is removed.

In cost-containment environments, the author has used a custom 35-cm, 7.1-French polyethylene pigtail catheter (Cook). The P7.1-38 series catheter has a 0.062-in. inner lumen and is adequate for most abscess fluids. The catheter is easy to track and has good kink resistance. It is durable enough to use for approximately 2 weeks. It does not have a locking pigtail suture, which is a minor disadvantage. Its low cost makes it ideal for short-term drainage procedures in cooperative patients or as a bridging catheter in a patient who will undergo multiple treatment stages. Because this catheter lacks a locking pigtail, extra care is needed in securing it so as to avoid accidental dislodgment.

Many drainage catheters come with the option of being deployed as a trocar by putting in an inner stylet. The author uses the trocar technique only for percutaneous cholecystostomy procedures that require a catheter with a small locking loop to facilitate the catheter traversing the often leatherlike gallbladder wall.

With regard to preventing inadvertent catheter dislodgment, the author has had good results using two adhesive-type catheter securing devices in tandem. One is placed where the catheter leaves the skin, and a second is situated closer to the catheter hub if room is available.

All extrathoracic drainage catheters should be placed to gravity drainage or bulb suction. The author has had good results with the Tru-Close system (Uresil), which comes either with or without a built-in suction bellows. A Pleur-Evac–type device may be considered for intrapleural collections to prevent tension pneumothorax. The catheter is flushed at least once per nursing shift with 10-20 mL of normal saline. The volume and frequency of catheter flush should adjusted on the basis of cavity size, drainage quality, and quantity.

Guide wires

Various guide wires are available for PAD, with different properties and prices. Guide wires should have the following characteristics:

  • Stiff enough to guide dilators and catheter into abscess
  • Not so stiff as to prevent easy coiling of wire shaft within the abscess
  • Floppy-tipped enough to encourage the wire to coil within the abscess and not perforate the abscess wall
  • Short enough to make use convenient

These requirements usually are satisfied by a 145-cm-long "heavy-duty" stainless steel wire with an outside diameter of 0.038 in., such as the Coons Interventional wire (THSF-38-145-COONS, Cook). Its shaft stiffness is between that of a standard angiographic wire and that of an Amplatz Extra/Super Stiff (Cook/Boston Scientific). It has a long floppy tip and costs less than $20.

Guide-wire stiffness depends on construction, material, and diameter. Stainless steel is cheaper and stiffer than nitinol but does not have nitinol's kink resistance. Wire stiffness is related by the fourth power of the radius. A guide wire's stiffness is derived from the inner mandrel, not the outer wire coiling. Most rigid guide wires are constructed as a solid mandrel with a thin outer coating. Stainless steel Lunderquist Exchange wire (Cook) and the newer nitinol-based Nitrix (EV3) are examples.