Diagnostic Peritoneal Lavage

Updated: Oct 20, 2017
  • Author: Liudvikas Jagminas, MD, FACEP; Chief Editor: Vikram Kate, MBBS, MS, PhD, FACS, FACG, FRCS, FRCS(Edin), FRCS(Glasg), FIMSA, MAMS, MASCRS  more...
  • Print


At one time, diagnostic peritoneal lavage (DPL), described by Root in 1965, [1]  was the diagnostic test of choice to detect bleeding within the abdominal cavity after trauma. However, the advent and widespread availability of computed tomography (CT), which carries near-comparable sensitivity and superior specificity, [2]  have limited the use of DPL to the diagnosis of intra-abdominal hemorrhage in unstable trauma patients.

Currently, DPL is performed infrequently, having been largely replaced by focused assessment with sonography for trauma (FAST) and CT. The American College of Surgeons (ACS) adopted FAST into the Advanced Trauma Life Support (ATLS) protocol, and the ninth edition of ATLS made DPL an optional skill station, owing to the widespread use of FAST. Still, each of these modalities has unique advantages and disadvantages. DPL retains its usefulness, especially in the hemodynamically unstable trauma patient who has a negative or equivocal FAST examination. (See Technical Considerations.)



DPL can be used to evaluate both blunt and penetrating abdominal trauma in patients who are hemodynamically unstable or who require urgent surgical intervention for associated extra-abdominal injuries. DPL can rapidly confirm or exclude the presence of intraperitoneal hemorrhage. Thus, the patient with a closed head injury, the unstable patient who has been in a motor vehicle accident, or the patient with a pelvic fracture and potential retroperitoneal hemorrhage can be appropriately triaged to emergency laparotomy.

A negative result on peritoneal aspiration allows the clinician to proceed to alternative management steps and allows the patient to forgo unnecessary laparotomy. Additionally, DPL can be used in nonemergency circumstances as a means of detecting solid-organ injury or HVI requiring laparotomy. [3, 4]

In the evaluation of patients with blunt abdominal trauma or penetrating anterior abdominal stab wounds, DPL provides the following benefits:

  • Rapid determination of the presence or absence of hemoperitoneum
  • Detection of intraperitoneal injury requiring laparotomy in stable patients
  • Rapid triage when the source of hemodynamic instability is unknown

Chereau studied DPL in 37 blunt abdominal trauma patients who had one or two CT signs predictive of small-bowel and mesenteric injuries. [5] A cell count ratio (CCR) was calculated, in which the ratio of white blood cells (WBCs) to red blood cells (RBCs) (WBC/RBC ratio) in the lavage fluid was divided by the ratio in peripheral blood. DPL was found to have a sensitivity of 100% but a specificity of only 43% for bowel injuries. The authors suggested that diagnosis might be improved by restricting the indications for exploratory laparotomy to patients with a CCR of 4 or higher.



An obvious need for laparotomy is the only absolute contraindication for DPL. Lack of training or familiarity with performing DPL, prior abdominal surgery, abdominal-wall infections, coagulopathy, morbid obesity, and second- or third-trimester pregnancy are all relative contraindications. [6]


Technical Considerations

Best practices

A positive FAST examination (hemoperitoneum) is useful and reliable in the hemodynamically unstable blunt trauma patient. However, if the FAST examination is negative or equivocal, it should be followed by DPL. DPL is 100% accurate for intra-abdominal injury in hemodynamically unstable patients, whereas FAST is positive only 45% of the time. [7]  However, DPL also takes 10-15 minutes, and the patient must be stable for the test to be carried out.

In the hemodynamically stable patient, CT is preferred because it is noninvasive and highly accurate. If CT is unavailable, either FAST or DPL may be used. DPL should also be considered in patients who have an unreliable examination or those at high risk for hollow-viscus injury (HVI), particularly when CT or ultrasonography detects minimal fluid or when the patient manifests fever, peritonitis, or both. This circumstance usually occurs 6-12 hours after an HVI.

DPL, though lacking organ specificity, remains the most sensitive test for mesenteric injury and HVI. FAST is rapid, noninvasive, and can be repeated multiple times; however, it is more user-dependent than DPL or CT. Both FAST and DPL fail to evaluate retroperitoneal and diaphragmatic injuries and poorly identify solid-organ injuries. Abdominopelvic CT requires a hemodynamically stable patient, is costly, and carries a small but significant lifetime risk of malignancy. However, CT reliably diagnoses solid-organ injuries and evaluates the retroperitoneum, but it is less sensitive and specific for HVIs and mesenteric injuries than DPL is. [8]

As a result of these differences (see Table 1 below), all three tests continue to play important roles in the evaluation of a trauma patient for abdominal injuries. [9]

Table 1. Comparison Parameters for DPL, FAST, and CT (Open Table in a new window)

Parameter DPL FAST CT
Time 10-15 min 2-4 min Variable
Repeatability Possible, but rarely done Easy and frequently done Yes
Reliability Not organ specific Operator dependent Obesity, movement
Sensitivity High Medium High
Specificity Low High High
Advantages Inexpensive, mobile, detects bowel injury Noninvasive, rapid, mobile, moderately expensive (equipment) Noninvasive, highly accurate, fixed, expensive (equipment)
Disadvantages Invasive, misses retroperitoneal and diaphragm injuries Hampered by subcutaneous or intra-abdominal air, obesity, pelvic fractures Misses diaphragm, small bowel, and pancreatic injuries; radiation

Kumar et al, in a prospective randomized trial comparing DPL (n=102) with FAST (n=98) in 200 consecutive patients (mean age, 28.3 years) who had sustained blunt (n=124) or penetrating (n=76) trauma to the torso, found DPL to be significantly superior to FAST for detecting bowel injuries, though it also took significantly longer to perform. [10]