BACKGROUND: Penetrating abdominal injuries without clear indication for laparotomy are a dilemma for surgeons. Delayed celiotom increases the morbidity and mortality while unnecessary laparotomies are inefficient. Diagnostic peritoneal lavage (DPL) is re-explored to determine its accuracy in penetrating abdominal trauma.

METHODS: All adult patients with penetrating abdominal injuries without indications for laparotomy at the Philippine General Hospital from September 1999 to March 2001 were included in the study. All underwent DPL and standard management in monitoring for penetrating abdominal injuries and await indications for laparotomy. Using various cut-offs for lavage red blood cell count (LRBC) as criteria in interpreting DPL results and, using either intraoperative findings or follow-up patient examination findings (for those who were not operated on) as gold standard, measures of accuracy were estimated. Receiver operating curves (ROC) were generated using various cut-offs ad the most clinically acceptable criteria (cut-offs) were selected.

RESULTS: Of the 213 cases, 69% were operated on and the rest where followed up for 2 weeks to determine if there were significant intra-abdominal injuries. The ideal cut-off for LRBC count was greater than or equal to 5000/cu mm (greater than or equal to 5000/cu mm as positive and <5,000/cu mm as negative) with a sensitivity of 93.2% (95% CI: 87.2, 96.7), specificity of 87.5% (95% CI: 77.8, 93.5) and a false positive rate of 12.5%. LWBC count showed no correlation to significant injuries.

CONCLUSIONS: DPL may be utilized as part of the selective management of penetrating abdominal injuries without clear indications of laparotomy. Using the LRBC count criterion of 5,000/cu mm, sensitivity is high and the false positive rate is acceptable.