Open or closed diagnostic peritoneal lavage for abdominal trauma? A meta-analysis

Objectives: To perform a meta-analysis of prospective, randomized controlle d trials comparing the closed and open technique of diagnostic peritoneal l avage (DPL) in trauma patients to determine whether there are any differenc e in outcomes. Methods: A search of MEDLINE database of English language articles publishe d from 1977 to 1999 was conducted by using the terms diagnostic peritoneal lavage, trauma, and randomized controlled trials, A manual search and Cochr ane Library database search was also conducted. Seven randomized controlled trials, including a total of 1,126 patients were identified that compared closed versus open technique. Two reviewers assessed the trials independent ly. Trial quality was critically appraised by using the Jadad Instrument, a validated published quality scale. Data extraction of major complications, technical difficulties, procedure times, and false-negative and false-posi tive rates was carried out. The fixed effects model was used for statistica l analysis. The Peto odds ratio (OR), weighted mean differences and 95% con fidence intervals (95% CI) were calculated. Results: The overall quality of studies was poor (mean, 2.4/7), Major compl ications did not differ significantly between closed versus open technique (OR, 0.65; 95% CI, 0.15 to 2.92, Technical failures and difficulties were s ignificantly higher in the closed group, i.e., OR 4.33 (95% CI, 1.96 to 9.5 6) and OR 4.19 (95% CI, 2.842 to 6.19), respectively. Accuracy of closed an d open DPL was comparable with no difference in false-negative or false-pos itive rates between the two techniques. Procedure time was consistently low er in the closed technique. Conclusions: The closed DPL technique is comparable to the standard open DP L technique in terms of accuracy and major complications. The advantage of reduced time to perform the closed DPL is offset by the increased technical difficulties and failures of this group. Therefore, any significant benefi t of routine dosed DPL in improving outcomes can be excluded with more conf idence based on pooled data than by the individual trials alone.