Tu1555 Impact of Needle Type and Gauge on the Performance of Endoscopic Ultrasound-Guided Sampling: A Meta-Analysis

Abstract

Impact of Needle Type and Gauge on the Performance of Endoscopic Ultrasound-Guided Sampling: A Meta-Analysis Young Oh*, Jeffrey L. Jackson Medical College of Wisconsin, Milwaukee, WI; Clement J. Zablocki VA Medical Center, Milwaukee, WI Background: There has been discordant data in the literature regarding which type and size of needle used for endoscopic ultrasound (EUS)-guided sampling has superior performance characteristics. AIM: The aim of this study was to comprehensively review and summarize the available literature on the performance of EUS-guided sampling based on the type and size of needle (19 gauge Trucut as well as 19, 22, and 25 gauge fine needle aspiration [FNA] needles) as well as sampling location. Methods: Published manuscripts until September 2011 regarding the use of EUS FNA or Trucut needles were identified in PubMed using relevant search terms. A review of the references from retrieved articles and abstracts from gastrointestinal meetings was also performed. Only studies that were published in English and compared two or more needles were included. Two reviewers independently reviewed each study for suitability to be included in the analysis. Data was pooled using a random effects model with variance based on exact binomial methods. Accuracy, sensitivity and specificity with 95% confidence intervals were calculated. Results: Forty-one articles and abstracts that included 3905 patients were identified that met the inclusion criteria. The combination of the Trucut needle and 22 gauge needle if necessary yielded the highest accuracy (0.95, 95% CI 0.93-0.97, p 0.03) and the 22 gauge needle alone yielded the lowest accuracy (0.76, 95% CI 0.72-0.80, p 0.001). A trend toward higher accuracies with the 25 gauge needle was observed, which did not reach statistical significance. The sensitivity using the Trucut needle was statistically worse than others (0.68, 95% CI 0.60-0.78, p 0.001) and no significant differences in specificity were observed with any of the needles. There appeared to be no differences by specific site of biopsy. Conclusions: The use of the Trucut needle in combination with the 22 gauge needle if necessary yielded the highest accuracy. For the use of single needles, a trend toward the highest accuracy was observed with the 25 gauge needle but this did not reach statistical significance.