Abstract
Background and study aims Despite the widespread use of endoscopic ultrasound (EUS)-guided tissue acquisition, the choice of optimal suction technique remains a subject of debate. Multiple studies have shown conflicting results with respect to the four suction techniques: Dry suction (DS), no suction (NS), stylet slow-pull (SSP) and wet suction (WS). Thus, the present network meta-analysis (NMA) was conducted to compare the diagnostic yields of above suction techniques during EUS-guided tissue acquisition. Methods A comprehensive literature search from 2010 to March 2022 was done for randomized trials comparing the aspirated sample and diagnostic outcome with various suction techniques. Both pairwise and network meta-analyses were performed to analyze the outcomes: sample adequacy, moderate to high cellularity, gross bloodiness and diagnostic accuracy. Results A total of 16 studies (n=2048 patients) were included in the final NMA. WS was associated with a lower odd of gross bloodiness compared to DS (odds ratio 0.50, 95% confidence interval 0.24-0.97). There was no significant difference between the various suction methods with respect to sample adequacy, moderate to high cellularity and diagnostic accuracy. On meta-regression, to adjust for the effect of needle type, WS was comparable to DS in terms of bloodiness when adjusted for fine-needle aspiration needle. Surface under the cumulative ranking analysis ranked WS as the best modality for all the outcomes. Conclusions The present NMA did not show superiority of any specific suction technique for EUS-guided tissue sampling with regard to sample quality or diagnostic accuracy, with low confidence in estimates.