A 20-gauge celiac plexus neurolysis (CPN) needle or a standard 19-gauge needle was used for performing celiac plexus block or neurolysis. An on-site cytopathologist was available for rendering diagnosis in all cases. Diagnostic adequacy was defined as the ability to establish a preliminary diagnosis based on on-site analysis of FNA specimens. Technical failure was

defined as the need for use of more than one needle because of its dysfunction or the inability to successfully access and/or sample an organ or a lesion in an individual patient. At phase I, 625 needles were used in 548 patients (diagnostic FNAs = 487, interventions = 61), with an overall technical failure rate of 11.5% (TABLE 1 and TABLE 2). Of the 63 technical failures, 53 were FNAs and 10 were therapeutic interventions. Reasons for technical failure in the 53 diagnostic FNA cases were failure to deploy the needle out of the sheath in 38, kinking of the biopsy needle

check details at the handle in 3, bent needle tip that precluded adequate needle visualization in 9 (FNA of solid masses), and stylet dysfunction in 3. Reasons for technical failure in the 10 interventions were inability to deploy the needle out of the sheath in 7 and the needle being bent out of shape, thereby precluding adequate visualization see more in 3. Overall, more technical failures were observed with the use of 19-gauge versus 22- and/or 25–gauge needles (19.7% vs 8.8%; P = .004) and with transduodenal versus other routes (24.4% vs 5.2%; P < .001) for both diagnostic (technical failure in 10.9%) and therapeutic (technical failure in 16.4%) procedures. Of the 63 technical failures, 44 (70%) were encountered during transduodenal procedures. When evaluating technical failures

by the type of needle and route, compared to 25-gauge, a higher proportion of failures were observed with 19- and 22–gauge needles when the transduodenal route was navigated: 15 of 28 (53.6%) versus 12 of 14 (85.7%) and 17 of 21 (81.0%), respectively (P = .012). The overall diagnostic adequacy was 97.1%. Based on these these observations, an algorithm (Fig. 1) was developed with the objective of improving technical outcomes and resource use. As in phase I, all FNAs for tissue acquisition via the duodenum were performed by using the same 25-gauge needle and all other routes with a 22-gauge needle. Although all cyst aspirations (>2 cm in size) and interventions via the duodenum were performed by using the newly developed Flexible 19-gauge needle (Boston Scientific, Natick, Mass), a standard 19-gauge needle was used to perform these indications via other routes. Cyst lesions ≤2 cm in size were aspirated by using a 22-gauge needle, irrespective of its location. As in phase I, all celiac plexus blocks and neurolysis were undertaken by using a 20-gauge CPN or standard 19-gauge needle. This algorithm was then applied prospectively in phase II (September 2011 to April 2012) by 3 endosonographers.