The institutional review board of the University of Alabama at Birmingham approved the study and granted a waiver of written informed consent, given that standard U.S. Food and Drug Administration–approved accessories were used for approved indications, and the only technical function was assessed during standard-of-care procedures. The main outcome measure was BLZ945 supplier to compare rates of technical failure between phases I and II. The secondary measures were to compare the rates of diagnostic adequacy and procedural complications and the average cost of an FNA needle

per individual patient. Baseline patient characteristics, procedure outcomes, and average cost of needle per individual patient were calculated for phases I and II. For comparison of categorical data between the two phases, a chi-square or Fisher exact test was used as indicated. For continuous data, the 2-sample t test was performed for comparison of patient age, and the Wilcoxon rank-sum test was

used for comparison of the needle cost data. Statistical significance was determined to be a P value of less than .05. Datasets were compiled by using Microsoft Excel (Microsoft, Redmond, WA, USA), and all statistical analyses were performed by using Stata 10 (StataCorp, College Station, TX, USA). In phase II, 500 consecutive patients underwent EUS-FNA and/or interventions over the 7-month period. With the exception of age, there was no difference in patient demographics or procedural indications between phases I and II (TABLE 1 and TABLE 2). By adapting the algorithm, compared to phase I, more 19- and 22–gauge needles were used see more in phase II (Table 2). More patients in phase II underwent transduodenal FNAs compared with patients in phase

I. After exclusion of patients who required sampling of more than one site (n = 4), the overall rate of technical failure in phase II was found to be significantly Parvulin less compared with that of phase I, 1.6% versus 11.5% (P < .001). This difference in technical failure was significant for both diagnostic FNAs (10.9% vs 1.8%; P < .001) and therapeutic interventions (16.4% vs 0%; P = .001) between phases I and II, respectively. All 8 technical failures in phase II were encountered during diagnostic FNA procedures that included stylet dysfunction in 1 patient who underwent a transgastric cyst aspiration by using a standard 19-gauge needle and the 25-gauge needle not being able to exit the sheath during transduodenal FNAs in 7 patients. When technical failures were evaluated based on needle type, compared with phase I, needle dysfunction was less common for both 19- and 22–gauge needles in phase II, 19.7% versus 0.8% (P < .001) and 12.3% versus 0% (P < .001), respectively. There was no difference in rates of technical failure for the 25-gauge needle between phases I and II at 7.3% versus 3.