The diagnostic accuracy of 22-gauge and 25-gauge needles in endoscopic ultrasound-guided fine needle aspiration of solid pancreatic lesions: a meta-analysis

• Introduction
• Methods
• Study identification
• Study eligibility
• Data extraction
• Outcomes for analysis
• Assessment of study quality
• Data synthesis and statistical analysis
• Results
• Study identifications and selection
• Description of variation in study methods
• Assessment of study quality
• Data synthesis
• Discussion
• References

Background and study aims: It is uncertain if needle gauge impacts the diagnostic accuracy of endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) of pancreatic mass lesions. Our aim was to use meta-analysis to more robustly define the diagnostic accuracy of EUS-FNA for pancreatic masses using 22 G and 25 G needles.

Patients and methods: Studies were identified by searching nine medical databases for reports published between 1994 and 2011, using a reproducible search strategy comprised of relevant terms. Only studies comparing the overall diagnostic accuracy of 22 G vs. 25 G EUS needles that used surgical histology or at least 6 months clinical follow up for a gold standard were included. Two reviewers independently scored the identified studies for methodology and abstracted pertinent data. When required, the original investigators were contacted to provide additional data. Pooling was conducted by both fixed-effects and random-effects models. Diagnostic characteristics (sensitivity, specificity, positive and negative likelihood ratios) with 95 % confidence intervals (CIs) were calculated.

Results: Eight studies involving 1292 subjects met the defined inclusion criteria. Of the 1292 patients, 799 were in the 22 G group and 565 were in the 25 G group (both needles were used in 72 patients). The pooled sensitivity and specificity of the 22 G needle were 0.85 (95 %CI 0.82 – 0.88) and 1 (95 %CI 0.98 – 1) respectively. The pooled sensitivity and specificity of the 25 G needle were 0.93 (95 %CI 0.91 – 0.96) and 0.97 (95 %CI 0.93 – 0.99) respectively. The bivariate generalized linear random-effect model indicated that the 25 G needle is associated with a higher sensitivity (P = 0.0003) but comparable specificity (P = 0.97) to the 22 G needle.

Conclusions: This meta-analysis suggests 25 G needle systems are more sensitive than 22 G needles for diagnosing pancreatic malignancy.

Introduction

Endoscopic ultrasound (EUS) provides detailed imaging of the pancreas and is useful for evaluating solid pancreatic lesions. When coupled with fine needle aspiration (FNA), EUS can provide both staging and cytological diagnosis of pancreatic malignancies. The diagnostic accuracy of EUS-FNA varies from 70 % – 98 % and may be influenced by multiple factors, including the location, size, and tissue firmness of the lesion, angulation of the endoscope tip, experience of the endoscopist, and availability of on-site cytopathology assessment [1] [2].

Different needle gauges have been used for pancreatic EUS-FNA including 19-gauge (19 G), 22-gauge (22 G), and 25-gauge (25 G) needles. The needle gauge may also impact diagnostic yield, though data are inconclusive [3]. It is often technically challenging to deploy 19 G needles in the duodenum and therefore these are used less commonly for pancreatic EUS-FNA. Several studies have assessed the performance characteristics of 22 G and 25 G needles for sampling pancreatic mass lesions, but most have failed to demonstrate superiority of either needle. The aim of this meta-analysis was to more robustly define the diagnostic accuracy of 22 G versus 25 G needles for EUS-FNA of pancreatic masses.

Methods

Study identification

All published studies that compared the diagnostic accuracy of 22 G vs. 25 G EUS needles in pancreatic masses that used surgical histology or at least 6 months of clinical follow-up as their gold standard were included. Studies were identified by searching nine medical databases including PubMed and Ovid MEDLINE for reports published between 1994 and October 2011. A reproducible search strategy was employed which combined the terms: “EUS,” “endosonography,” or “endoscopic ultrasound,” and “needle” or “FNA,” and “solid,” “mass,” “pancreas,” or “pancreatic,” and “22” or “22 g,” and “25” or “25 g.” References from retrieved articles were also manually reviewed.

Study eligibility

Two investigators independently evaluated studies for inclusion in the systematic review, and any disagreements were adjudicated by the senior investigator. Studies published only in abstract form were not included in this analysis. Investigators were not blinded to journal titles, author names, or institutional affiliations. Studies were included if they: (i) prospectively or retrospectively assessed patients with solid pancreatic masses using the two different needle gauges; (ii) used surgical histology or at least 6 months of clinical follow-up as their gold standard; and (iii) reported (or provided data allowing calculation of) the number of true positives, true negatives, false positives, and false negatives for each needle size.

Data extraction

Two study investigators extracted these data independently for each study: (i) publication year; (ii) country of origin; (iii) study design; (iv) patient demographics; (v) number of true-positive, true-negative, false-positive, and false-negative observations; (vi) number of passes required to achieve a diagnosis; (vii) needle manufacturer; (viii) data pertaining to technique including use of Doppler, use of negative suction, and number of to-and-fro passes; (ix) lesion size and location; (x) ease of use; and (xi) rate of needle malfunction.

Outcomes for analysis

The primary outcome measure for this meta-analysis was the comparison of the pooled sensitivities and specificities of the 22 G and 25 G needles in the diagnosis of pancreatic cancer in patients with solid pancreatic masses. Secondary outcomes included: positive and negative likelihood ratios; diagnostic accuracy by anatomic subsites (head, body/tail); the rates of needle malfunction; physician ease of use; and overall complications.

Assessment of study quality

The methodological quality of the selected studies was graded independently by two reviewers using the Quality Assessment of Diagnostic Accuracy Studies (QUADAS) tool, a validated tool for the quality assessment of diagnostic accuracy studies [4]. We performed component analysis using the review manager 5 tool [5], which was depicted as a proportional bar graph for each of the 11 individual criteria. Disagreement between the two extracting authors was resolved by consensus.

Data synthesis and statistical analysis

The sensitivity, specificity, positive likelihood ratio, and negative likelihood ratio were calculated for each study and were then pooled by using the Mantel–Haenszel fixed-effects method [6]. Estimates were also combined using the random-effects model by DerSimonian and Laird [7]. In the absence of significant heterogeneity (P > 0.1), the fixed-effects model results were presented. All pooled data were reported with the associated 95 % confidence intervals (CIs). A value of 0.5 was added to each cell in 2 × 2 tables with zero cell frequencies. All statistical tests were two-sided, and the significance level was set at 5 %. Heterogeneity was assessed by both χ² and I ² statistics [8] [9]. A P value < 0.10 (or a large χ² statistic relative to degrees of freedom) was considered evidence of heterogeneity beyond chance. An I ² value greater than 50 % was considered substantial heterogeneity.

The pooled sensitivity and specificity of the two needles were compared using the bivariate approach [10]. In addition, a summary receiver operating characteristic (SROC) curve was constructed using the DerSimonian – Laird random-effects model [7]. A perfect test has an area under the curve (AUC) close to 1; poor tests have AUCs close to 0.5. Meta-DiSc version 1.4 statistical software (Meta-DiSc, Unit of Clinical Biostatistics team of the Ramon y Cajal Hospital, Madrid, Spain) was used to calculate the sensitivity, specificity, positive likelihood ratio, and negative likelihood ratio for each study and then the pooled analysis [11]. It was also used to generate forest plots of sensitivity, specificity and to construct the SROC curves. SAS software version 9.2 (SAS Institute, Cary, North Carolina, USA) was used to compute the bivariate analysis.

Results

Study identifications and selection

The literature search yielded 36 potentially pertinent manuscripts for inclusion. Sixteen of these manuscripts were excluded immediately after initial review, most commonly because they did not pertain to the evaluation of pancreatic masses. Of the remaining twenty manuscripts, ten met the inclusion criteria for this meta-analysis ([Fig. 1]) [12] [13] [14] [15] [16] [17] [18] [19] [20] [21]. Two studies were subsequently excluded. One of these described the overall diagnostic accuracy, but reported insufficient data to allow calculation of specific operating characteristics (true positives, true negatives, false negatives and false positives) [19]. The other study reported results for overall diagnostic accuracy among two groups (pancreatic and others vs. submucosal lesions), so it was not possible to extract the data for pancreatic lesions alone [21]. Therefore eight studies were included in the final analysis.

Manual review of the references of retrieved manuscripts concerning EUS-FNA in pancreatic mass lesions did not yield any additional studies that met the inclusion criteria for this analysis.

Description of variation in study methods

Of the eight studies meeting inclusion criteria for the meta-analysis ([Table 1]), two were prospective nonrandomized studies that used both needles in all patients [12] [13], two were randomized controlled trials that used only one needle in each patient [14] [17], one was a randomized controlled trial with regard to the sequence of the two needles used in all patients [20], the remaining three were retrospective studies [15] [16] [18]. One study was published as a letter to the editor [16]; the rest were complete manuscripts [12] [13] [14] [15] [17] [18] [20]. Four studies were conducted in North America [13] [14] [15] [16], two in Italy [17] [20], and two in Japan [12] [18]. Six studies provided detailed patient demographic data [14] [15] [16] [17] [18] [20].

The number of patients included in these studies ranged from 10 to 842. All studies except one used histology and at least 6 months of follow-up as the reference standard; the other study used surgical histology only as its reference standard [13]. All studies used disposable 25 G and 22 G EUS-FNA needles, which incidentally all came from the same manufacturer (Echotip and Echotip Ultra, Cook Endoscopy, Winston Salem, North Carolina, USA). Six studies described the use of negative suction during FNA [12] [13] [14] [17] [18] [20]. One or more experienced endosonographers performed the procedures in six studies [12] [13] [14] [17] [18] [20], while the remaining studies did not describe the endoscopist(s) involved [15] [16]. Cytopathologists were blinded in five of the eight studies [12] [13] [14] [17] [20].

The primary author of one study was contacted to obtain additional data to allow calculation of the operating characteristics of the two needles [16]. Data regarding the number of patients with inadequate/nondiagnostic samples were mentioned in only four studies [12] [14] [15] [20]. All studies included patients with and without pancreatic malignancy with the exception of one study that included only patients with pancreatic malignancy; hence, only sensitivity could be calculated from this study [14].

In one study the total number of lesions was different from the total number of patients (120 vs. 115), as five patients had two discrete lesions. In our current analysis, each lesion was considered separately in determining the diagnostic characteristics [18].

Assessment of study quality

Overall, the quality of the reported studies was good and comparable ([Fig. 2]). The majority of the studies may be subject to review bias, as the investigators who used the patient’s clinical status at ≥ 6 months as a reference standard were generally not blinded to the result of the EUS-FNA. Differential verification bias could have been introduced in most of the included studies, as the index test result influenced the choice of reference standard (those with positive cytology were more likely to have surgery as reference standard compared with those with a negative test in whom clinical and radiographic follow-up was chosen). However, these two biases would have been difficult to avoid given the nature of the diagnostic process related to pancreatic masses. The limited period of clinical follow-up in these studies might also have resulted in misclassification bias.

Data synthesis

Eight studies involving 1292 subjects met the defined inclusion criteria. Of the 1292 patients, 799 were in the 22 G group and 565 were in the 25 G group (both needles were used in 72 patients). Eight studies were included in the sensitivity analysis and seven studies were included in the specificity analysis as one study comprised only patients with pancreatic cancer [14]. The between-study variability (i. e. heterogeneity) beyond what could be expected by sampling error was low with an I ² of 0.0 % for the pooled sensitivity of the 22 G needle and for the specificity of both needles. The I ² for the pooled sensitivity of the 25 G needle was 30 %, which was not considered substantial heterogeneity as per the a priori definition.

The pooled sensitivity and specificity of the 22 G needle were 0.85 (95 %CI 0.82 – 0.88) and 1 (95 %CI 0.98 – 1) respectively ([Fig. 3]). The pooled sensitivity and specificity of the 25 G needle were 0.93 (95 %CI 0.91 – 0.96) and 0.97 (95 %CI 0.93 – 0.99) respectively ([Fig. 4]). The bivariate generalized linear random-effect model indicated that the 25 G needle is associated with a higher sensitivity (P = 0.0003) but comparable specificity (P = 0.97) relative to the 22 G needle.

The pooled positive and negative likelihood ratio of the 22 G needle were 15.64 (95 %CI 4.03 – 60.63) and 0.16 (95 %CI 0.14 – 0.19) respectively. The pooled positive and negative likelihood ratio for the 25 G needle were 17.05 (95 %CI 8.35 – 34.86) and 0.09 (95 %CI 0.06 – 0.13) respectively. The symmetric curve shows a trade-off between sensitivity and specificity. The area under the SROC was 0.97 for the 22 G needle and 0.98 for the 25 G needle, which indicates high accuracy ([Fig. 5]).

Sensitivity analyses were performed and most of the mild heterogeneity noticed among the pooled sensitivity of the 25 G needle was related to the study by Uehara et al. [18]. However, removing this study from the analysis did not alter the results. Removing the study by Yusuf et al. (which was markedly larger than the other studies) yielded a similar conclusion, but just missed statistical significance. In this subanalysis, the pooled sensitivity of the 22 G needle was 0.87 (95 %CI 0.82 – 0.91) and the pooled sensitivity of the 25 G needle was 0.95 (95 %CI 0.91 – 0.98; P = 0.08).

We analyzed the retrospective and prospective studies separately. The pooled sensitivity estimates for the 22 G and 25 G needles among the retrospective studies were 0.84 (95 %CI 0.81 – 0.88) and 0.93 (95 %CI 0.89 – 0.96) respectively, and among the prospective studies were 0.87 (95 %CI 0.78 – 0.93) and 0.94 (95 %CI 0.89 – 0.97) respectively. These subgroup sensitivity estimates are concordant with the pooled sensitivity estimates for all of the studies combined.

The nature of the data reporting did not allow for evaluation of certain variables, such as diagnostic accuracy by anatomic subsites (head, body/tail) and physician ease of use. Three studies described the incidence of needle malfunction [12] [13] [14]. The rate of needle malfunction was similar between the two needles among these studies: 12.8 % (11/86) for the 22 G needle vs. 15.7 % (14/89) for the 25 G needle (P = 0.6). Of the five studies that reported adverse events, three described no complications with either needle [12] [13] [20], while one study reported a 2 % (11/540) risk of pancreatitis using the 22 G needle [15], and another reported two cases of pancreatitis (2 %) but did not specify the needle type [18].

Discussion

This systematic review and meta-analysis of EUS-FNA for solid pancreatic lesions included a large cohort of patients (n = 1292) and quantitatively summarizes the available evidence regarding the diagnostic performance of the two most commonly used needle gauges for EUS-FNA of solid pancreatic lesions. These results demonstrated that the 25 G needle has superior sensitivity to the 22 G needle in this clinical context.

The reason for the observed superior diagnostic accuracy of the 25 G needle in this analysis is uncertain. Anecdotally, 25 G needles are associated with fewer bloody aspirates than 22 G needles, which may potentially have a beneficial effect on cytological interpretation without compromising cellular yield. Also, in some solid lesions involving the uncinate process or medial aspect of the pancreatic head, needle passage may be technically easier with a 25 G needle than with a 22 G needle because of greater flexibility and less friction within the needle sheath. Finally, some pancreatic mass lesions are very firm, and achieving adequate to-and-fro traversal of the mass for the best possible acquisition of cytological material may be more easily achieved with a smaller caliber FNA needle in these cases.

The pooled results from this analysis are consistent with the trend from each of the individual studies, which lacked adequate sample size to reach statistical significance independently. The results of this analysis appear generalizable given the variety of clinical settings within the studies, as well as the types of patients, lesions, and needles used, which are similar to routine clinical practice.

Interestingly, all of the included studies used EUS-FNA needles from the same manufacturer. However, there is no obvious reason to suspect that needles of the same gauge from different suppliers would perform differently. Recently a needle has been developed that includes a reverse-facing bevel that may potentially facilitate shearing of tissue into the needle during retrograde movement of the needle [22]. Whether or not this needle will perform differently from conventional FNA needles for sampling of solid pancreatic lesions is not yet known.

The inclusion of only English language and fully published studies could have potentially excluded some relevant trials. Another potential limitation of this review is the use of imperfect criterion standards. The reference standards chosen were surgical histology or at least 6 months of clinical follow-up. It is conceivable that false-negative cytology, such as a missed cancer, might require longer than 6 months to become clinically apparent. While possible, this seems unlikely given the natural history of pancreatic adenocarcinoma and would probably not change the overall conclusion, as the same criteria were implemented for both needles.

Although selected trials were considered statistically homogeneous, variations in study characteristics with regard to study design, participant selection, use of the same needles on the same or different lesions, order of using the needles, presence or absence of blinded cytopathologists, and the number of endosonographers might all potentially introduce bias.

Most of the studies failed to provide detailed information regarding the handling of inadequate/nondiagnostic results. Between-needle variation in these cases might lead to a biased assessment of test performance. However, in the studies that have reported inadequate/nondiagnostic results, the number of these cases was limited and of similar proportion between the two needles.

A nonconcurrent comparison in study design might make some of the included studies susceptible to bias by changes over time in uncontrollable, confounding risk factors, such as different echoendoscope processors, or improvement in the experience of endoscopists and cytopathologists, which may account for the observed differences between the two needles. The study by Yusuf et al. might be subject to this bias as it was implemented over 7 years and the authors have clearly mentioned that the case distribution represents different periods of time according to the availability of the needle system. During the early period of the study only the 22 G needle was used for EUS-FNA; later on, when the 25 G needle became available, it was used exclusively.

In conclusion, the currently available literature demonstrates that a 25 G needle has better sensitivity and comparable specificity to a 22 G needle in EUS-FNA of solid pancreatic masses. As such, we suggest that this needle should be used in preference in clinical practice in appropriate patients. Future prospective studies may allow for eventual stratification of needle performance by anatomic subsite, elucidate the reasons for false-negative FNA, and provide a comparison of conventional needles with newer designs.

Competing interests: None.

• References

• 1 Bentz JS, Kochman ML, Faigel DO et al. Endoscopic ultrasound-guided real-time fine-needle aspiration: clinicopathologic features of 60 patients. Diagn Cytopathol 1998; 18: 98-109
  CrossrefPubMedSearch in Google Scholar
• 2 Mertz H, Gautam S. The learning curve for EUS-guided FNA of pancreatic cancer. Gastrointest Endosc 2004; 59: 33-37
  CrossrefPubMedSearch in Google Scholar
• 3 Itoi T, Itokawa F, Sofuni A et al. Puncture of solid pancreatic tumors guided by endoscopic ultrasonography: a pilot study series comparing Trucut and 19-gauge and 22-gauge aspiration needles. Endoscopy 2005; 37: 362-366
  Thieme ConnectPubMedSearch in Google Scholar
• 4 Whiting P, Rutjes AW, Reitsma JB et al. The development of QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews. BMC Med Res Methodol 2003; 3: 25
  CrossrefPubMedSearch in Google Scholar
• 5 Database C: Review Manager (RevMan). In: 31 August 2011 edn. Available at: http://ims.cochrane.org.accesdistant.sorbonne-universite.fr/revman Accessed: 30 October 2012
  PubMed
• 6 Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst 1959; 22: 719-748
  PubMedSearch in Google Scholar
• 7 DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986; 7: 177-188
  CrossrefPubMedSearch in Google Scholar
• 8 Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002; 21: 1539-1558
  CrossrefPubMedSearch in Google Scholar
• 9 Huedo-Medina TB, Sanchez-Meca J, Marin-Martinez F et al. Assessing heterogeneity in meta-analysis: Q statistic or I2 index?. Psychol Methods 2006; 11: 193-206
  CrossrefPubMedSearch in Google Scholar
• 10 Reitsma JB, Glas AS, Rutjes AW et al. Bivariate analysis of sensitivity and specificity produces informative summary measures in diagnostic reviews. J Clin Epidemiol 2005; 58: 982-990
  CrossrefPubMedSearch in Google Scholar
• 11 Zamora J, Abraira V, Muriel A et al. Meta-DiSc: a software for meta-analysis of test accuracy data. BMC Med Res Methodol 2006; 6: 31
  CrossrefPubMedSearch in Google Scholar
• 12 Imazu H, Uchiyama Y, Kakutani H et al. A prospective comparison of EUS-guided FNA using 25-gauge and 22-gauge needles. Gastroenterol Res Pract 2009; 2009: 546390
  PubMedSearch in Google Scholar
• 13 Lee JH, Stewart J, Ross WA et al. Blinded prospective comparison of the performance of 22-gauge and 25-gauge needles in endoscopic ultrasound-guided fine needle aspiration of the pancreas and peri-pancreatic lesions. Dig Dis Sci 2009; 54: 2274-2281
  CrossrefPubMedSearch in Google Scholar
• 14 Siddiqui UD, Rossi F, Rosenthal LS et al. EUS-guided FNA of solid pancreatic masses: a prospective, randomized trial comparing 22-gauge and 25-gauge needles. Gastrointest Endosc 2009; 70: 1093-1097
  CrossrefPubMedSearch in Google Scholar
• 15 Yusuf TE, Ho S, Pavey DA et al. Retrospective analysis of the utility of endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) in pancreatic masses, using a 22-gauge or 25-gauge needle system: a multicenter experience. Endoscopy 2009; 41: 445-448
  Thieme ConnectPubMedSearch in Google Scholar
• 16 Siddiqui AA, Lyles T, Avula H et al. Endoscopic ultrasound-guided fine needle aspiration of pancreatic masses in a veteran population: comparison of results with 22- and 25-gauge needles. Pancreas 2010; 39: 685-686
  CrossrefPubMedSearch in Google Scholar
• 17 Camellini L, Carlinfante G, Azzolini F et al. A randomized clinical trial comparing 22G and 25G needles in endoscopic ultrasound-guided fine-needle aspiration of solid lesions. Endoscopy 2011; 43: 709-715
  Thieme ConnectPubMedSearch in Google Scholar
• 18 Uehara H, Ikezawa K, Kawada N et al. Diagnostic accuracy of endoscopic ultrasound-guided fine needle aspiration for suspected pancreatic malignancy in relation to the size of lesions. J Gastroenterol Hepatol 2011; 26: 1256-1261
  CrossrefPubMedSearch in Google Scholar
• 19 Sakamoto H, Kitano M, Komaki T et al. Prospective comparative study of the EUS guided 25-gauge FNA needle with the 19-gauge Trucut needle and 22-gauge FNA needle in patients with solid pancreatic masses. J Gastroenterol Hepatol 2009; 24: 384-390
  CrossrefPubMedSearch in Google Scholar
• 20 Fabbri C, Polifemo AM, Luigiano C et al. Endoscopic ultrasound-guided fine needle aspiration with 22- and 25-gauge needles in solid pancreatic masses: a prospective comparative study with randomisation of needle sequence. Dig Liver Dis 2011; 43: 647-652
  CrossrefPubMedSearch in Google Scholar
• 21 Kida M, Araki M, Miyazawa S et al. Comparison of diagnostic accuracy of endoscopic ultrasound-guided fine-needle aspiration with 22- and 25-gauge needles in the same patients. J Interv Gastroenterol 2011; 1: 102-107
  PubMedSearch in Google Scholar
• 22 Giovannini M, Monges GM, Iglesias-Garcia J et al. Prospective multicenter evaluation of a novel 22-G Echo-Tip Procore Histology EUS-needle in patients with a solid pancreatic mass. Gastrointest Endosc 2011; 73: AB152-AB153 (Abstract)
  PubMedSearch in Google Scholar

Corresponding author

• References

• 1 Bentz JS, Kochman ML, Faigel DO et al. Endoscopic ultrasound-guided real-time fine-needle aspiration: clinicopathologic features of 60 patients. Diagn Cytopathol 1998; 18: 98-109
  CrossrefPubMedSearch in Google Scholar
• 2 Mertz H, Gautam S. The learning curve for EUS-guided FNA of pancreatic cancer. Gastrointest Endosc 2004; 59: 33-37
  CrossrefPubMedSearch in Google Scholar
• 3 Itoi T, Itokawa F, Sofuni A et al. Puncture of solid pancreatic tumors guided by endoscopic ultrasonography: a pilot study series comparing Trucut and 19-gauge and 22-gauge aspiration needles. Endoscopy 2005; 37: 362-366
  Thieme ConnectPubMedSearch in Google Scholar
• 4 Whiting P, Rutjes AW, Reitsma JB et al. The development of QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews. BMC Med Res Methodol 2003; 3: 25
  CrossrefPubMedSearch in Google Scholar
• 5 Database C: Review Manager (RevMan). In: 31 August 2011 edn. Available at: http://ims.cochrane.org.accesdistant.sorbonne-universite.fr/revman Accessed: 30 October 2012
  PubMed
• 6 Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst 1959; 22: 719-748
  PubMedSearch in Google Scholar
• 7 DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986; 7: 177-188
  CrossrefPubMedSearch in Google Scholar
• 8 Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002; 21: 1539-1558
  CrossrefPubMedSearch in Google Scholar
• 9 Huedo-Medina TB, Sanchez-Meca J, Marin-Martinez F et al. Assessing heterogeneity in meta-analysis: Q statistic or I2 index?. Psychol Methods 2006; 11: 193-206
  CrossrefPubMedSearch in Google Scholar
• 10 Reitsma JB, Glas AS, Rutjes AW et al. Bivariate analysis of sensitivity and specificity produces informative summary measures in diagnostic reviews. J Clin Epidemiol 2005; 58: 982-990
  CrossrefPubMedSearch in Google Scholar
• 11 Zamora J, Abraira V, Muriel A et al. Meta-DiSc: a software for meta-analysis of test accuracy data. BMC Med Res Methodol 2006; 6: 31
  CrossrefPubMedSearch in Google Scholar
• 12 Imazu H, Uchiyama Y, Kakutani H et al. A prospective comparison of EUS-guided FNA using 25-gauge and 22-gauge needles. Gastroenterol Res Pract 2009; 2009: 546390
  PubMedSearch in Google Scholar
• 13 Lee JH, Stewart J, Ross WA et al. Blinded prospective comparison of the performance of 22-gauge and 25-gauge needles in endoscopic ultrasound-guided fine needle aspiration of the pancreas and peri-pancreatic lesions. Dig Dis Sci 2009; 54: 2274-2281
  CrossrefPubMedSearch in Google Scholar
• 14 Siddiqui UD, Rossi F, Rosenthal LS et al. EUS-guided FNA of solid pancreatic masses: a prospective, randomized trial comparing 22-gauge and 25-gauge needles. Gastrointest Endosc 2009; 70: 1093-1097
  CrossrefPubMedSearch in Google Scholar
• 15 Yusuf TE, Ho S, Pavey DA et al. Retrospective analysis of the utility of endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) in pancreatic masses, using a 22-gauge or 25-gauge needle system: a multicenter experience. Endoscopy 2009; 41: 445-448
  Thieme ConnectPubMedSearch in Google Scholar
• 16 Siddiqui AA, Lyles T, Avula H et al. Endoscopic ultrasound-guided fine needle aspiration of pancreatic masses in a veteran population: comparison of results with 22- and 25-gauge needles. Pancreas 2010; 39: 685-686
  CrossrefPubMedSearch in Google Scholar
• 17 Camellini L, Carlinfante G, Azzolini F et al. A randomized clinical trial comparing 22G and 25G needles in endoscopic ultrasound-guided fine-needle aspiration of solid lesions. Endoscopy 2011; 43: 709-715
  Thieme ConnectPubMedSearch in Google Scholar
• 18 Uehara H, Ikezawa K, Kawada N et al. Diagnostic accuracy of endoscopic ultrasound-guided fine needle aspiration for suspected pancreatic malignancy in relation to the size of lesions. J Gastroenterol Hepatol 2011; 26: 1256-1261
  CrossrefPubMedSearch in Google Scholar
• 19 Sakamoto H, Kitano M, Komaki T et al. Prospective comparative study of the EUS guided 25-gauge FNA needle with the 19-gauge Trucut needle and 22-gauge FNA needle in patients with solid pancreatic masses. J Gastroenterol Hepatol 2009; 24: 384-390
  CrossrefPubMedSearch in Google Scholar
• 20 Fabbri C, Polifemo AM, Luigiano C et al. Endoscopic ultrasound-guided fine needle aspiration with 22- and 25-gauge needles in solid pancreatic masses: a prospective comparative study with randomisation of needle sequence. Dig Liver Dis 2011; 43: 647-652
  CrossrefPubMedSearch in Google Scholar
• 21 Kida M, Araki M, Miyazawa S et al. Comparison of diagnostic accuracy of endoscopic ultrasound-guided fine-needle aspiration with 22- and 25-gauge needles in the same patients. J Interv Gastroenterol 2011; 1: 102-107
  PubMedSearch in Google Scholar
• 22 Giovannini M, Monges GM, Iglesias-Garcia J et al. Prospective multicenter evaluation of a novel 22-G Echo-Tip Procore Histology EUS-needle in patients with a solid pancreatic mass. Gastrointest Endosc 2011; 73: AB152-AB153 (Abstract)
  PubMedSearch in Google Scholar