A Prerotational, Simulation-Based Workshop Improves the Safety of Central Venous Catheter Insertion

Transcription

1 CHEST Original Research A Prerotational, Simulation-Based Workshop Improves the Safety of Central Venous Catheter Insertion Results of a Successful Internal Medicine House Staff Training Program CRITICAL CARE Hiroshi Sekiguchi, MD ; Joji Erik Tokita, MD ; Taro Minami, MD ; Lewis Ari Eisen, MD, FCCP ; Paul Henry Mayo, MD, FCCP ; and Mangala Narasimhan, DO, FCCP Background: The purpose of this study was to evaluate the effectiveness of a simulation-based workshop with ultrasonography instruction in reducing mechanical complications associated with central venous catheter (CVC) insertion. Methods : A single-center prospective cohort study was conducted in the medical ICU and respiratory step-down unit of an urban teaching hospital. Fifty-six medical house staff members were trained prior to their rotations over a 6-month period. The data on mechanical complication rates after the implementation of the workshop were compared with previous experience when no structured educational program existed. Results : There were 334 procedures in the preeducation period compared to 402 procedures in the posteducation period. The overall complication rate, including placement failure, in the preeducation and posteducation period was 32.9% and 22.9%, respectively ( P,.01). Placement failure rate decreased from 22.8% to 16.2% ( P 5.02), and arterial punctures decreased from 4.2% to 1.5% ( P 5.03). Ultrasonography usage increased from 3.0% to 61.4% ( P,.01). Multivariate analysis demonstrated that interns were more likely to cause overall mechanical complications compared with fellows and attending physicians in the preeducation period ( P 5.02); however, this trend was not observed in the posteducation period. Catheter site and ultrasonography usage significantly affected the overall complication rate in both periods, and ultrasound-guided femoral CVC was the safest procedure in the posteducation period. Conclusions : Implementation of a prerotational workshop significantly improved the safety of CVC insertion, especially for CVCs placed by inexperienced operators. We suggest that simulationbased training with ultrasonography instruction should be conducted if house staff members are responsible for CVC placement. CHEST 2011; 140(3): Abbreviations: CVC 5 central venous catheter; IJ 5 internal jugular; MICU 5 medical ICU; RSDU 5 respiratory step-down unit Central venous catheter (CVC) placement is associated with mechanical complications in 5% to 19% of patients. 1-4 Prior studies have shown that the operator s level of training significantly affects overall complication rate. 4,5 Placement of a CVC often is performed by house staff in teaching hospitals, and physician training to improve technique is essential for patient safety. Studies also have demonstrated that the use of two-dimensional ultrasonography decreases mechanical complication rates Despite its proven 652 benefits, training in ultrasound-guided CVC insertion is not mandatory in internal medicine residency training. Recently, simulation-based training has gained increasing attention in various fields of medicine. Implementation of simulation-based training on CVC placement decreased both mechanical 11 and infectious complications. 12,13 There is emerging evidence on efficacy but no consensus on timing, duration, and content of training for medical house staff on CVC placement. Original Research

2 We hypothesized that a brief, but organized prerotational, simulation-based workshop leads to reduction in mechanical complications associated with CVC insertion. We also hypothesized that after limited structured training, more house staff would use ultrasonography for CVC insertion. Our goal was to compare mechanical complication rates after the implementation of the prerotational educational workshop to our previous experience in when no structured educational program existed. Materials and Methods The study was conducted at Beth Israel Medical Center, a 697-bed urban teaching hospital, and at the Manhattan campus of the Albert Einstein College of Medicine. Data were collected on patients admitted to the 16-bed medical ICU (MICU) and the 12-bed respiratory step-down unit (RSDU). We designed and provided our educational workshop to 56 medical house staff members with a 2:1 trainee-to-faculty ratio within 1 week of the start of their MICU/RSDU rotations over a 6-month period. Data on mechanical complications were prospectively collected beginning December We compared our data to a prior 6-month study conducted at the same institution. 14 The prior study data were collected retrospectively from a quality improvement database. All patients were aged 18 years. There were no exclusion criteria. The Beth Israel Medical Center institutional review board approved the study (Project ) with a requirement for informed consent from physicians. All the participating house staff members provided informed consent prior to the workshop. A requirement for informed consent from patients was waived. Training started with a pretraining test to evaluate knowledge required for CVC placement and ultrasonography usage (e-appendix 1). After the pretraining test, 1 h of interactive teaching was administered using computer slides and video demonstration. This component covered indications for CVC placement, contraindications, sterile technique, use of a central line bundle, review of relevant anatomy for all sites, and identification of vascular stenosis and DVT. Following the 1-h interactive teaching, participants completed 45 min of hands-on training. Two ultrasound-compatible torso Manuscript received December 27, 2010; revision accepted May 18, Affiliations: From the Division of Pulmonary and Critical Care Medicine (Dr Sekiguchi), Mayo Clinic, Rochester, MN; Division of Nephrology (Dr Tokita), Mount Sinai School of Medicine, New York, NY; Division of Pulmonary, Critical Care and Sleep Medicine (Dr Minami), Memorial Hospital of Rhode Island, Pawtucket, RI; Division of Critical Care Medicine (Dr Eisen), Department of Medicine, Albert Einstein College of Medicine, Montefiore Medical Center, Bronx, NY; and the Division of Pulmonary, Critical Care and Sleep Medicine (Drs Mayo and Narasimhan), North Shore Long Island Jewish Medical Center, New Hyde Park, NY. The abstract of this article was presented at CHEST 2008, Philadelphia, Pennsylvania, on October 27, Funding/Support: The authors have reported to CHEST that no funding was received for this study. Correspondence to: Hiroshi Sekiguchi, MD, Division of Pulmonary and Critical Care Medicine, Mayo Clinic, 200 First St SW, Rochester, MN 55905; sekiguchi.hiroshi@mayo.edu. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians ( site/misc/reprints.xhtml ). DOI: /chest mannequins (Blue Phantom; Redmond, Washington); one handmade thigh mannequin; and two ultrasound machines, MicroMaxx with 13-6 MHz transducer (SonoSite, Inc; Bothell, Washington) and Acuson Cypress C/V with MHz transducer (Siemens Medical Solutions Inc; Malvern, Pennsylvania) were used. The medical house staff were trained in landmark subclavian, ultrasoundguided internal jugular (IJ), and ultrasound-guided and landmark femoral CVC insertions. Neither landmark IJ nor ultrasoundguided subclavian insertion was recommended because of potential risks and limited evidence. 7 Following hands-on training, a posttraining test was given to confirm understanding of the procedure and the use of ultrasonography. In accordance with hospital policy, house staff members were supervised while placing CVCs until deemed competent to perform the procedure independently. Competency required placement of 10 subclavian, 10 IJ, and five femoral venous catheters or 10 ultrasound-guided procedures. The supervision criteria were the same between preeducation and posteducation periods. Medical house staff members were allowed to place CVCs with the ultrasound-guided technique or the landmark technique at their discretion. Choice was guided by the clinical situation, ultrasonography availability, patient characteristics, and personal experience. The choice of CVC sites was also at the discretion of the operator in both supervised and independently placed procedures. Removal of femoral CVCs within 48 h was advised because of concerns of increased infectious and thrombotic risks. 1 We constructed a prospective database of consecutive CVC placement attempts over a 6-month period beginning December 2007 and reviewed our previous database of a 6-month period from 2003 to Only the first attempt was recorded to avoid confounding of patient factors. Patient characteristics included age, sex, BMI, and level of consciousness. Operator and supervisor characteristics were the number of years of postgraduate training. Procedure characteristics consisted of use of ultrasonography guidance, site of insertion, time of procedure (day defined as 7:00 am to 7:00 pm, night as 7:00 pm to 7:00 am ), number of percutaneous punctures, the patient care unit where the procedure was performed, and the level of emergency (emergent or elective). Overall complications included placement failure and other mechanical complications. Data on anatomic variance and presence of DVT reported during the procedure were collected. CVC infection rates were gathered from our institutional data. Only partial data in the preeducation period were retrievable; however, the number of CVC infections in MICU and RDSU were all captured in the posteducation period. Quantitative parameters are expressed as mean SD. Potential risk factors for mechanical complications were analyzed in a bivariate analysis by using the x 2 test. Characteristics associated with complications with P,.02 were entered into a multivariate logistic regression analysis, with complications as the dependent variable and various parameters as the independent variables. All P values are based on two-tailed tests of significance. P,.05 is considered statistically significant. All computations were performed by JMP, version 8.0 (SAS Institute Inc; Cary, North Carolina) software on data imported from a Microsoft Excel 2003 (Microsoft Corporation; Redmond, Washington) spreadsheet. Results There were 402 CVC placements over a 6-month period during 2007 to 2008 when the formal educational workshop took place (posteducation period) compared with 334 placements during 2003 to CHEST / 140 / 3 / SEPTEMBER,

3 in the preeducation period. Baseline characteristics of the two study periods are presented in Table 1. Patient consciousness, level of training of operators and supervisors, patient care unit (MICU or RSDU), level of emergency under which the procedure was performed, and CVC type (insertion site and ultrasonography usage) were significantly different between the two periods. The number of various types of emergent or elective CVC insertions performed by each level of operator is further described in detail in e-tables 1 and 2. The emergent procedures significantly increased from 27.3% in the preeducation period to 44.8% in the posteducation period ( P,.01). The proportion of overall procedures performed by fellows and attending physicians changed from 19.5% in the preeducation period to 31.1% in the posteducation period ( P,.01). The number of CVCs placed by attending physicians was only two and one in the preeducation and posteducation period, respectively. Overall, in the preeducation period, the most common CVC placement procedure was landmark subclavian (59.9%) followed by landmark femoral (32.0%). The most commonly performed CVC placement procedure in the posteducation period was ultrasound-guided femoral (33.3%) followed by landmark subclavian (31.3%) and ultrasound-guided IJ (26.4%). Mechanical complications in each period are presented in Figure 1. The overall complication rate, including placement failure, in the preeducation and posteducation period was 32.9% and 22.9%, respectively ( P,.01). There was a trend toward a decrease in nonfailure mechanical complications (10.2%-6.7%, P 5.09). Placement failure rate was 22.8% and 16.2%, respectively ( P 5.02). The arterial puncture rate significantly decreased from 4.2% to 1.5% ( P 5.03). Other less-common complications were not significantly different between periods. The number of percutaneous punctures was significantly lower in the posteducation period (2.6 vs 1.5, P,.01). Mechanical complications by type of CVC (insertion site and ultrasonography usage) are described in Table 2. In both periods, overall complication and placement failure rates were significantly affected by type of CVC insertion performed. Landmark IJ CVC placement had the highest overall complication and failure rates in either period. In the preeducation period, ultrasound-guided IJ CVC placement had the lowest overall complication rate (n 5 10) followed by landmark femoral (n 5 107). In the posteducation period, ultrasound-guided femoral CVC placement had the lowest overall complication and failure-toplace rates (n 5 134). Other complication rates were small in number and not affected by the type of line in either time period. Mechanical complications from the use of ultrasonography are presented in Figure 2. In the preeducation period, ultrasound-guided CVC placement was small in number (n 5 10) and ultrasonography usage did not decrease overall complications, including placement failure (10.0% vs 33.6%, P 5.17), nonfailure mechanical complications (10.0% vs 10.2%, P ), or placement failure rates (0.0% vs 23.5%, P 5.12). In the posteducation period, ultrasonography usage significantly decreased the overall complications (12.2% vs 40.0%, P,.01) and placement failure rate (6.5% vs 31.6%, P,.01) compared with landmark CVC placement. Nonfailure mechanical complication rate was not significantly different between ultrasound-guided and landmark CVC placement (5.67% vs 8.39%, P 5.29). In order to further analyze the effect of ultrasonography use, landmark and ultrasound-guided CVC insertions were compared in the posteducation period. Because the numbers of landmark IJ and ultrasound-guided subclavian CVC insertions were small (n 5 3 and n 5 4, respectively), only landmark femoral (n 5 26) and ultrasound-guided femoral placements (n 5 134) were analyzed. The use of ultrasonography in femoral CVC insertion significantly decreased overall complications (7.5% vs 34.6%, P,.01), placement failure (5.2% vs 26.9%, P,.01), and arterial puncture (0.75% vs 7.7%, P,.01). Mechanical complication by level of operators was analyzed. In the preeducation period, the level of training of operators had a significant effect on overall complication rate (interns, 41.2%; residents, 34.3%; fellows and attending physicians, 20.0%; P 5.03). In the posteducation period, it did not have a significant effect on overall complication rate (interns, 17.7%; residents, 26.3%; fellows and attending physicians, 17.6%; P 5.13). The overall complication rate in CVC placements performed by interns (postgraduate year 1) decreased from 41.2% in the preeducation period to 17.7% in the posteducation period ( P 5.03). There was no significant difference in mechanical complication rates in CVC placements performed by residents (postgraduate years 2 and 3) between the two periods (32.3% vs 26.3%, P 5.07). In either education period, a higher number of percutaneous punctures resulted in a higher overall complication, nonfailure complication, and placement failure rates. CVC insertions with more than two percutaneous punctures had overall complication rates of. 50% in either period. Variables associated with complication rates with P,.2 were entered into a multivariate analysis ( Table 3 ). Catheter type (insertion site and ultrasonography usage) significantly affected the overall complication rate in either study period. In the preeducation period, landmark femoral and ultrasound-guided 654 Original Research

4 Table 1 Baseline Characteristics of Two Study Periods Characteristics Preeducation (n ) Posteducation (n )P Value Patient characteristics Age, y Male sex 203 (60.8) 225 (56.0).19 BMI Unconscious 241 (72.2) 257 (63.9).02 Operator characteristics,.01 Interns (PGY-1) 68 (20.4) 34 (8.5) Residents (PGY-2 and PGY-3) 201 (60.2) 243 (60.4) Fellows and attending physicians 65 (19.5) 125 (31.1) Supervisor characteristics,.01 No supervisors 97 (29.0) 139 (34.6) Residents (PGY-2 and PGY-3) 102 (30.5) 66 (16.4) Fellows and attending physicians 135 (40.4) 197 (49.0) Procedure characteristics MICU 312 (93.4) 389 (96.8).04 Day (7:00 am-7:00 pm ) 232 (69.5) 281 (69.9).90 Emergency 91 (27.3) 180 (44.8),.01 Catheter insertion type,.01 Landmark IJ 17 (5.1) 3 (0.7) Ultrasound IJ 10 (3.0) 109 (26.4) Landmark subclavian 200 (59.9) 126 (31.3) Ultrasound subclavian 0 (0) 4 (1.0) Landmark femoral 107 (32.0) 26 (6.5) Ultrasound femoral 0 (0) 134 (33.3) Total ultrasound usage 10 (3.0) 247 (61.4),.01 Data are presented as mean SD or No. (%). IJ 5 internal jugular; MICU 5 medical ICU; PGY 5 postgraduate year. IJ CVC placements were the safest procedures. In the posteducation period, ultrasound-guided femoral CVC placement was the safest procedure. In the preeducation period, interns were more likely to cause mechanical complications compared with fellows and attending physicians ( P 5.02). A similar trend toward an increased complication rate was seen in the multivariate analysis comparing residents to fellows and attending physicians ( P 5.05). In the posteducation period, no difference was observed in overall mechanical complication rate by level of training of operators. One incidental DVT was found in the femoral vein after a landmark femoral CVC placement attempt failed. Two cases of IJ vein stenosis were found during ultrasound-guided IJ CVC placement, which ended up in one placement failure and one success with repositioning. There were three central line-associated bloodstream infections reported during the 6-month posteducation period, which correspond to an infection rate of 1.6 of 1,000 device days. The data were obtained from the institutional monthly infection data. Complete data were not available during the preeducation period; however, the central lineassociated bloodstream infection rate from September through November in 2003 and 2004 were 6.0 of 1,000 device days and 8.0 of 1,000 device days, respectively. Fifty-six house staff members underwent pretraining and posttraining tests during the workshop. The mean SD pretraining and posttraining scores were and , respectively, out of a maximum of 10 ( P,.01). Discussion Implementation of a prerotational workshop with simulation and ultrasonography guidance significantly Figure 1. Mechanical complications by study period. The overall complication rate in the preeducation and posteducation period was 32.9% and 22.9%, respectively ( P,.01). Placement failure rate was 22.8% and 16.2%, respectively ( P 5.02). The arterial puncture rate was 4.2% to 1.5%, respectively ( P 5.03). Comp 5 complication. CHEST / 140 / 3 / SEPTEMBER,

5 Table 2 Overall Mechanical Complication and Placement Failure Rates by Catheter Type Complication Types Landmark IJ Ultrasound IJ Landmark SC Ultrasound SC Landmark FM Ultrasound FM P Value a Overall complications, including placement failure Preeducation 8/17 (47.1) 1/10 (10.0) 76/200 (38.0) 0/0 25/107 (23.4) 0/0.01 Posteducation 2/3 (66.7) 19/109 (17.4) 51/126 (40.5) 1/4 (25.0) 9/26 (34.6) 10/134 (7.5),.01 P Valueb N/A.24 N/A N/A Placement failure Preeducation 7/17 (41.2) 0/10 (0) 52/200 (26) 0/0 17/107 (15.9) 0/0.02 Posteducation 2/3 (66.7) 8/109 (7.3) 40/126 (31.8) 1/4 (25.0) 7/26 (26.9) 7/134 (5.2),.01 P Valueb N/A.25 N/A N/A Data are presented as No. (%), unless otherwise indicated. FM 5 femoral; N/A 5 not available; SC 5 subclavian. See Table 1 legend for expansion of the other abbreviation. a Comparison of catheter types within each study period. b Comparison of preeducation and posteducation within each catheter type. reduced overall mechanical complication rates, including placement failure and arterial puncture rates, for CVC placement. The number of percutaneous punctures per CVC placement was significantly lower in the posteducation period, supporting our primary hypothesis that a prerotational, simulation-based workshop significantly improves the safety of CVC insertion. Previous studies showed a decreased catheter-related infection rate after simulation-based training. 12,13 A recent study also showed that the simulation-based training reduced mechanical complications. 11 The present program was unique in its relatively short training time (1 h 45 min), effective trainee-to-faculty ratio (2:1), detailed instruction on ultrasonography usage, and image interpretation. Multivariate analysis demonstrated that interns were more likely to cause mechanical complications than fellows and attending physicians in the preeducation period. However, in the posteducation period, multivariate analysis showed no difference in overall mechanical complication rate among the three different levels of operators. The bivariate analysis demonstrated that the overall complication rate in CVC placements performed by interns significantly decreased in the posteducation period compared to the preeducation period. To our knowledge, this report is the first to demonstrate that a simulation-based workshop significantly enhances the safety of CVC placements by relatively inexperienced operators. In addition to the reduction in mechanical complication rates, we observed notable changes in the type of CVC placement. Selection of CVC insertion site and ultrasonography usage were at the operator s discretion. After the workshop, ultrasonography usage increased significantly (from 3.0% to 61.4%), and the percentage of landmark subclavian and femoral CVC placements decreased significantly. This decrease coincided with an increase in ultrasound-guided femoral and IJ CVC placements. These changes may be explained by the following hypotheses. First, competency in ultrasonography technique is relatively easy for house staff to acquire. Second, the ultrasound machine is readily available for use 24 h/d. Third, the increased number of emergent procedures (from 27.3% to 44.8%) plays a role. When the IJ site was not available during an emergency (ie, intubation, the patient was sitting upright for dyspnea, etc), the femoral site may have been preferred. In fact, the most common emergent procedures in the posteducation period were ultrasound-guided femoral and IJ CVC placements (n 5 91, 50.6%, and n 5 32, 17.8%, respectively) (e-table 2). Among elective procedures, the proportion of femoral procedures significantly decreased from 28.4% to 20.3% ( P 5.04) (e-tables 1, 2). We believe that our workshop changed physician behavior with regard to choice Figure 2. Mechanical complications by use of ultrasonography in the two study periods. In the preeducation period, ultrasound guidance did not show a decrease in overall complication (10.0% vs 33.6%, P 5.17) or placement failure (0.0% vs 23.5%, P 5.12) rates. In the posteducation period, the use of ultrasonography significantly decreased the overall com plication (12.2% vs 40.0%, P,.01) and placement failure (6.5% vs 31.6%, P,.01) rates. Land 5 landmark; US 5 ultrasonography. See Figure 1 legend for expansion of other abbreviation. 656 Original Research

6 Table 3 Multivariate Analysis for Overall Mechanical Complications Predictors OR 95% CI P Value Preeducation Male vs female sex Level of operators.07 Interns vs fellows and attending physicians Residents vs fellows and attending physicians Level of supervisors.74 No supervisors vs fellows and attending physicians Residents vs fellows and attending physicians RSDU vs MICU Catheter insertion type.02 Landmark IJ vs landmark femoral Ultrasound IJ vs landmark femoral Landmark subclavian vs landmark femoral Posteducation Day vs night Emergent vs elective Level of operators.09 Interns vs fellows and attending physicians Residents vs fellows and attending physicians Level of supervisors.26 No supervisors vs fellows and attending physicians Residents vs fellows and attending physicians Catheter insertion type,.01 Landmark IJ vs ultrasound femoral ,.01 Ultrasound IJ vs ultrasound femoral Landmark subclavian vs ultrasound femoral ,.01 Ultrasound subclavian vs ultrasound femoral Landmark femoral vs ultrasound femoral ,.01 RSDU 5 respiratory step-down unit. See Table 1 legend for expansion of other abbreviations. of CVC site and use of ultrasound guidance, significantly affecting mechanical complication rates. Another notable change was the practice of CVC insertion in the posteducation period. The number of procedures performed by interns decreased, and more procedures were performed by fellows and attending physicians. Interestingly, in the landmark procedures, there was no difference in the proportion of CVC placements performed by fellows and attending physicians between the two study periods (preeducation, 17.3%; posteducation, 20.7%; P 5.37) (e-tables 1, 2). This finding can be explained by our new institutional policy that recommends that house staff and fellows perform at least 10 supervised ultrasound-guided procedures for credentialing. Because ultrasound-guided CVC placement was officially introduced in our educational workshop, more senior residents and fellows chose ultrasoundguided CVC placement. In fact, the proportion of the ultrasound-guided femoral and IJ CVC placements performed by the group of fellows and attending physicians was high (41.0% and 32.1%, respectively) in the posteducation period (e-table 2). The present study has several limitations. Our database did not include previous surgeries, radiotherapy, or catheter attempts, which may be associ- ated with an increased mechanical complication rate. 3 When investigators or supervisors were not present at the time of CVC placement, complications may have been underreported. Our data set was not powered to detect a difference in pneumothorax, which can be seen in the subclavian approach in 2% to 3% of attempts. 14 Recent institutional changes on credentialing ultrasound-guided CVC placements affected the operator and supervisor characteristics significantly compared with the preeducation period. Finally, the CVC infection rate was collected from institutional data, and complete data were available only for the posteducation period. In conclusion, our prerotational, simulation-based workshop improved the safety of CVC insertion, especially for placements by relatively inexperienced operators. We also showed that using ultrasonography for CVC placement is significantly safer than the landmark technique. We suggest that simulationbased training should be conducted prior to starting ICU rotations if house staff will be primarily responsible for CVC placement. We believe that our training model is reproducible and provides a system for teaching centers to bring inexperienced house staff to an appropriate level of competence for CVC placement using an efficient multimedia hands-on approach. CHEST / 140 / 3 / SEPTEMBER,

7 Acknowledgments Author contributions: Dr Sekiguchi had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Dr Sekiguchi: contributed to the concept, training of house staff, data collection, statistical analysis, primary authorship, composition of manuscript, and review of the final manuscript prior to submission. Dr Tokita: contributed to the training of house staff, data collec tion, crucial review, editing of the manuscript, and review of the final manuscript prior to submission. Dr Minami: contributed to the training of house staff, data collection, crucial review, editing of the manuscript, and review of the final manuscript prior to submission. Dr Eisen: contributed to the data collection, statistical analysis, crucial review, editing of the manuscript, and review of the final manuscript prior to submission. Dr Mayo: contributed to the crucial review and editing of the manuscript and review of the final manuscript prior to submission. Dr Narasimhan: contributed to the concept, data collection, crucial review, editing of the manuscript, and review of the final manuscript prior to submission. Financial/nonfinancial disclosures: The authors have reported to CHEST that no potential conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article. Other contributions: The study was performed at Beth Israel Medical Center, New York, NY. Additional information: The e-appendix and e-tables can be found in the Online Supplement at org/content/140 /3/652/suppl/DC1. References 1. Merrer J, De Jonghe B, Golliot F, et al ; French Catheter Study Group in Intensive Care. Complications of femoral and subclavian venous catheterization in critically ill patients: a randomized controlled trial. JAMA ;286(6): McGee DC, Gould MK. Preventing complications of central venous catheterization. N Engl J Med ;348(12): Mansfield PF, Hohn DC, Fornage BD, Gregurich MA, Ota DM. Complications and failures of subclavian-vein catheterization. N Engl J Med ;331(26): Sznajder JI, Zveibil FR, Bitterman H, Weiner P, Bursztein S. Central vein catheterization. Failure and complication rates by three percutaneous approaches. Arch Intern Med ; 146 (2 ): Fares LG II, Block PH, Feldman SD. Improved house staff results with subclavian cannulation. Am Surg ; 52 ( 2 ): Karakitsos D, Labropoulos N, De Groot E, et al. Real-time ultrasound-guided catheterisation of the internal jugular vein: a prospective comparison with the landmark technique in critical care patients. Crit Care ; 10 ( 6 ): R ccforum.com/content/10/6/r162. Accessed October 1, Hind D, Calvert N, McWilliams R, et al. Ultrasonic locating devices for central venous cannulation: meta-analysis. BMJ ;327 (7411 ): Randolph AG, Cook DJ, Gonzales CA, Pribble CG. Ultrasound guidance for placement of central venous catheters: a meta-analysis of the literature. Crit Care Med ;24 (12 ): Denys BG, Uretsky BF, Reddy PS. Ultrasound-assisted cannulation of the internal jugular vein. A prospective comparison to the external landmark-guided technique. Circulation ;87 (5 ): Mallory DL, McGee WT, Shawker TH, et al. Ultrasound guidance improves the success rate of internal jugular vein cannu la tion. A prospective, randomized trial. Chest ; 98 (1 ): Barsuk JH, McGaghie WC, Cohen ER, O Leary KJ, Wayne DB. Simulation-based mastery learning reduces complications during central venous catheter insertion in a medical intensive care unit. Crit Care Med ;37 (10 ): Khouli H, Jahnes K, Shapiro J, et al. Performance of medical residents in sterile techniques during central vein catheterization: randomized trial of efficacy of simulation-based training. Chest ;139 (1 ): Barsuk JH, Cohen ER, Feinglass J, McGaghie WC, Wayne DB. Use of simulation-based education to reduce catheter-related blood stream infections. Arch Intern Med ; 169 ( 15 ): Eisen LA, Narasimhan M, Berger JS, Mayo PH, Rosen MJ, Schneider RF. Mechanical complications of central venous catheters. J Intensive Care Med ;21 (1 ): Original Research