18.10 Novel Electromechanical Simulator for Training in Cardiopulmonary Bypass

Posted on December 22, 2014

J. Fernandez1, P. Frank1, J. Resnik2, P. Benharash1  1David Geffen School Of Medicine, University Of California At Los Angeles,Division Of Cardiothoracic Surgery,Los Angeles, CA, USA 2David Geffen School Of Medicine, University Of California At Los Angeles,Department Of Anesthesiology And Perioperative Medicine,Los Angeles, CA, USA

Introduction:  Training of physicians and healthcare professionals in the management of cardiopulmonary bypass (CPB) poses major challenges. These include the complexity of the cardiovascular system as well as the rarity of catastrophic events. In light of these challenges, a widely available and realistic simulator for bypass is needed and is yet lacking. We aimed to design an artificial, computer-controlled CPB simulator that would generate the responses of the human body to various hemodynamic states. Further, we hoped that this CPB simulator could be coupled to the CPB machine that is typically used by cardiac perfusionists during cardiac surgery. 

Methods:  The CPB simulator was driven by computer control and feedback. LabVIEW 2013 software (National Instruments, Austin TX) was used to design a graphic user interface similar to a patient monitor. The computer also controlled a pump via a customizable pulse-width modulated waveform to simulate cardiac ejection of arbitrary strength and duration. A series of control valves were placed in the circuit to simulate the systemic vascular resistance. 

Results: A computer algorithm was able to generate normal and abnormal scenarios that commonly present in the emergency department or operating room. The CPB simulator uses an algorithm for increasing and/or reducing the cardiac output of the system based on heart rate and preload without user input. The system generated various conditions, including hypovolemic, cardiogenic and distributive shocks. The CPB simulator was able to respond to virtual administration of chronotropes, inotropes, vasoconstrictors, and vasodilators in real time. The system was successfully connected to a clinical bypass machine and initiation and weaning of bypass was performed. The simulator was found to be realistic by a cardiac surgeon and a perfusionist.  

Conclusion: We have demonstrated the feasibility of a simulator that incorporates computer algorithms, displays, and actual mechanical fluid dynamics to provide a suitable platform for cardiopulmonary bypass practice. Given its simplicity and low cost, this CBP simulator should be used in individual and team scenarios. Routine practice may potentially be lifesaving in cases of pump emergencies such as power failure and air embolism. The simulator also provides an ideal setup for team training that would include members of surgical, anesthesia, and perfusion training programs.

18.11 Low-Cost Cardiothoracic Skills Simulator for Mentored Remote Practice

Posted on December 22, 2014

A. J. Doud1, E. Roselli2, G. Loor1  1University Of Minnesota,Cadiothoracic Surgery,Minneapolis, MN, USA 2Cleveland Clinic,Cleveland, OH, USA

Introduction:  Traditional methods of passing surgical expertise from mentor to student have relied on observation and graduated repetition of core component tasks, built together gradually into a broader surgical competency. While simulation in medical training has emerged as a supplemental alternative means of polishing surgical skill, ultimately time spent in the operating room with a surgical mentor is what knits together these skills into a meaningful body of applicable knowledge. However, direct contact with a surgical mentor is limited by availability, caseload and the educational needs of other students. Ideally, surgical mentorship would focus on the critical elements of a student’s deficits, offer timely feedback for improvement, and be readily available in a student’s down-time. Here we present an early-concept system for mentored surgical practice, which may be assembled with 3D-printable and off-the-shelf components.  The system may be supported by a web-based video exchange system, which allows students to upload video of their surgical training from the home or training lab, receive critique from surgical mentors reviewing their videos remotely, and monitor their improvement over time. 

Methods:  Tasks included in the trainer are focused on cardiothoracic surgical practice.  Stations present in the trainer include coronary anastomosis, mitral valve replacement, aortic valve repair, aortic grafting and cannulating. Dimensions of obstacle and barrier pieces were modified from prototype models produced from averaged CT anatomical data for difficult cases in each sub-competency of surgical practice. Vascular simulation material was constructed from a variety of fabrics selected for realism of feel when suturing.  Attention was paid to approximating the spatial and tactile constraints of working within the mediastinum.

Results: All components of the cardiothoracic trainer were fabricated from either off-the-shelf components or 3D-printable components. These components were printed in PLA plastic using a Makerbot Replicator 2 printer. The machine was able to preserve the geometry of all components and produce a set of pieces needed for a trainer in under 10 hours of unsupervised machine time.

Conclusion: Previous prototypes of the surgical trainer required prohibitively costly large-scale fabrication methods to produce the parts needed for the trainer, which in its primary iteration obviated widespread resident use.  We took a new approach in attempting to replicate the functionality of the prior prototype, while exchanging “off-the-shelf”, or small volume 3D-printed components for more costly machined or manufactured parts. Over 50% cost reduction per unit was achievable independent of production volumes . The next round of investigation will focus on in-home use of the system by surgical residents using a novel tele-mentoring system actively under development.
 

18.12 Fabrication of High Fidelity Simulated Breast Skin: A Comparison of Three Silicone Elastomers

Posted on December 22, 2014

K. Maciolek1, A. Terry1, R. Ray1, S. Laufer1, P. Barlow1, C. Pugh1  1University Of Wisconsin School Of Medicine And Public Health,Department Of Surgery,Madison, WI, USA

Introduction:  The clinical breast examination (CBE) poses a unique teaching challenge. Students are uncomfortable and ill prepared to perform the CBE at the end of training. Simulation technology provides a readily available learning platform for training CBE skills. However, current commercial simulators have a homogenous appearance and consistency that lack realism crucial for trainees to achieve competency. The study aim was to assess the visual and tactile realism of three silicone breast skins for improvement of the skin currently used for one CBE simulator.

Methods:  Study participants (N=42) included a convenience sample of clinicians at an academic training program: nurses (N=23), medical students (N=9), pharmacy staff (N=6) and physicians (N=3). Participants performed CBEs on three breast models prepared from the same mold but using different silicone materials (Smooth-On, Inc., Easton, PA). Material A was a single layer of Soma Foma 15, material B was a single layer of EcoFlex 30, and material C was three layers, two made with Dragon Skin 10 and a middle layer containing 200% Slacker. On day one, the breast skins were filled with a hard breast positive, and on day two a soft silicone mastectomy insert (Nearly Me Technologies, LLC, Waco, TX). After performing the CBE, participants completed a survey to assess the realism of the shape, softness and elasticity of the breast and nipple (4-point Likert scale: 0 = Not at all realistic; 3= Highly realistic) and ranking the models on visual appearance, tactile characteristics and overall realism. Repeated measures ANOVA tested if the samples differed in realism. A Chi-Square was performed on the rank questions of visual, realistic feel and overall preference.

Results: All forty-two participants (90.4% female) performed CBEs on all three of the different materials. Soma Foma was rated as much more realistic than EcoFlex and Dragon skin on all measures (shape, softness and elasticity) with the exception of the shape of the nipple, p>.05 (Table 1). Soma Foma was ranked higher on visual appearance, tactile characteristics and overall realism compared to EcoFlex and Dragon skin (p<.001). These results were independent of the breast insert that was placed under the skin (p>0.05).

Conclusion: This study was successful at collecting useful discriminating information about different silicone materials used to fabricate CBE simulator skins. Overall, participants preferred the skin comprised of Soma Foma. Review of participant comments suggests that Soma Foma’s superiority may result from the idealistic youthful appearance and overall heat retention at room temperature. Further work is needed to assess the role of how psychosocial factors effect evaluations of simulator realism.
 

18.13 Commercial Video Camera Use in the Operating Room: GoPro HERO3+, Contour+2, and TASER AXON Flex

Posted on December 22, 2014

S. N. Graves1, S. Davidson1, A. Langerman2  1University Of Chicago,Pritzker School Of Medicine,Chicago, IL, USA 2University Of Chicago,Department of Surgery,Chicago, IL, USA

Introduction:
Video recording in the operating room acts as a record of operative performance, provides useful data for quality improvement, and serves as a platform for sharing surgical technique and knowledge. There are many challenges to collecting video in the operating room to include user interface, physical obstruction of field of view, and contrasting light conditions. Commercially available, off-the-shelf camera systems can facilitate surgical education and record keeping in the operating room by providing a low cost solution for high-definition point-of-view video capture.

Methods:
The GoPro HERO3+ Black Edition, Contour+2, and TASER AXON Flex cameras were field tested for suitability of use in the operating room. Each camera was tested in a clinical and non-clinical setting and evaluated for quality of video and ease of operation. Non-clinical testing consisted of app interface and camera adjustments to ensure comfort, stability, and video quality. Clinical testing consisted of recording surgical procedures, further adjustment of camera settings to compensate for the live operating room environment, and optimizing video collection workflow.

Results:
Head mounting each camera using the respective manufacturer’s head mounting devices accomplished the best point-of-view video collection. The TASER AXON Flex camera was the most comfortable camera, but camera settings could not be adjusted to achieve high quality video capture. The GoPro and Contour cameras produced the best, usable footage in terms of resolution and ability to compensate for the contrasting light environment surrounding the surgical field, with the GoPro producing the best overall footage of surgical procedures. The Contour+2 is only capable of wide-angle field of view and did not adequately capture minute procedures performed during surgical procedures. The optimal video settings for the GoPro HERO3+ Black Edition were found to be a resolution of 1080p with the Protune function enabled, field of view set to ‘narrow’, and white balance set to automatic. Optimal settings for the Contour+2 were determined to be a resolution of 1080p with high video quality, automatic white balance, and contrast, sharpness, and exposure set to 62, 3, and -4, respectively. 

Conclusion:
Commercially available off-the-shelf cameras are a viable option for video collection in the operating room. Cameras must be capable of compensating for contrasting light conditions, provide adequate data security, and be easily operated in order to ensure optimal video capture. The GoPro HERO3+ and Contour+2 cameras are technically sophisticated enough to capture high definition footage in the operating room environment, while the GoPro HERO3+ produced the highest quality video capture of the cameras tested during surgical procedures.
 

18.14 Cost-Effective Vessel Ligation Simulator Develops Transferrable Surgical Skills

Posted on December 22, 2014

Y. Hu1, I. A. Le1, R. N. Goodrich1, K. D. Brooks1, B. L. Edwards1, J. R. Gillen1, A. T. Schroen1, P. W. Smith1, S. K. Rasmussen1  1University Of Virginia,Surgery,Charlottesville, VA, USA

Introduction: Simulation training facilitates operative teaching while safeguarding patient safety. However, a cost-effective model for open vessel ligation does not exist. We hypothesized that a durable vessel ligation simulator that effectively evaluates and instructs junior trainees can be constructed and implemented for less than 100 USD.

Methods: VesselBox was designed to simulate vessel ligation using expired surgical gloves as surrogate vessels. Construction cost was 30 USD; flexible cost was 0.17 USD per ligation. Model construct validity was previously demonstrated through assessments of students, residents, and faculty. To assess model effectiveness, graduating fourth-year medical students were video-recorded while performing ligations using VesselBox. These pre-test recordings were evaluated by three blinded surgical faculty using the Objective Structured Assessments of Technical Skills global rating scale (GRS) and a task-specific checklist. Subsequently, each student was trained using VesselBox in an adaptive practice session tailored to learning speed through cumulative sum. Lastly, post-tests were performed on fresh human cadavers and assessed with blinding toward pre-test results. Inter-rater correlation was assessed by Spearman’s correlation coefficient, and changes in proficiency were measured using the signed rank sum test.

Results: Among 16 students, pre-test scores averaged 2.29 out of 5 for the GRS (interquartile range, IQR 2.01-2.5) and 4.83 out of 8 for the task-specific checklist (IQR 4.17-6.33). Pre-test inter-rater correlation for the two metrics were 0.92 (p < 0.001) and 0.67 (p = 0.025), respectively. VesselBox practice sessions averaged 21.8 min per participant (IQR 19.5 – 27.7), and consisted of between 8 and 13 ligation repetitions. Average training cost was 1.87 USD per participant. Post-tests demonstrated skill transferability and improvement, as measured by both GRS (3.23 vs 2.29, p < 0.001, Figure 1A) and checklist metrics (7.33 vs 4.83, p < 0.001, Figure 1B). Median speed improved from 128.2s to 97.5s per ligation (p = 0.001).

Conclusions: VesselBox is an innovative, cost-effective model for open vessel ligation. Ideally suited for graduating medical students and junior surgical residents, VesselBox has proven utility in both evaluative and practice settings. In the interest of patient safety, simulators for open surgical skills should be used to certify baseline proficiency in a pre-clinical setting.

18.15 Unidentified Retained Foreign Object Simulation: a training for residents on root cause analysis

Posted on December 22, 2014

N. Young1, D. Patterson1, G. A. Merica1, R. Grim1, T. Bell1, V. Ahuja1  1York Hospital,Surgery,York, PA, USA

Introduction:  Gawande et al. (2003 NEJM; Risk Factors for Retained Instruments and Sponges after Surgery) concluded that intra-operative high blood loss, unplanned change in operation, and multi-operative teams increases the risk of an unidentified retained foreign object (URFO). To this end, our community hospital’s URFO policy was revised to mandate a radiologic film in addition to counts. A simulation was created to educate trainees on systematic approaches to preventing URFO in the operating room (OR) and utilizing root cause analysis to learn from sentinel events.

Methods:  A multi-disciplinary team including residents and OR staff, designed a scenario that would simulate an emergency that is susceptible to URFO error-a laparoscopic procedure that is emergently converted to open due to trocar placement arterial injury. The scenario had massive blood loss requiring mass transfusion protocol with multiple operative teams. Purposeful environment of poor communication and team work with high noise level was created. The team joined the residents in the lecture hall who watched the proceedings via videoconference. The Patient Safety Officer (PSO) gave a report that 5 days after abdominal closure a retained sponge was found on abdominal x-ray. The group was led through a root cause analysis process by the PSO. An Event Flow Diagram was described followed by an Event Causation Diagram (Fig. 1).

Results: Using a survey, the majority of the trainees recognized URFO as the sentinel event during the simulation, 11/14 (78.6%) (p = .057). 78.6% of trainees indicated the sentinel event was due to “Failure of Crew Resource Management Performance,” and indicated “Performing Radiologic Films” could prevent or reduce the number of these events. However, 100% of the trainees indicated that team work was not an issue. Out of a maximum of 5 points, trainees rated “overall communication” a mean of 4.1. Communication from anesthesia to team had the lowest score (M=2.9), followed by surgeon to nurse (M=4.3) and nurse to surgeon (M=4.4). After the scenario, our focus was root cause analysis. 78.6% of the trainees correctly indicated that “finding the individual responsible for the mistake” was not the goal, but understanding the primary focus of the root cause analysis of this scenario was “system design.”

Conclusion: The trainees were given the opportunity to learn the concept of root cause analysis and identify factors that lead to URFO. Communication and adherence to policy and procedures is essential in the operating room to reduce sentinel events. Simulation is a methodology that needs to be explored further to develop a framework for patient safety curriculum in graduate medical education. 

18.16 Residents’ Perception of Skill Reduction during Dedicated Research Time

Posted on December 22, 2014

R. D. Ray1, P. B. Barlow1, A. D. D’Angelo1, C. M. Pugh1  1University Of Wisconsin,Department Of Surgery, School Of Medicine And Public Health,Madison, WI, USA

Introduction: Surgery residents may take years away from clinical responsibilities for dedicated research time. Evaluating the potential reduction in surgical skill and knowledge is critical to ensuring competent, independent performance at the end of training. As part of a longitudinal project, the study aim was to investigate residents’ perceptions of clinical skill reduction during dedicated research time. Our hypothesis was that residents would perceive a greater potential reduction in skill during research time for procedures they were less confident in performing prior to entering the lab.

Methods: Surgical residents (PGY 2-3) from Midwestern training programs participated in four simulated clinical procedures (Laparoscopic ventral hernia (LVH) repair, Bowel anastomosis, Subclavian central line insertion, and Urinary catheterization). Prior to performing the simulated procedures, participants rated procedures in terms of difficulty (1=Not difficult to 5=Extremely difficult) and confidence (1=Not confident to 5=Extremely confident). They also rated what reduction, if any, they believed their time in the lab would have on their ability to perform the four procedures (1=No reduction to 5=Very large reduction). Analysis of variance tested differences in ratings among surgical tasks, and Pearson correlations calculated the relationship between months spent in the lab, confidence in performing the procedures, and estimated skill reduction.

Results:Twenty-five residents (60% female) completed the four clinical simulators. Residents had completed between 0-36 months in a lab (M=9.5 months, SD=10.3). Table 1 shows the pairwise comparisons of participants’ average perceived difficulty, confidence, and skills reduction. As the number of months in the lab increased, Confidence ratings for the bowel anastomosis and urinary catheterization decreased (p<0.05). Also, aside from LVH (r=.010), residents' perceived skill reduction was significantly, negatively related to their confidence on the surgical tasks for Bowel anastomosis (r=-.489, p<.005); Central line (r=-.688, p<.001); and Urinary catheter (r=-.531, p<.005).

Conclusion:Residents who were less confident in performing the bowel anastomosis, subclavian central line insertion and urinary catheterization perceived a greater skills reduction during research time. This relationship was not true for the LVH repair, which may result from participants’ not considering themselves to have sufficient baseline skill to accurately predict any future reduction. Future work is required to understand how resident’s perception of skills decay relates to actual changes in performance during dedicated research time and confidence when re-entering clinical practice.

18.17 Approaching surgery simulation education from a patient-centric pathway

Posted on December 22, 2014

K. W. Miyasaka1, R. Aggarwal2  1University Of Pennsylvania,Department Of Surgery,Philadelphia, PA, USA 2McGill University,Department Of Surgery,Montreal, QC, Canada

Introduction: Approaches to simulation education in surgery have traditionally focused on technical skills, recreating isolated aspects of operative or procedural situations. However, clinical practice is a patient-centered continuum of care, and competence consists of a blend of both technical and non-technical skills. In order to deliver an educational experience that is realistic and relevant to clinical practice, we implemented a curriculum built around “pathways” – simulated sequences of preoperative, intraoperative, and postoperative encounters – for first-year general surgery residents.

Methods: The pathway simulation begins with residents seeing a standardized patient (SP) in clinic, performing a preoperative evaluation and consent. They proceed to an operating room containing a procedure-specific porcine or synthetic model. A confederate assistant and anesthesiologist are also present in this fully immersive operative simulation. The same SP reprises their role for the final postoperative encounter in a simulated ward or clinic setting.

Residents are observed and evaluated by attending faculty watching via video in a separate room. Evaluations utilize rating scales endorsed by the American Board of Surgery (Pre-op: CAMEO, Intra-op: OPRS, Post-op: Mini-CEX). Both faculty and the SP provide feedback to each resident at the conclusion of the pathway.

We developed a curriculum that was implemented as a series of three-day educational modules to complement these pathway simulations for four surgical divisions at our institution (biliary, foregut, colorectal, and acute care).

Results: Repeating the pathway on the first and final day of each module enabled serial assessment for efficacy of the educational intervention in addition to residents’ level of achievement. Furthermore, the modular nature of the curriculum allowed for multiple iterations, accommodating all 16 first-year surgery residents as small groups over the course of an academic year without undue disruption to clinical services.

Conclusion: Adopting a patient-centric approach to surgery simulation, we recreated a complete sequence of patient care encounters in a realistic clinical setting, encompassing a pathway of care. Relevant training as well as objective evaluation of clinical competence could be performed in a time-efficient manner. The role of a dedicated simulation program with leadership resources to provide structure and oversight to participating faculty and residents was critical in the success and sustainability of the curriculum, which is now in its second year and in the process of being deployed at a second institution.

18.18 Robotic Simulator Curriculum Validation Study

Posted on December 22, 2014

J. L. Miller1, S. M. Novak1, D. L. Bartlett1, A. H. Zureikat1, H. J. Zeh1, M. E. Hogg1  1University Of Pittsburgh,Surgical Oncology/Surgery/Medicine,Pittsburgh, PA, USA

Introduction:

Robotic surgery is becoming widely used by general surgery and its specialties such as surgical oncology. Our institution has performed over 1,000 robotic surgical oncology cases and has identified the learning curve for several complex robotic resections. Critical to successful dissemination of the platform is a better understanding of how new surgeons learn the technology. This study aims to evaluate content and predictive validity of robotic simulation in surgery.  We hypothesize mastery-based simulation is a valid tool to train residents and fellows toward operative proficiency.

Methods:

A mastery-based simulation curriculum was performed in a virtual reality environment. Mastery was defined as 90% proficiency on each module. A pretest/posttest experimental design utilized virtual reality M score tasks (Match Board 3, Ring & Rail 2, Tubes and Continuous Suture) and inanimate environments using video analysis (Ring Rollercoaster 4, Around the World and Interrupted Suture) to evaluate technical improvement. Prior robotic training experience and curriculum assessment was self-reported in a survey; fellows were asked to rate modules based on difficulty and utility using a Likert scale of 5 (5 being greatest).

Results:

11 fellows enrolled in the curriculum. Prior robotic simulator experience showed: min=0, max=31, median=0.375 and mode=0 hours. Prior robotic case experience demonstrated: min=0, max=50, median=12 and mode=4 cases. 9 fellows (82%) completed the mandatory curriculum. 7 fellows (77.8%) achieved mastery on all 24 modules (one deficient on 4, one deficient on 9). Individual test scores improved; overall time and errors decreased (Table). Of the 24 modules, frequency to mastery demonstrated: min=1, max=17, median=2. Simulator hours spent completing curriculum showed: min=1.1, max=6.6 and med=4.2. 9 (100%) fellows continued modules beyond mastery. Fellows rated modules between 1 and 5 for difficulty and 3 and 5 for utility. Needle driving and Endowrist 2 modules were perceived as most difficult; needle driving modules were most useful. 8 (89%) fellows perceived improvement in robotic skills after completing the curriculum.

Conclusion:

This pilot study is limited by sample size; however, these preliminary results show overall score improvement, decrease in errors and decrease in total time. Time to complete the curriculum is manageable.To increase power for statistical comparison, the study is ongoing to include incoming fellows, senior general surgery residents and other fellowship programs. Ultimately, the study will assess correlation between performance on simulator curriculum with inanimate biotissue curriculum and operative improvement to assess content and predictive validity.

18.19 Prehospital Trauma Care Education for First Responders in Western Rajasthan

Posted on December 22, 2014

A. Aekka2, M. V. Hollis2, E. M. Boudiab2, G. P. Laput2, H. Purohit3, A. K. Vyas2,4, D. Vyas1,2  1Michigan State University,Department Of Surgery,Lansing, MI, USA 2Michigan State University,College Of Human Medicine,Lansing, MI, USA 3Arogyaa.com,Meerut, UP, India 4Michigan State University,Department Of Pediatrics,Lansing, MI, USA

Introduction:

The burden of trauma and injury is particularly devastating for developing nations such as India. The crux of the problem lies in the astonishing lack of prehospital trauma services: 80% of trauma victims in India cannot access medical care within the first hour. Existing health education initiatives fail to engage first responders and neglect the local context. 

Methods:
A 2-day hands-on prehospital trauma management training program with video lectures was developed for first responders in the local language. The course consists of 10 interactive sessions dealing with essential prehospital trauma care concepts, such as airway establishment, hemorrhage control, CPR, fracture stabilization, triage, and communication. Extensive self-learning videos, which help to overcome the language barrier, and high-fidelity simulation, which presents the most realistic training experience, provide a level of engagement that traditional didactic methods cannot offer. Video-debriefing serves as a valuable evaluation method. First responders are further introduced to advanced tools, such as the King LT airway and pulse oximeter, but are instructed in improvised management as resources are limited on the field. A comprehensive, but concise manual and specialized tool kit were also developed for trainees.
 
48 participants from Jodhpur, Rajasthan, including police officers, firemen, ambulance and taxi drivers, EMTs, hospital staff, and nursing staff and students, attended the pilot course. 18 instructors were recruited from Jodhpur and included medical students/residents and faculty, private practice physicians, and police officials. These individuals were trained in a 4-hour session prior to the program regarding the course content, materials, expectations, methods of engagement, and principles, such as collaborative learning, positive reinforcement, and the use of native, lay-person language.
 

Results:
Pre- and post-training surveys were used to evaluate participants' competence in managing 10 prehospital trauma matters. Statistically significant increases in competence were demonstrated for all topics: Airway (35.0%), Hemorrhage (36.1%), Fractures (32.0%), Cervical Spine Injury (45.4%), Chest Injury (41.3%), IV Line Placement (29.9%), Extrication (18.6%), Scene Assessment (35.1%), Triage (26.5%), and Communication (25.4%). The greatest increases were observed in cervical spine and chest injury management. The lowest, but still significant, increases were observed in extrication and communication. A six-month post-training survey will be conducted.

Conclusion:
First responder training in the native language with simulation and video-debriefing improves understanding and skills in all essential aspects of prehospital care, however, results suggest that discussion of extrication and communication should be strengthened. The goal is to develop a program that engages and prepares first responders in the procedural, cognitive, and affective aspects of prehospital trauma management.

18.20 Novel Simulation Course for Application of Resuscitative Endovascular Balloon Occlusion of Aorta

Posted on December 22, 2014

R. A. Lawless1, J. D. Love1  1University Of Texas Health Science Center At Houston,Acute Care Surgery,Houston, TX, USA

Introduction:
Noncompressible truncal hemorrhage (NCTH) remains a major cause of death from traumatic injuries both in military and civilian populations.  Recent interest in resuscitative endovascular balloon occlusion of the aorta (REBOA) has shown promise as an alternative to resuscitative thoracotomy and adjunct to preperitoneal packing for NCTH.  General surgery residents and fellows are exposed to this technique in the elective and emergent setting during abdominal aortic aneurysm repair.  However, this skill is not always maintained in surgeons not planning on completing a vascular surgery fellowship.  The recent interest in REBOA has brought a need for training in this technique for both new trainees and current trauma attendings.

Methods:
Between the University of Texas Health Science Center and Texas Medical Center the Advanced Surgical Skills for Exposure in Trauma (ASSET) is held four times per year.  The curriculum is followed as determined by the ACS Committee on Trauma.  Following course completion, a didactic session, anatomic demonstration, and simulation session for the REBOA was introduced.

Results:
Participants included PGY-3, PGY-4, PGY-5, PGY-6, general surgery faculty, and trauma/critical care faculty.  From June 2012 to March 2014, 64 participants completed the simulation course.  Following the course a number of participants have utilized the technique learned from the course in the care of their trauma patients.  

Conclusion:
The REBOA technology can be taught to trainees and surgeons in practice and immediately applied in the trauma setting.  Course participants are confident in this technique of minimally invasive control of NCTH.  Future and ongoing plans for the course include a pre-test, post-test, self-assessment, and course survey.

16.17 30-day Unplanned Readmission After Lower Extremity Bypass: Is Diabetes An Independent Predictor?

Posted on December 22, 2014

A. Najafian1, S. Selvarajah1, E. B. Schneider1, M. B. Malas2, B. Ehlert3, K. C. Orion3, A. H. Haider1, C. J. Abularrage3  1Johns Hopkins University School Of Medicine,Center For Surgical Trials And Outcomes Research, Department Of Surgery,Baltimore, MD, USA 2Johns Hopkins University School Of Medicine,Department Of Surgery,Baltimore, MD, USA 3Johns Hopkins University School Of Medicine,Division Of Vascular Surgery And Endovascular Therapy, Department Of Surgery,Baltimore, MD, USA

Introduction:  Readmission negatively affects reimbursement and increases both morbidity and mortality. Lower extremity bypass (LEB) has been shown to be associated with readmission. This study aimed to compare 30-day unplanned readmission after LEB between diabetics and non-diabetics.

Methods:  Patients undergoing LEB in the 2011-12 ACS-NSQIP database were divided into 3 groups: non-diabetics (non-DM), non-insulin-dependent diabetics (NIDDM) and insulin dependent diabetics (IDDM). Unplanned readmission was compared and multivariate logistic regression was used to evaluate the influence of diabetes status on 30-day readmission. 

Results: A total of 9207 patients (5155 (56%) non-DM, 1690 (18%) NIDDM and 2362 (26%) IDDM) underwent LEB. Unplanned readmission was observed in 1448 (16%) patients. IDDM had significantly higher crude postoperative complication (30% non-DM, 36% NIDDM vs. 43%, P<0.001) and unplanned readmission rates (14% non-DM, 16% NIDDM vs. 20%, P<0.001). Concomitant cardiac disease significantly modified the association between diabetes and unplanned readmission. On multivariable analysis, IDDM was an independent predictor of unplanned readmission in the absence of cardiac disease (OR=1.21; 95% CI [1.01-1.44]; P=0.03). However, this association did not remain significant in the presence of cardiac disease (OR=1.16; 95% CI [0.88-1.53]; P=0.28). Subgroup analysis of IDDM patients revealed that regardless of cardiac status, postoperative complications were the strongest independent predictors of unplanned readmission. In IDDM without concomitant cardiac disease, the only preoperative independent predictors of unplanned readmission were dialysis (OR=1.57; 95% CI [1.08-2.28]; P=0.01) and anemia  (OR=1.51; 95% CI [1.09-2.08]; P=0.01). 

Conclusion: While unplanned readmission after LEB was more common among diabetics, only IDDM in the absence of cardiac disease was independently associated with unplanned readmission. In IDDM patients without cardiac disease, dialysis dependence and anemia significantly increased the likelihood of unplanned readmission. Regardless of cardiac status, postoperative complications were strongly associated with readmission.

 

16.18 Predictors of Surgical Site Infection after Discharge in Patients Undergoing Major Vascular Surgery

Posted on December 22, 2014

M. L. Barnes1, J. T. Wiseman1, S. Saha1, J. Havlena1, S. Fernandes-Taylor1, K. C. Kent1  1University Of Wisconsin School Of Medicine And Public Health,Wisconsin Surgical Outcomes Research Program,Madison, WI, USA

Introduction: Surgical site infection (SSI) after vascular surgery is one of the most common post-operative complications and is the leading cause of unplanned, potentially preventable hospital readmissions among surgical patients. Transitional care interventions to detect SSI after hospital discharge may help stem the burden of readmissions and morbidity associated with SSI; however little is known regarding the classification of patients at high-risk for development of SSI occurring after hospital discharge (postSSI).   

Methods: Patients who underwent major vascular surgery (abdominal aortic aneurysm, open aortoiliac, and lower extremity revascularization procedures) from 2005-2012 were identified from the American College of Surgeons National Surgical Quality Improvement Program Participant Use Files. Patients were categorized as having no SSI, SSI while in-hospital (preSSI), or postSSI. Multivariable logistic regression was performed using patient demographics, preoperative health characteristics and comorbidities, and operative variables to predict preSSI and postSSI.   

Results: Of the 50,091 patients who underwent major vascular surgery, 4,481 (9.0%) were diagnosed with SSI (2.1% preSSI; 6.8% postSSI). Multivariate predictors of both preSSI and postSSI include obesity vs. normal BMI (preSSI odds ratio [OR]:1.6; 95% confidence interval [CI]: 1.3-2.0) (postSSI OR: 2.2; 95% CI: 2.0-2.5) an existing open wound or wound infection (preSSI OR:1.2; 95% CI: 1.0-1.5) (postSSI OR:1.2; 95% CI: 1.1-1.3). Multivariate predictors unique to patients who experienced preSSI include totally dependent functional status (OR: 2.7; 95% CI: 1.8-4.0), emergency case (OR: 2.5, 95% CI: 1.9-3.1), chronic obstructive pulmonary disease (OR:1.4; 95% CI: 1.1-1.8) and prolonged operative time (OR:1.3; 95% CI: 1.0-1.7). Multivariate predictors unique to patients who experienced postSSI include female gender (OR:1.4; 95% CI 1.3-1.5), overweight vs. normal BMI (OR:1.3; 95% CI: 1.2-1.5), insulin dependent and non-dependent diabetes mellitus (OR:1.4; 95% CI: 1.2-1.5) (OR:1.3; 95% CI: 1.1-1.4), dyspnea with moderate exertion (OR:1.1; 95% CI: 1.0-1.3), rest pain/gangrene (OR:1.4; 95% CI: 1.3-1.5), coronary artery disease (OR:1.1; 95% CI: 1.0-1.2), hypertension requiring treatment (OR: 1.2; 95% CI: 1.1-1.4), peripheral vascular disease (OR:1.3; 95% CI: 1.2-1.4), smoking (OR:1.2; 95% CI: 1.1-1.3) and neurological disease (OR:1.1; 95% CI: 1.0-1.3).   

Conclusions: Predictors of preSSI after major vascular surgery are largely acute non-modifiable conditions whereas predictors of postSSI are primarily chronic comorbidities. Appropriate identification of these different sets of risk factors may improve both in-hospital wound surveillance and subsequent transitional care efforts to improve wound monitoring

16.19 Risk factors of Surgical Site Infection after Open Abdominal Aortic Aneurysm Repair

Posted on December 22, 2014

J. T. Wiseman1, S. Fernandes-Taylor1, K. C. Kent1  1University Of Wisconsin,Wisconsin Surgical Outcomes Research Program,Madison, WI, USA

Objective: Surgical site infection (SSI) is a major cause of morbidity after open abdominal aortic aneurysm repair (AAA). Incidence of SSI can lead to ventral hernia, graft infection, hospital readmission, increased costs, and in certain cases mortality. Identification of SSI risk factors at the time of open AAA repair may facilitate selection of patients at high-risk for SSI development, allowing for a higher level of in-hospital wound surveillance and subsequent implementation of appropriate transitional care interventions. The objective of this study is to describe the incidence and risk factors for SSI after open AAA repair using national data.

Methods: Patients who underwent open AAA repair from 2005-2012 were identified from the American College of Surgeons National Surgical Quality Improvement Program Participant Use Files. Multivariable logistic regression was performed using patient demographics, preoperative health characteristics and comorbidities, and operative variables to predict SSI.

Results: Of the 8,192 patients who underwent open AAA repair, 386 (4.3%) were diagnosed with SSI within 30 days of surgery. On multivariate analysis, patient preoperative health characteristics and comorbidities that predicted SSI included obesity vs. normal BMI (odds ratio [OR]: 2.1; 95% confidence interval [CI]: 1.6-2.7), overweight vs. normal BMI (OR: 1.4; 95% CI: 1.0-1.8), an existing open wound or wound infection (OR: 2.2; 95% CI: 1.2-3.9), dyspnea (OR: 1.5; 95% CI: 1.2-1.9), and smoking (OR: 1.3; 95% CI: 1.1-1.7). Operative factors that predicted SSI included emergency operation (OR: 1.8; 95% CI: 1.3-2.4) and operative time >5 hours vs. 3-5 hours (OR: 1.6; 95% CI: 1.3-2.1).

Conclusions: SSI after open AAA repair is a common post-operative complication. Our data suggest that there are readily identifiable patient and operative characteristics known prior to and at completion of open AAA repair that significantly predict increased risk of SSI, suggesting that targeted in-hospital wound surveillance and subsequent transitional care efforts may improve outcomes in this patient population.

 

16.20 Pancreatic lipoma: does it need treatment?

Posted on December 22, 2014

M. T. Fohtung1, K. Sandrasegaran1, N. Zyromski1  1Indiana University School Of Medicine,General Surgery,Indianapolis, IN, USA

Introduction:  Pancreatic lipomas are a rare and benign form of the mesenchymal neoplasms that make up 1%-2% of all pancreatic neoplasms. They are often diagnosed incidentally during radiographic imaging and treatment in the 69 reported cases in the literature has mostly been conservative although there is no consensus on the histopathologic significance of lipomas or whether surgical intervention is warranted. In this study, we describe 74 cases of intrapancreatic lipomas from a single institution over a 12-year period to elucidate the natural history including associated symptoms, the need for intervention and the potential for dedifferentiation to liposarcoma.

Methods:  In the period from January 2001 to December 2013, we selected patients over the age of 18 diagnosed with pancreatic lipoma based on Ultrasound, CT scan and MR imaging. Clinical data was coupled with radiographic images and reviewed. Age at the time of diagnosis, gender, presenting symptoms, location of lipoma in pancreas, size on imaging, surgical intervention and pathological findings were evaluated. 

Results: Pancreatic lipomas were identified in 74 patients including 41 women and 33 men aged 31-88 (median age 64). Most of the neoplasms were located in the head of the pancreas (n=32) followed by the body (n=16), tail (n=16), neck (n=6) and uncinate (n=6). The lipomas ranged in size from 0.1cm to 4.8cm. The majority of patients had no symptoms at presentation (n=36) while others had abdominal pain (n=26), dysphagia (n=1), jaundice (n=1), pelvic pain (n=1), chest pain (n=2), nausea/vomiting (n=2), dyspnea (n=1), constipation (n=1). One lipoma was pathologically confirmed and no patients underwent surgical resection.

Conclusion: Pancreatic lipomas are rare benign mesenchymal neoplasms with increasing incidence as the use of radiographic imaging continues to expand.  Most reported cases are managed conservatively with follow-up imaging especially when the lipoma has well defined margins and causes no obstruction to the pancreatic duct or common bile duct. Given the lack of consensus, questions remain about the histopathological significance concerning dedifferentiation to liposarcoma and whether surgical resection should then be considered as an intervention. This large single center study will help shed more light on these questions.

17.02 Trauma System Regionalization across State Borders

Posted on December 22, 2014

J. J. Sumislawski1, S. A. Savage1, B. L. Zarzaur2  1University Of Tennessee Health Science Center,Memphis, TN, USA 2Indiana University School Of Medicine,Indianapolis, IN, USA

Introduction: Organization of trauma centers into a state trauma system is associated with reduced mortality. Development of a trauma system could affect a neighboring state’s trauma center. In 2010, an inclusive trauma system was implemented that allowed centers in bordering states to participate. This study was designed to examine the effect of the development of this new trauma system on a participating out-of-state Level 1 trauma center.

Methods: Patients referred to a participating out-of-state trauma center were included. Using a difference-in-differences approach, residents of the state with the new trauma system (TSystem patients) were compared with residents of the state with the trauma center (TCenter patients) PRE (2008-2009) and POST (2011-2012) implementation of the trauma system.

Results: TCenter patients decreased 3% PRE versus POST while TSystem patients increased 39% (Table 1). Injury severity did not change for TCenter patients but decreased for TSystem patients PRE (mean Injury Severity Score 15) versus POST (12, p<0.05). Transfers from referring hospitals increased from both states. For TSystem patients, air arrivals decreased and payer status did not change. Compared with TCenter patients, odds of mortality for TSystem patients decreased PRE (OR 0.97; 95% CI 0.72, 1.31) versus POST (OR 0.73; 95% CI 0.53, 0.99). When only those with ISS >10 were analyzed, mortality did not change PRE (OR 0.85; 95% CI 0.60, 1.20) versus POST (OR 0.96; 95% CI 0.67, 1.37). Secondary overtriage increased PRE (OR 0.67; 95% CI 0.54, 0.83) versus POST (OR 1.53; 95% CI 1.32, 1.78) for TSystem patients.

Conclusion: Development of a state trauma system resulted in an increase in less severely injured patients referred to an out-of-state trauma center without change in payer status. For patients of higher injury severity, there was no change in odds of mortality PRE versus POST for either state. Trauma-system implementation promoted overtriage of in-state patients to the out-of-state trauma center. Despite increases in volume and overtriage, allowing participation of a neighboring state’s trauma center did not result in increased mortality for either state’s residents.

 

17.03 Break a Leg Not the Bank: Should We Treat Simple Fractures in Trauma Centers?

Posted on December 22, 2014

F. Gani1, N. Nagarajan1, H. Alshaikh1, C. K. Zogg1, H. Alturki1, S. Selvarajah1, A. Najafian1, L. Kodadeck1, C. G. Velopulos2, D. T. Efron2, E. B. Schneider1, A. H. Haider1  1Johns Hopkins University School Of Medicine,Center For Surgical Trials And Outcomes Research, Department Of Surgery,Baltimore, MD, USA 2Johns Hopkins University School Of Medicine,Department Of Surgery,Baltimore, MD, USA

Introduction:  Healthcare policy efforts are increasingly geared toward providing higher quality care at lower costs. For fractures alone, healthcare spending in the United States is estimated to be >$20 billion per year. Previous data for severely injured patients suggest an improvement in survival at trauma centers. Little is known about possible differences in patient charges for the management of non-life threatening conditions at trauma centers vs. non-trauma center hospitals. Using uncomplicated, closed tibial fractures as an index condition, this study examined possible differences in patient charges at level 1 trauma centers (TC) compared with non-trauma centers (NTC).

Methods:  Data from the 2006-2011 HCUP Nationwide Emergency Department Sample (NEDS) were queried and patients with a primary diagnosis of closed tibial fracture who underwent routine same-day discharge from the emergency department (ED) were identified using ICD-9-CM diagnosis codes. Patients with an Abbreviated Injury Scale (AIS) extremity score of ≤2 were included in the study cohort. Patients with other major concomitant injuries listed in diagnosis positions 2-15 were excluded as were individuals who had a calculated  AIS >0 in any region except “extremity.” Patient demographics, injury and hospital characteristics were compared between TC and NTC using χ² and t-tests. Wilcoxon rank-sum tests examined TC vs. NTC difference in median patient charges. A generalized linear model with a gamma distribution and robust error variances (adjusted for age, sex, insurance status, Charlson comorbidity index, income quartile, mechanism of injury and hospital region) examined differences in mean total charges between patients treated at TC and NTC.

Results: A total of 15,773 patients met inclusion criteria. 1,845 patients were treated at TC and 13,361 patients at NTC. Patients at TC were younger compared to those at NTC, median age 44 (IQR = 32-56) and 48 years (IQR = 38-60) respectively. Proportionally fewer female patients were treated at TC vs. NTC (44.77% vs. 51.21%, p<0.001). Median total charges were higher at TC vs. NTC [$2,278, (IQR $1,259-$4074) vs. $1,351, (IQR $848-$2,313), p<0.001]. Adjusted charges for management in the ED were 94% higher at TC vs. NTC [$3,781 (95%CI $3,548-$4,013) vs. $1,951 (95% CI $1,902-$2,000) p<0.001].

Conclusion: Patients undergoing routine same-day discharge after ED treatment of an uncomplicated tibial fracture at TC incur substantially higher charges than otherwise similar patients treated at NTC. Better understanding the factors underlying the differences in charges observed between TC vs. NTC facilities may enable substantial savings to the healthcare system.

 

17.04 The Relationship Between Blood Alcohol Level and Injury Severity: Is the Floppy Patient Myth True?

Posted on December 22, 2014

C. Valdez1, C. Renne1, M. Radomski1, R. Amdur1, J. Dunne1, B. Sarani1  1George Washington University School Of Medicine And Health Sciences,General Surgery,Washington, DC, USA

Introduction:
The impact of inebriation on severity of injury is unclear.  Conventional teaching suggests increasing blood alcohol level (BAL) may be associated with fewer injuries because the patient is limp due to intoxication at the time of injury. The few studies to date on this topic are limited to a particular mechanism of injury (MOI), injury pattern, or BAL. Therefore we sought to determine the impact of BAL on injury pattern and severity across all MOI. We hypothesize that there is no relationship between BAL and injury severity when controlling for MOI. 

Methods:
Following IRB approval, a retrospective study was performed at an adult trauma center from January 1, 2011 to December 31, 2012. All MOI were included. Injury severity was assessed using the injury severity score (ISS). Chi square and ANOVA were used to examine the relationship between BAC, injury pattern, and ISS within each MOI. Multivariate regression model (MVR) examined the BAC-ISS association adjusting for MOI, gender, and age. 

Results:
Of 1397 patients, the mean age was 44±19, mean ISS was 7.5±6.8, mean BAL was 92±130 mg/dL and 70% were male. Overall mortality rate was 1.3%. Rib fracture (p=0.002) and hemo/pneumothorax (p=0.0009) were negatively associated with BAL, while concussion and laceration had a positive association with BAL (p<0.0001). An increasing BAL had a negative correlation with ISS following fall from standing (p<0.001).  Across all MOI combinde, there was no significant association between BAL and ISS. 

Conclusion:
Inebriated patients have a decreased risk of rib fractures and hemo/pneumothoax, and an increased risk of concussion or laceration.  Increased BAL is associated with a lower ISS in the specific MOI of fall from standing. However, across all MOI, there was no significant association between BAL and ISS when controlling for MOI.  Inebriated patients should be triaged and approached with the same clinical index of suspicion for injury as sober patients. 

17.05 Trends in the treatment of pelvic fractures 2008-2010: Where do we stand?

Posted on December 22, 2014

C. Chu2, L. Tennakoon1, D. Spain1, K. Staudenmayer1  1Stanford University,Surgery,Palo Alto, CA, USA 2University Of South Carolina School Of Medicine,Columbia, SC, USA

Introduction: Bleeding from pelvic fractures can be life-threatening. Treatment for bleeding pelvic fractures involves noninvasive means (pelvic binders) and invasive procedures such as angioembolization (AE) and external fixation (EXFIX).  It is not known how frequently these modalities are used in U.S. trauma centers or whether there have been trends over time.  We hypothesized that there would be an increase in the use of AE and a decrease in the use of EXFIX over time. We also sought to determine if the procedures were associated with a reduction in mortality.

Methods: We used the National Trauma Databank (NTDB) from 2008-2010.   Patients were included in the study if they had the International Classification of Diseases, 9th edition and Clinical Modification (ICD-9-CM) codes for pelvic fractures.  Patients were excluded if they were <18 years, had an isolated acetabular fracture, were not admitted to the hospital, or had an ISS<15.  Only centers that had demonstrated an ability to perform AE or EXFIX were included in the analysis.  The primary outcomes were whether the patient had an AE or EXFIX within the first 24 hours after admission. The secondary outcome was mortality. Univariate analyses and multi-level logistic regression (to control for center effects) were used.

Results: A total of 22,568 met inclusion and exclusion criteria.  Patients were predominantly male (59.6%), white (70.3%), and between the ages of 18 and 44 (50.7%).   Overall, AE and EXFIX were performed in 746 (3.3%) and 663 (2.9%) of patients, respectively.  Patients who received AE and EXFIX were different across all measures in unadjusted analyses, but after adjusting for known confounders, only age, injury severity, physiologic instability, and diagnosis of shock were associated with receiving a procedure.    Over the study period, there was an increase in the use of AE (2.5% in 2007 to 3.7% in 2010, p<0.001), which remained significant in adjusted analysis (OR per year 1.15, p=0.008).  There was no significant trend for EXFIX.  AE and EXFIX were associated with a higher mortality in unadjusted analyses compared to those who did not receive a procedure (11.0% for no procedure vs. 20.5% and 13.4% for AE and EXFIX, respectively; p<0.001).  In adjusted analyses, AE remained associated with higher mortality (OR 1.63, p<0.001), whereas EXFIX was associated with a slightly lower risk (OR 0.95, p<0.01).

Conclusion: The use of AE in severely injured pelvic fracture patients is increasing.  However, this procedure is associated with a higher mortality.  It is possible that AE is used more often in patients at high risk of death, but that its use does not reduce this risk.  We should carefully examine the use of this expensive resource in future studies.

 

17.06 Trauma System Funding is Associated WIth Increased Numbers of Level 3 Trauma Centers

Posted on December 22, 2014

E. Kelly1, E. R. Kiemele2, G. Reznor1, J. M. Havens1, Z. Cooper1, A. Salim1  1Brigham And Women’s Hospital,Boston, MA, USA 2Harvard School Of Medicine,Brookline, MA, USA

Introduction:
Taken as a group, state trauma systems are associated with beneficial effects, such as reduction in mortality, but not all trauma systems are the same. Some states allocate a yearly budget in support of its system's activities, some states do not. It has not been shown that the benefits of a trauma system accrue equally to all states, or if the prescence of  funding leads to beneficial effects not seen in states without a budget. It is also not known whether funding for a trauma system is associated with financial benefits that produce a return on the investment of budgetary funds. The objective of this study was to determine if f states with funded trauma systems are associated with an increase in access to trauma care (as defined as numbers of trauma centers per capita), or cost effectiveness of trauma care (as defined by numbers of trauma centers per Gross Domestic Product) compared to states without trauma system funding in place.

Methods:
A retrospective population based study was performed.  Data for the number of verified trauma centers in 2010 were obtained from the American College of Surgeons (ACS) and for state-designated trauma centers from official reports from state departments of health. Only adult centers were examined. Populations and Gross Domestic Product (GDP) were obtained from the US Census. The main outcome measure was the number of trauma centers per state per population and per GDP. Statistical analysis was carried out using the Chi Square Test and Poisson Regression; p values <.05 were reported as significant.

Results:
There was no statistically significant correlation between the presence of a funded system and the numbers of Level 1 or Level 2 trauma centers. However, there was a statistically significance increase in the number of Level 3 centers in states with funded trauma systems per state GDP and population.  In funded states, the number of Level 3 trauma centers per GDP and state population were 72.5±14.2/$100 Billion and 65.2±13.2/Million people compared to 4.31±1.7/$100 Billion and 1.60±0.60/Million people for non-funded states (p < 0.05). Poisson multivariate regression identified system funding as an independent predictor of number of Level 3 centers.  Data expressed as mean ± SEM.

Conclusion:
Our study shows that the number of Level 3 trauma centers significantly and independently correlated with the presence of a funded trauma system. The number of Level 1 and 2 centers showed no such correlation. As Level 3 trauma centers are a key point of entry for trauma care, further study is warranted to determine if increased Level 3 access leads to improved time to definitive care or other clinical outcomes.  Furthermore, our study shows that states that allocate funds for trauma systems operation have a greater number of Level 3 centers per dollar of GDP, even in states with lower tax bases, resulting in more cost efficient access.
 

« Previous Page
Next Page »