Eric J. Keller

All Publications

• Research Ethics in IR: The Intersection Between Care and Progress. Journal of vascular and interventional radiology : JVIR Bozorghadad, S., Newton, I. G., Perez, A. W., Makary, M. S., Keller, E. J. 2020; 31 (5): 846?48

  View details for DOI 10.1016/j.jvir.2020.02.014

  View details for PubMedID 32359529

• Informed Consent: Beating a Dead Horse or an Opportunity for Quality Improvement? Journal of vascular and interventional radiology : JVIR Keller, E. J., Perez, A. W., Makary, M. S. 2020; 31 (1): 139?40

  View details for DOI 10.1016/j.jvir.2019.08.010

  View details for PubMedID 31864541

• Efficacy of endovascular Z-configuration stenting for malignant versus nonmalignant caval obstruction. Journal of vascular surgery. Venous and lymphatic disorders Han, T. M., Bondarev, S., Keller, E. J., Vogelzang, R. L., Resnick, S. A. 2020

  Abstract

  The objective of this study was to assess factors associated with symptom resolution after endovascular stenting for superior or inferior vena cava syndrome.Eighty-six consecutive vena cava Z-configuration stent placements in 82 patients (53 ± 14 years old) at a single institution were reviewed for patient demographics, comorbidities, and durability of stent patency (also evaluated were persistent or recurrent symptoms, stent occlusion, and need for repeated stenting). Logistic regression was used to identify independent factors associated with stent patency, and ? coefficients and analysis of variance were used to compare cases subdivided by lesion location (superior vena cava, inferior vena cava) and the presence or absence of malignant disease.Clinical follow-up was available in 77 of 86 (90%) cases. Technical success with clinical failure (persistent symptoms) occurred in 40% of these cases with a median follow-up of 67 (interquartile range, 14-570) days and mortality rate of 63% during this period. Malignant obstructions had a significantly higher clinical failure rate of 54% compared with 15% for nonmalignant obstructions (? = 0.34; P = .002). However, only metastatic disease was independently associated with clinical failure when controlling for demographics, other comorbidities, and differential follow-up (adjusted odds ratio, 8.27; 95% confidence interval, 2.79-24.50).Vena cava Z-stenting effectively resolves symptoms in 85% of nonmalignant obstructions compared with only 46% of malignant obstructions. Patients should be counseled accordingly, and those with malignant obstructions may require closer follow-up to evaluate the need for reintervention and goals of care.

  View details for DOI 10.1016/j.jvsv.2020.02.022

  View details for PubMedID 32414673

• You're Performing My Procedure: Teamwork and Tribalism in IR. Journal of vascular and interventional radiology : JVIR Bozorghadad, S., Perez, A. W., Makary, M. S., Keller, E. J. 2019

  View details for DOI 10.1016/j.jvir.2019.09.021

  View details for PubMedID 31771897

• Predictors of Disease Recurrence after Venoplasty and Stent Placement for May-Thurner Syndrome JOURNAL OF VASCULAR AND INTERVENTIONAL RADIOLOGY Bondarev, S., Keller, E. J., Han, T., Young, V. A., Gupta, S., Vogelzang, R. L., Eskandari, M., Resnick, S. A. 2019; 30 (10): 1549?54

  Abstract

  To identify factors independently associated with disease recurrence after venoplasty and stent placement for May-Thurner syndrome (MTS).Fifty-nine consecutive patients (age, 47 y ± 15; 93% female) were identified who had undergone endovascular stent placement for MTS. Patient charts were reviewed for demographic data, risk factors for venous thrombosis, comorbidities, and venous inflow or outflow at first follow-up (3 wk to 6 mo after treatment). Logistic regression was used to identify independent predictors of symptom recurrence or repeat intervention, and multivariate analysis of variance and receiver operator characteristic curve analysis were used to assess relationships between degrees of in-stent stenosis and other variables in the 73% of patients with available cross-sectional imaging. Median follow up was 20.7 months (interquartile range, 4.7-49.5 mo).All procedures were technically successful. Disease recurrence, defined as symptom recurrence following initial postprocedural resolution, was observed in 38% of patients. No preprocedural variable was found to be independently predictive of disease recurrence; however, poor venous inflow or outflow were both strongly associated with recurrent disease, with adjusted odds ratios and 95% confidence intervals of 38.02 (3.76-384.20; P = .002) and 7.00 (1.15-42.71; P = .04), respectively. Higher degrees of in-stent stenosis were also associated with symptom recurrence, with an area under the curve of 0.93 (P = .000002) and 39%-41% stenosis being 78%-83% sensitive and 88%-92% specific for symptom recurrence.These results suggest that cross-sectional imaging can help differentiate patients in whom closer follow-up may be warranted after venoplasty and stent placement for MTS and also guide counseling regarding prognosis.

  View details for DOI 10.1016/j.jvir.2019.07.012

  View details for Web of Science ID 000490349600005

  View details for PubMedID 31526576

• Gluteal Vein Anatomy: Location, Caliber, Impact of Patient Positioning, and Implications for Fat Grafting. Aesthetic surgery journal Turin, S. Y., Fracol, M., Keller, E., Markl, M., Collins, J., Krochmal, D., Kim, J. Y. 2019

  Abstract

  BACKGROUND: Deaths in gluteal autografting occur due to gluteal vein injuries, but data is lacking on the precise location and caliber of these veins.OBJECTIVES: To present the first in-vivo study of gluteal vein anatomy using MRI.METHODS: MRI venography of 16 volunteer hemi-sections was conducted in the supine, prone, prone with a bump (jack-knife), left and right decubitus positions in one session after a single contrast administration. Caliber and course of the superior and inferior gluteal veins (SGV/IGV) were analyzed versus bony landmarks and position changes.RESULTS: The SGV has a very short submuscular course before splitting into 2 smaller branches superolaterally. The IGV runs immediately deep to the gluteus maximus in the center of the buttock as a single large trunk, on average 56mm deep (mean 27mm of muscle belly and 30mm subcutaneous fat). No intramuscular or subcutaneous branches >2mm were found. In the prone position, the IGV and SGV have an average caliber of 5.96mm and 5.63mm. Vessel caliber decreased by 21% and 27%, respectively in the jack-knife position and by 14% and 15% in lateral decubitus.CONCLUSIONS: The SGV and IGV are immediately deep to gluteus maximus approximately 6 cm deep with a caliber on the order of 6mm in the prone position. The distribution of these vessels suggests there is no "safe zone" in the intramuscular or submuscular planes. The jackknife or lateral decubitus positions can decrease vein caliber by up to 27%, possibly reducing the risk of injury due to either traction or direct cannula impact.

  View details for DOI 10.1093/asj/sjz260

  View details for PubMedID 31574144

• Reconsidering Requests-Futility in IR. Journal of vascular and interventional radiology : JVIR Clark, J. M., Keller, E. J. 2019; 30 (6): 961?62

  View details for DOI 10.1016/j.jvir.2019.01.025

  View details for PubMedID 31126607

• Reflect and Remember: The Ethics of Complications in Interventional Radiology. Seminars in interventional radiology Keller, E. J. 2019; 36 (2): 104?7

  Abstract

  Clinicians, particularly those in procedural specialties, tend to feel personally responsible when complications occur. Medical errors among surgeons have been studied and provide an insightful window into the ethics of complications. Ethically we must consider what we owe patient and families, ourselves, and our colleagues. To some degree, the answers are similar: reflection, confession, and resolution . We owe patients and families an explanation and lack of abandonment; we owe ourselves thoughtful reflection on what caused the complication and how we can learn from it; and we owe our colleagues assistance managing our complications and a chance to learn from our mistakes. As a specialty that prides itself on innovation and novel therapies, interventional radiology has a unique relationship with complications that has not been well developed. As the specialty grows, it will be important to provide forums for further understanding the ethical challenges in interventional radiology.

  View details for DOI 10.1055/s-0039-1688423

  View details for PubMedID 31123380

• Cost-Effectiveness of a Guided Peripherally Inserted Central Catheter Placement System: A Single-Center Cohort Study JOURNAL OF VASCULAR AND INTERVENTIONAL RADIOLOGY Keller, E. J., Aragona, E., Molina, H., Lee, J., Salem, R., Resnick, S. A., Chrisman, H., Collins, J. D. 2019; 30 (5): 709?14

  Abstract

  To assess the cost-effectiveness of peripherally inserted central catheter (PICC) placements using an ultrasound and electrocardiogram-guided system versus external measurements and confirmatory chest X-rays (CXRs).Sixty-eight guided PICC placements were performed in 63 outpatients (mean age, 43 ± 13 years; 50% male) and compared to 68 propensity score-matched PICC placements (mean age, 44 ± 13 years; 56% male) performed using external measurements by the same operators. Post-placement CXRs were used to confirm final catheter tip positioning. Cohorts were compared in terms of repositioning rates, desired tip positioning rates (in the lower third of the superior vena cava or at the cavoatrial junction), and estimated cost per PICC positioned as desired using manufacturer quotes, Medicare reimbursement rates, and hourly wages for staff time. Agreement between tip positioning according to the guided system versus CXR was also assessed.Guided PICC placements required less repositioning (1.5% vs 10.3%, P = .03) and resulted in more catheters positioned as desired (86.8% vs 67.6%, P = .01) than the external measurement approach. The cost per PICC positioned as desired was lower for guided placements ($318.54 vs $381.44), suggesting that the guided system was cost-effective in this clinical setting. Guided system-CXR agreement for tip position was poor (?=0.25, P = .002) due to tips being slightly farther from the cavoatrial junction on CXR than indicated by the guided system.The guided PICC placement system was cost-effective in outpatients treated by a single division of interventional radiology at an academic institution.

  View details for DOI 10.1016/j.jvir.2018.07.032

  View details for Web of Science ID 000468259000015

  View details for PubMedID 30773436

• The growing pains of physician-administration relationships in an academic medical center and the effects on physician engagement PLOS ONE Keller, E. J., Giafaglione, B., Chrisman, H. B., Collins, J. D., Vogelzang, R. L. 2019; 14 (2): e0212014

  Abstract

  Physician engagement has become a key metric for healthcare leadership and is associated with better healthcare outcomes. However, engagement tends to be low and difficult to measure and improve. This study sought to efficiently characterize the professional cultural dynamics between physicians and administrators at an academic hospital and how those dynamics affect physician engagement.A qualitative mixed methods analysis was completed in 6 weeks, consisting of a preliminary analysis of the hospital system's history that was used to purposefully recruit 20 physicians across specialties and 20 healthcare administrators across management levels for semi-structured interviews and observation. Participation rates of 77% (20/26) and 83% (20/24) were achieved for physicians and administrators, respectively. Cohorts consisted of equal numbers of men and women with experience ranging from 1 to 35 years within the organization. Field notes and transcripts were systematically analyzed using an iterative inductive-deductive approach. Emergent themes were presented and discussed with approximately 400 physicians and administrators within the organization to assess validity and which results were most meaningful.This investigation indicated a professional cultural disconnect was undermining efforts to improve physician engagement. This disconnect was further complicated by a minority (10%) not believing an issue existed and conflicting connotations not readily perceived by participants who often offered similar solutions. Physicians and administrators felt these results accurately reflected their realities and used this information as a common language to plan targeted interventions to improve physician engagement. Limitations of the study included its cross-sectional nature with a modest sample size at a single institution.A qualitative mixed methods analysis efficiently identified professional cultural barriers within an academic hospital to serve as an institution-specific guide to improving physician engagement.

  View details for DOI 10.1371/journal.pone.0212014

  View details for Web of Science ID 000458761300069

  View details for PubMedID 30759151

  View details for PubMedCentralID PMC6373942

• Understanding Bias: A Look at Conflicts of Interest in IR. Journal of vascular and interventional radiology : JVIR Clark, J. M., Anderson, D., Makary, M. S., Keller, E. J. 2019; 30 (5): 765?66

  View details for DOI 10.1016/j.jvir.2019.01.002

  View details for PubMedID 31029391

• Who We Are and What We Can Become: The Anthropology of IR and Challenges of Forming a New Specialty JOURNAL OF VASCULAR AND INTERVENTIONAL RADIOLOGY Keller, E. J., Vogelzang, R. L. 2018; 29 (12): 1703?4

  View details for DOI 10.1016/j.jvir.2018.07.021

  View details for Web of Science ID 000453341200012

  View details for PubMedID 30502878

• Single-Center Retrospective Review of Radiofrequency Wire Recanalization of Refractory Central Venous Occlusions Keller, E. J., Gupta, S. A., Bondarev, S., Sato, K. T., Vogelzang, R. L., Resnick, S. A. ELSEVIER SCIENCE INC. 2018: 1571?77

  Abstract

  To retrospectively review the effectiveness and safety of radiofrequency (RF) wire recanalization of refractory central venous occlusions (CVOs) and compare recurrent and nonrecurrent CVOs in terms of patient and occlusion characteristics.Twenty CVOs were treated in 18 patients (age 40 y ± 13; 9 women) with 11 superior vena cava (SVC) or brachiocephalic vein occlusions (ie, supradiaphragmatic) and 9 inferior vena cava or iliac vein occlusions (ie, infradiaphragmatic). Indications included pain, edema, ulceration, and/or dialysis arteriovenous fistula dysfunction peripheral to the CVO(s). All patients had multiple venous thrombotic risk factors, including mechanical venous compression, endothelial injury, and/or coagulopathies. CVO traversal was first attempted with standard and advanced techniques before RF wire recanalization and followed up with computed tomographic venography and clinic visits approximately 1, 3, 6, and 12 months after treatment.Sixteen CVOs (80%) were successfully transversed and associated with symptom relief. One major complication occurred involving SVC perforation into the pericardial space. Primary CVO patency rate was 56% at a median follow-up of 14.1 months (interquartile range [IQR], 9.2-20.0 mo). Recurrent CVOs tended to be infradiaphragmatic (71% vs 12% for supradiaphragmatic; P = .02), longer (12.9 cm ± 10.0 vs 2.3 cm ± 1.3; P < .01), and associated with implanted venous stents, filters, or cardiac pacer/defibrillator leads (86% vs 22%; P = .01). Median time to restenosis/occlusion was 1.5 months (IQR, 1.1-6.1 mo).RF wire recanalization is a relatively effective and safe option for refractory CVOs. Patients with longer, infradiaphragmatic CVOs associated with indwelling devices may require closer follow-up for CVO recurrence.

  View details for DOI 10.1016/j.jvir.2018.06.017

  View details for Web of Science ID 000450542400012

  View details for PubMedID 30293732

• Reinforcing the Importance and Feasibility of Implementing a Low-dose Protocol for CT-guided Biopsies ACADEMIC RADIOLOGY Keller, E. J., Lewandowski, R. J., Goodwin, L., Yaghmai, V., Nemcek, A., Carr, J. C., Collins, J. D. 2018; 25 (9): 1146?51

  Abstract

  This study sought to more definitely illustrate the impact and feasibility of implementing a low-dose protocol for computed tomography (CT)-guided biopsies using size-specific dose estimates and multivariate analyses.Fifty consecutive CT-guided lung and extrapulmonary biopsies were reviewed before and after implementation of a low-dose protocol (200 patients total, mean age 61?±?15 years, 128 women). Analyses of variance with Bonferroni correction were used to compare standard and low-dose protocols in terms of patient demographics, physician experience, target lesion size, total dose-length product, total acquisitions, size-specific dose estimate, signal-to-noise ratio, contrast-to-noise ratio, and lesion conspicuity ratings. All procedures were performed on the same 16-slice CT scanner.Voluntary protocol adherence was 100% (lung) and 89% (extrapulmonary). The low-dose protocol achieved significantly lower total average dose-length product [(lung) 735.6?±?599.4?mGy?×?cm to 252.1?±?101.9?mGy?×?cm, P?

  View details for DOI 10.1016/j.acra.2018.01.003

  View details for Web of Science ID 000454183700007

  View details for PubMedID 29426686

• Perceptions of Quality in Interventional Oncology JOURNAL OF VASCULAR AND INTERVENTIONAL RADIOLOGY Keller, E. J., Kennedy, K. Y., Patel, A. P., Ivanovic, M., Collins, J. D., Sato, K., Thornburg, B. G., Riaz, A., Nemcek, A. A., Desai, K., Lewandowski, R. J., Salem, R., Vogelzang, R. L., Hickey, R. 2018; 29 (3): 367?72

  Abstract

  To inductively characterize perceptions of quality in interventional oncology (IO) based on values and experiences of patients and referring providers.Brief ethnographic interviews were completed with referring providers and patients before and after a variety of liver-directed procedures about their experiences, concerns, and perceptions of IO services at a single institution. Constructivist grounded theory was used to systematically analyze interview transcripts for themes until thematic saturation was achieved. All transcripts were analyzed by a reviewer with 3-years of experience performing such analyses, and 50% were randomly selected to be coded by 2 additional blinded reviewers. Interreviewer agreement was assessed via Cohen ?.Interviews with 22 patients (mean age, 65 y ± 13; 9 women) and 12 providers (mean age, 54 y ± 9; 6 women) were required to reach and confirm thematic saturation. Interreviewer agreement for interview themes was excellent (? = 0.78; P < .001). Perceptions of high-quality IO care relied on interventional radiologists being responsive, friendly, and open; engaging in multidisciplinary collaboration; having thoughtful, dedicated support staff; and facilitating well-coordinated care after procedures and follow-up more than technical expertise and periprocedural comfort. Patient and provider perceptions of quality differed, but disjointed care after procedures was the most common critique among both groups.An inductive qualitative approach effectively characterized specific aspects of perceptions of high-quality IO care among patients and referring providers.

  View details for DOI 10.1016/j.jvir.2017.10.033

  View details for Web of Science ID 000427215900013

  View details for PubMedID 29395900

• The consistency of myocardial strain derived from heart deformation analysis INTERNATIONAL JOURNAL OF CARDIOVASCULAR IMAGING Keller, E. J., Fang, S., Lin, K., Freed, B. H., Smith, P. M., Spottiswoode, B. S., Davids, R., Carr, M., Jolly, M., Markl, M., Carr, J. C., Collins, J. D. 2017; 33 (8): 1169?77

  Abstract

  The purpose of this study was to assess the consistency of semi-automated myocardial strain analysis by prototype software across field strengths, temporal resolutions, and examinations. 35 volunteers (48?±?13 years; 20% women) and 25 patients (54?±?12 years; 44% women) without significant cardiac dysfunction underwent cine cardiac magnetic resonance imaging (CMR) at 1.5 T with a temporal resolution of 39.2 msec. 34 subjects also underwent imaging at 3.0 T; 16 had repeat examinations within 14 days; and 9 underwent CMR with temporal resolutions of 12.5 and 39.2 msec on the same day. Prototype heart deformation analysis (HDA) software was used to retrospectively quantify strain from segmented balanced steady state free precession (bSSFP) cinegraphic images. Myocardial contours were automatically generated on short axis images and drawn at end-diastole by two independent reviewers on long-axis images. Contours were automatically propagated throughout the cardiac cycle. Global and regional peak systolic strain were compared across observers, field strengths, temporal resolutions, and examinations. Inter-observer agreement was excellent (ICC?>?0.87, p??0.05). Normal global peak systolic strain values with HDA were -25.0 (-24.0 to -26.1)% (LV circumferential), 60.5 (55.3 to 65.6)% (LV radial), -22.3 (-20.5 to -?24.0)% (LV longitudinal), and -26.0 (-23.8 to -28.2)% (RV longitudinal). HDA prototype software enabled efficient and consistent quantification of myocardial strain from conventional bSSFP cine CMR data, demonstrating clinical feasibility.

  View details for DOI 10.1007/s10554-017-1090-6

  View details for Web of Science ID 000405226700009

  View details for PubMedID 28239799

• Four-Dimensional Flow MRI-Based Splenic Flow Index for Predicting Cirrhosis-Associated Hypersplenism AMERICAN JOURNAL OF ROENTGENOLOGY Keller, E. J., Kulik, L., Stankovic, Z., Lewandowski, R. J., Salem, R., Carr, J. C., Schnell, S., Markl, M., Collins, J. D. 2017; 209 (1): 46?53

  Abstract

  The objective of this study is to evaluate the ability of spleen volume, blood flow, and an index incorporating multiple measures to predict cirrhosis-associated hypersplenism.A total of 39 patients (14 women and 25 men; mean [± SD] age, 52 ± 10 years) with cirrhosis and sequelae of portal hypertension underwent 4D flow MRI and anatomic 3-T MRI performed before and after contrast administration. Unenhanced 4D flow MRI was used to assess abdominal hemodynamics, and splenic volumes were measured on T1-weighted gradient-recalled echo MRI. Relationships among demographic characteristics, blood component counts, splenic volume, arterial flow, venous flow, and the percentage of shunted portal flow were assessed in 29 consecutive patients (i.e., the derivation group), to develop a splenic flow index. This index was assessed along with splenic volume and blood flow alone in 10 additional consecutive patients (i.e., the validation group) via ROC curve analysis, to identify platelet counts of less than 50 × 103 cells/?L, leukocyte counts of less than 3.0 × 103 cells/?L, or both.In the derivation cohort (platelet count, 129 ± 76 × 103 cells/?L), splenic volume, arterial flow, venous flow, and the percentage of shunted portal flow were inversely correlated with platelet counts (? = -0.68, -0.68, -0.56, and -0.36, respectively; p < 0.05). Adding splenic volume to arterial flow and the product of venous flow and the percentage of shunted portal flow indexed to the body surface area yielded superior correlations with platelet counts, leukocyte counts, and the degree of severity of hypersplenism (? = -0.75, -0.48, and -0.75, respectively; p ? 0.001) and predicted severe hypersplenism (sensitivity, 100%; specificity, 100%) in the validation cohort (platelet count, 93 ± 71 × 103 cells/?L).A splenic flow index that incorporates both splenic volume and blood flow is a better indicator of hypersplenism than is splenic volume alone.

  View details for DOI 10.2214/AJR.16.17620

  View details for Web of Science ID 000404057000019

  View details for PubMedID 28463524

• Who We Are and What We Can Become: An Analysis of Professional Identity Formation in IR JOURNAL OF VASCULAR AND INTERVENTIONAL RADIOLOGY Keller, E. J., Mcgee, K. A., Resnick, S. A., Trerotola, S. O., Valji, K., Johnson, M. S., Collins, J. D., Vogelzang, R. L. 2017; 28 (6): 850?56

  Abstract

  To characterize the unique experiences, values, and perspectives of interventional radiology (IR) fellows.Sixteen fellows from 4 US vascular and IR programs were interviewed within 2 months of beginning and 2-3 months following their 2015-2016 fellowships about patient interactions, training experiences, and views of IR and other specialties. Interviews were systematically analyzed for dominant themes by using constructivist grounded theory. Four interviews with 2015-2016 interventional cardiology fellows, 16 interviews with IR attending physicians, and online descriptions of IR were also analyzed for context. Themes were compared qualitatively and quantitatively.Interobserver agreement was good for interview themes (? = 0.70; P < .0001). IR fellows' professional identity emerged primarily from radiologic and surgical interests, with distinct emphasis on being "innovators," "thinking differently," and "needing to adapt and advertise abilities to survive." Fellows' descriptions of patient care were more clinically focused than past interviews with attending physicians (P = .05), but clinical interests common in medical specialties were limited, and descriptions of "nonprocedural patient care" were primarily periprocedural (81%). Descriptions of the future of the field conveyed competing pressures, loose role definition, and disconnect between academic and private-practice IR.IR fellows share professional interests, views of their field and others, and descriptions of patient care, but there is uncertainty regarding future roles of the specialty and a need for more specific and unified definitions of nonprocedural patient care in IR.

  View details for DOI 10.1016/j.jvir.2017.02.006

  View details for Web of Science ID 000403132500012

  View details for PubMedID 28292636

• Why Vascular Surgeons and Interventional Radiologists Collaborate or Compete: A Look at Endovascular Stent Placements CARDIOVASCULAR AND INTERVENTIONAL RADIOLOGY Keller, E. J., Collins, J. D., Crowley-Matoka, M., Chrisman, H. B., Milad, M. P., Vogelzang, R. L. 2017; 40 (6): 814?21

  Abstract

  To understand how cultural differences between vascular surgeons (VSs) and interventional radiologists (IRs) affect their clinical decision making and inter-specialty relationships.Twenty-four conversational interviews were conducted with IRs and VSs about their approaches to patient care, views of their specialty and others, and solutions to any expressed concerns. Interview transcripts were systematically analyzed to identify and compare key themes according to the constructivist grounded theory and content analysis using NVivo 10 software. These data were supplemented with a retrospective analysis of 3658 endovascular stent placements performed at a large medical academic center over 11 years. Aggregate counts were divided by provider specialty, and trends were assessed via correlation coefficients.Endovascular stent placements were relatively equally divided between IR and VS over 11 years with some variability from placements by cardiology. IRs tend to lay claim to treatments as masters of procedures, whereas VSs base their claims on being masters of the treated diseases, leading to collaboration in some practices and bitter competition in others. The level of perceived competition was most associated with specialists' awareness of and appreciation for specialty-specific values rather than differences in practice structure/reimbursement.Understanding cultural differences between IRs and VSs is imperative for fostering better collaboration to grow shared territory rather than competing for the same slice of the pie.

  View details for DOI 10.1007/s00270-017-1570-z

  View details for Web of Science ID 000400443200002

  View details for PubMedID 28062898

• Specialty-Specific Values Affecting the Management of Symptomatic Uterine Fibroids JOURNAL OF VASCULAR AND INTERVENTIONAL RADIOLOGY Keller, E. J., Crowley-Matoka, M., Collins, J. D., Chrisman, H. B., Milad, M. P., Vogelzang, R. L. 2017; 28 (3): 420?28

  Abstract

  To better understand why interventional radiologists and gynecologists differ in their approaches to symptomatic uterine fibroids.Conversational interviews were conducted with 26 interventional radiologists and gynecologists about their professional roles, clinical reasoning, and practice variation within and outside their specialty. Interview transcripts were systematically analyzed using NVivo 10 software (QSR International, Burlington, Massachusetts) according to grounded theory and content analysis to identify key themes and compare themes across specialties and practice environments. Data were supplemented with retrospective analysis of 7,659 patients with symptomatic uterine fibroids treated at a large academic center over 11 years.Interventional radiologists' shares of symptomatic uterine fibroid treatment and endovascular stent treatments have remained constant (P > .05) for 11 y at a large medical center, whereas minimally invasive gynecologic fibroid treatments and the percentage of interventional radiology (IR) procedures reimbursed by Medicaid/Medicare have increased significantly (r > .90, P < .001 and r = .93, P < .001). Interventional radiologists and gynecologists shared a commitment to do "the right thing" for patients, but each group possessed distinct professional values affecting how they viewed medical evidence, outcomes, and their colleagues. When differences were apparent and concerning, physicians tended to suspect ulterior motives not in patients' best interests.Interventional radiologists and gynecologists demonstrated wide-ranging perspectives regarding their role in caring for patients with symptomatic uterine fibroids. To promote genuine collaboration and adoption of shared goals, stakeholders should seek and promote a deeper understanding of specialty-specific values and culture.

  View details for DOI 10.1016/j.jvir.2016.11.008

  View details for Web of Science ID 000397075000016

  View details for PubMedID 28082073

• Superior Abdominal 4D Flow MRI Data Consistency with Adjusted Preprocessing Workflow and Noncontrast Acquisitions ACADEMIC RADIOLOGY Keller, E. J., Collins, J. D., Rigsby, C., Carr, J. C., Markl, M., Schnell, S. 2017; 24 (3): 350?58

  Abstract

  To assess the impact of an alternative preprocessing workflow on noncontrast- and contrast-enhanced abdominal four-dimensional flow magnetic resonance imaging (4D flow MRI) data consistency.Twenty patients with cirrhosis and portal hypertension (5 women; 53?±?10 years old) underwent 4D flow MRI at 3.0T before and after administration of 0.03?mmol/kg of gadofosveset trisodium with velocity sensitivities of 100 and 50?cm/s for arterial and venous flow quantifications, respectively. 4D flow MRI data were preprocessed using the conventional workflow (workflow 1), applying noise filters prior to eddy current correction, and an alternative workflow (workflow 2), first correcting for eddy currents and using noise filtering only if needed for anti-aliasing. Vessel segmentation quality was ranked by independent reviewers and compared via Wilcoxon signed-rank tests. Flow quantification and conservation of mass at two portal and one arterial branch points were compared via paired t tests.Segmentation quality was significantly higher for workflow 2 (P??0.05). Noncontrast acquisitions yielded significantly smaller portal flow values (P?

  View details for DOI 10.1016/j.acra.2016.10.007

  View details for Web of Science ID 000394727700014

  View details for PubMedID 27940231

• Fostering better policy adoption and inter-disciplinary communication in healthcare: A qualitative analysis of practicing physicians' common interests PLOS ONE Keller, E. J., Crowley-Matoka, M., Collins, J. D., Chrisman, H. B., Milad, M. P., Vogelzang, R. L. 2017; 12 (2): e0172865

  Abstract

  In response to limited physician adoption of various healthcare initiatives, we sought to propose and assess a novel approach to policy development where one first characterizes diverse physician groups' common interests, using a medical student and constructivist grounded theory.In 6 months, a medical student completed 36 semi-structured interviews with interventional radiologists, gynecologists, and vascular surgeons that were systematically analyzed according to constructivist grounded theory to identifying common themes. Common drivers of clinical decision making and professional values across 3 distinct specialty groups were derived from physicians' descriptions of their clinical decision making, stories, and concerns.Common drivers of clinical decision making included patient preference/benefit, experience, reimbursement, busyness/volume, and referral networks. Common values included honesty, trustworthiness, loyalty, humble service, compassion and perseverance, and practical wisdom. Although personal gains were perceived as important interests, such values were easily sacrificed for the good of patients or other non-financial interests. This balance was largely dependent on the incentives and security provided by physicians' environments.Using a medical student interviewer and constructivist grounded theory is a feasible means of collecting rich qualitative data to guide policy development. Healthcare administrators and medical educators should consider incorporating this methodology early in policy development to anticipate how value differences between physician groups will influence their acceptance of policies and other broad healthcare initiatives.

  View details for DOI 10.1371/journal.pone.0172865

  View details for Web of Science ID 000394688200157

  View details for PubMedID 28235088

  View details for PubMedCentralID PMC5325554

• Reduction of aberrant aortic haemodynamics following aortic root replacement with a mechanical valved conduit INTERACTIVE CARDIOVASCULAR AND THORACIC SURGERY Keller, E. J., Malaisrie, S., Kruse, J., McCarthy, P. M., Carr, J. C., Markl, M., Barker, A. J., Collins, J. D. 2016; 23 (3): 416?23

  Abstract

  Previous work suggests that aortic root and valve prostheses alter blood flow patterns in the ascending aorta, creating aberrant haemodynamics compared with those of healthy volunteers. Various valve designs have been proposed to better restore physiological haemodynamics. In this study, magnetic resonance imaging (MRI) was used to non-invasively assess three-dimensional (3D) ascending aortic haemodynamics after aortic root replacement (ARR) with a mechanical valved conduit postulated to create less turbulent blood flow.Ten patients (40 ± 9 years) underwent transthoracic echocardiography and contrast-enhanced multidimensional four-dimensional (4D) flow MRI at 1.5 T after ARR with an On-X mechanical valved conduit. Preoperative 4D flow MRI was available in 7 patients. Ten age- and gender-matched healthy volunteers (42 ± 13 years) were also analysed to characterize physiological flow. The presence of vortex/helix formation was graded by two blinded observers. Peak transvalvular pressure gradients were computed using the simplified Bernoulli equation. Patients' postoperative pressure gradients and helicity/vorticity grades were compared with preoperative gradients and those from healthy volunteers.Intra- and interobserver ratings showed good agreement (? = 0.93, P < 0.01 and ? = 0.84, P < 0.01, respectively). Highly helical and/or vortical flow was observed in all patients preoperatively, which was significantly reduced postoperatively (P < 0.01 and <0.01, respectively), restoring similar flow patterns similar to those seen in volunteers (P = 0.56 and 0.56). Peak transvalvular pressure gradients (?P) were also significantly reduced [43 ± 21 vs 12 ± 7 mmHg, P < 0.05 (Echo); 48 ± 22 vs 16 ± 9 mmHg, P < 0.05 (MRI)], but remained significantly higher than those of volunteers (6 ± 1 mmHg, P < 0.01).Preliminary evidence suggests that ARR with an On-X mechanical valve significantly reduces aberrant aortic haemodynamics, producing flow patterns that resemble those in healthy volunteers.

  View details for DOI 10.1093/icvts/ivw173

  View details for Web of Science ID 000383902600012

  View details for PubMedID 27245620

  View details for PubMedCentralID PMC6390273

• Physicians' professional identities: a roadmap to understanding "value" in cardiovascular imaging JOURNAL OF CARDIOVASCULAR MAGNETIC RESONANCE Keller, E. J., Vogelzang, R. L., Freed, B. H., Carr, J. C., Collins, J. D. 2016; 18: 52

  Abstract

  Quality improvement efforts in cardiovascular imaging have been challenged by limited adoption of initiatives and policies. In order to better understand this limitation and inform future efforts, the range clinical values related to cardiovascular imaging at a large academic hospital was characterized.15 Northwestern Medicine physicians from internal medicine, cardiology, emergency medicine, cardiac/vascular surgery, and radiology were interviewed about their use of cardiovascular imaging and imaging guidelines. Interview transcripts were systemically analyzed according to constructivist grounded theory and combined with 56 previous interviews with interventional radiologists, interventional cardiologists, gynecologists, and vascular surgeons to develop a model describing specialty-specific values. This model was applied to the 15 pilot interviews focused on cardiovascular imaging, highlighting specialty specific differences in values and practice patterns. Transcripts were also reviewed independently by a cardiologist and 2 radiologists followed by a group discussion to assess reproducibility and achieve a consensus regarding the results.Differences in perceived value of cardiovascular imaging and use of guidelines among physicians were well explained by three value-associated identity categories (managers, diagnosticians, and fixers) that were further differentiated along three axes (broad v. focused-thinkers, complex v. definitive-answer-seekers, and public visibility).Quality improvement in cardiovascular imaging may be limited by a lack of understanding and incorporation of the complexity of medical culture into ongoing initiatives. Both individually and during policy development, it is important to first understand the complexity of stakeholders' diverse perceptions of "value," "quality," and "appropriateness."

  View details for DOI 10.1186/s12968-016-0274-x

  View details for Web of Science ID 000383227100002

  View details for PubMedID 27566058

  View details for PubMedCentralID PMC5002193

• Providing Context: Medical Device Litigation and Inferior Vena Cava Filters SEMINARS IN INTERVENTIONAL RADIOLOGY Keller, E. J., Vogelzang, R. L. 2016; 33 (2): 132?36

  Abstract

  Over the last few years, an increasing number of lawsuits have been filed involving inferior vena cava filters. This has prompted the U.S. Judicial Panel on Multidistrict Litigation to centralize these lawsuits into two multidistrict litigations: one for Cook's filters and one for Bard's. Both sets of cases share similar questions of facts, in particular whether these filters' design and manufacturing practices made them unreasonably prone to serious complications. The resolution of these cases will add to a larger legal debate concerning how much legal protection the 1976 Medical Device Amendments should offer firms from tort liability. As a specialty that often relies on medical devices, it is not only important for interventional radiologists to have a general understanding of medical device litigation but also to reflect upon the approaches to informed consent regarding these devices.

  View details for DOI 10.1055/s-0036-1581086

  View details for Web of Science ID 000376217000011

  View details for PubMedID 27247482

  View details for PubMedCentralID PMC4862861

• Spatio-temporal Visualization of Regional Myocardial Velocities Eurographics Workshop on Visual Computing for Biology and Medicine Sheharyar, A., Chitiboi, T., Keller, E. J., Rahman, O., Schnell, S., Markl, M., Bouhali, O., Linsen, L. 2016: 89?98
• Philosophy in Medical Education: A Means of Protecting Mental Health ACADEMIC PSYCHIATRY Keller, E. J. 2014; 38 (4): 409?13

  Abstract

  This study sought to identify and examine less commonly discussed challenges to positive mental health faced by medical students, residents, and physicians with hopes of improving current efforts to protect the mental health of these groups. Additionally, this work aimed to suggest an innovative means of preventing poor mental health during medical education.Literature on medical student, resident, and physician mental health was carefully reviewed and a number of psychiatrists who treat physician-patients were interviewed.The culture of medicine, medical training, common physician psychology and identity, and conflicting professional expectations all seem to contribute to poor mental health among medical students, residents, and physicians. Many current efforts may be more successful by better addressing the negative effects of these characteristics of modern medicine.Programs aimed at promoting healthy mental lifestyles during medical education should continue to be developed and supported to mitigate the deleterious effects of the challenging environment of modern medicine. To improve these efforts, educators may consider incorporating philosophical discussions on meaning and fulfillment in life between medical students and faculty. Through medical school faculty members sharing and living out their own healthy outlooks on life, students may emulate these habits and the culture of medicine may become less challenging for positive mental health.

  View details for DOI 10.1007/s40596-014-0033-y

  View details for Web of Science ID 000339803500005

  View details for PubMedID 24477902

• Ethical Considerations Surrounding Survival Benefit-Based Liver Allocation LIVER TRANSPLANTATION Keller, E. J., Kwo, P. Y., Helft, P. R. 2014; 20 (2): 140-146

  Abstract

  The disparity between the demand for and supply of donor livers has continued to grow over the last 2 decades, and this has placed greater weight on the need for efficient and effective liver allocation. Although the use of extended criteria donors has shown great potential, it remains unregulated. A survival benefit-based model was recently proposed to answer calls to increase efficiency and reduce futile transplants. However, it was previously determined that the current allocation system was not in need of modification and that instead geographic disparities should be addressed. In contrast, we believe that there is a significant need to replace the current allocation system and complement efforts to improve donor liver distribution. We illustrate this need first by identifying major ethical concerns shaping liver allocation and then by using these concerns to identify strengths and shortcomings of the Model for End-Stage Liver Disease/Pediatric End-Stage Liver Disease system and a survival benefit-based model. The latter model is a promising means of improving liver allocation: it incorporates a greater number of ethical principles, uses a sophisticated statistical model to increase efficiency and reduce waste, minimizes bias, and parallels developments in the allocation of other organs. However, it remains limited in its posttransplant predictive accuracy and may raise potential issues regarding informed consent. In addition, the proposed model fails to include quality-of-life concerns and prioritize younger patients. We feel that it is time to take the next steps toward better liver allocation not only through reductions in geographic disparities but also through the adoption of a model better equipped to balance the many ethical concerns shaping organ allocation. Thus, we support the development of a similar model with suggested amendments.

  View details for DOI 10.1002/lt.23780

  View details for Web of Science ID 000331194500003

  View details for PubMedID 24166860