Network
Latest external collaboration on country level. Dive into details by clicking on the dots.
Publication
Featured researches published by Ann H. Kim.
Annals of Vascular Surgery | 2015
Pamela A. Moorehead; Ann H. Kim; Claire P. Miller; Tejas V. Kashyap; Daniel E. Kendrick; Vikram S. Kashyap
BACKGROUND Bovine aortic arch (BA) occurs in approximately 15-35% of the US population and is regarded as a clinically insignificant, normal variant. The aim of this study was to assess the prevalence of types I (type I bovine arch [T1BA], common origin of innominate and/or left common carotid artery) and II (type II bovine arch [T2BA], left common carotid originating from innominate) bovine arch in patients with and without thoracic aortic pathology. METHODS We retrospectively reviewed all serial computed tomography images (n = 817) performed at our institution over 4 months to determine the overall prevalence of BA. Thoracic aorta and/or arch vessels were visualized, with images read by certified radiologists. A separate analysis compared a series of 156 consecutive patients with thoracic pathology (dissection or aneurysm ≥ 4.0 cm), from a 25-month period, with 757 control patients without pathology from the original sample. Statistical analysis included a chi-squared contingency table. RESULTS Analysis revealed a bovine arch prevalence of 31.1% (n = 254), including 14.9% T1BA and 16.2% T2BA. Patients with thoracic aortopathy (n = 156) had aortic dissection (n = 26) or aneurysm (n = 130). These patients were older and had an increased prevalence of hypertension, hyperlipidemia, and aortic calcification. In addition, there was increased prevalence of T2BA in the pathology group (23.7%) compared with controls (15.9%; P = 0.03). T1BA was not significantly different between groups (11.5% vs. 14.9%; P = 0.59). When thoracic disease was stratified by pathology type, T2BA occurred more frequently in patients with thoracic aortic aneurysm (24.6% vs. 15.9%; P = 0.04). BA trended upward, in patients with thoracic aortic dissection (42.3% vs. 30.8%; P = 0.28). CONCLUSIONS Our analyses revealed a prevalence of bovine arch of 31% in our patient population. BA occurred more frequently in patients with thoracic aortopathy than controls. Therefore, patients with BA may be associated with higher levels of thoracic aortic pathology and may benefit from increased clinical vigilance.
Journal of Vascular Surgery | 2016
John C. Wang; Ann H. Kim; Vikram S. Kashyap
Acute limb ischemia (ALI) is one of the most common vascular emergencies, with high risk for limb loss if it is not treated expediently. Endovascular therapy is less invasive and used increasingly because of patient factors that disfavor open surgery despite limited quality data to support its safety and efficacy. This evidence summary reviews literature from 1990 to 2014, comparing contemporary surgical and endovascular revascularization. Systematic review was performed with emphasis on acuity of presentation, study design, revascularization techniques, limb salvage and mortality rates, and complications. There were 2999 articles identified and 563 abstracts reviewed; 68 articles were reviewed fully and 26 critically appraised. Limb salvage, amputation-free survival, overall survival and mortality, and treatment complications were elucidated, including Medicare outcomes data. Risk factors for amputation and mortality were identified. Surgical or endovascular revascularization for ALI is achievable with acceptable limb salvage and amputation rates, which are not markedly different between the two modalities in the short term. Endovascular therapy and surgery are complementary rather than competing strategies for ALI. Further good-quality clinical trial data are needed to define longer term outcomes.
Journal of Vascular Surgery | 2016
Daniel E. Kendrick; Claire P. Miller; Pamela A. Moorehead; Ann H. Kim; Henry Baele; Virginia L. Wong; David W. Jordan; Vikram S. Kashyap
OBJECTIVE Endovascular intervention exposes surgical staff to scattered radiation, which varies according to procedure and imaging equipment. The purpose of this study was to determine differences in occupational exposure between procedures performed with fixed imaging (FI) in an endovascular suite compared with conventional mobile imaging (MI) in a standard operating room. METHODS A series of 116 endovascular cases were performed over a 4-month interval in a dedicated endovascular suite with FI and conventional operating room with MI. All cases were performed at a single institution and radiation dose was recorded using real-time dosimetry badges from Unfors RaySafe (Hopkinton, Mass). A dosimeter was mounted in each room to establish a radiation baseline. Staff dose was recorded using individual badges worn on the torso lead. Total mean air kerma (Kar; mGy, patient dose) and mean case dose (mSv, scattered radiation) were compared between rooms and across all staff positions for cases of varying complexity. Statistical analyses for all continuous variables were performed using t test and analysis of variance where appropriate. RESULTS A total of 43 cases with MI and 73 cases with FI were performed by four vascular surgeons. Total mean Kar, and case dose were significantly higher with FI compared with MI. (mean ± standard error of the mean, 523 ± 49 mGy vs 98 ± 19 mGy; P < .00001; 0.77 ± 0.03 mSv vs 0.16 ± 0.08 mSv, P < .00001). Exposure for the primary surgeon and assistant was significantly higher with FI compared with MI. Mean exposure for all cases using either imaging modality, was significantly higher for the primary surgeon and assistant than for support staff (ie, nurse, radiology technologist) beyond 6 feet from the X-ray source, indicated according to one-way analysis of variance (MI: P < .00001; FI: P < .00001). Support staff exposure was negligible and did not differ between FI and MI. Room dose stratified according to case complexity (Kar) showed statistically significantly higher scattered radiation in FI vs MI across all quartiles. CONCLUSIONS The scattered radiation is several-fold higher with FI than MI across all levels of case complexity. Radiation exposure decreases with distance from the radiation source, and is negligible outside of a 6-foot radius. Modern endovascular suites allow high-fidelity imaging, yet additional strategies to minimize exposure and occupational risk are needed.
Journal of Endovascular Therapy | 2016
Daniel E. Kendrick; Matthew T. Allemang; Andre F. Gosling; Anil Nagavalli; Ann H. Kim; Setsu Nishino; Sahil A. Parikh; Hiram G. Bezerra; Vikram S. Kashyap
Purpose: To examine the hypothesis that alternative flush media could be used for lower extremity optical coherence tomography (OCT) imaging in long lesions that would normally require excessive use of contrast. Methods: The OPTical Imaging Measurement of Intravascular Solution Efficacy (OPTIMISE) trial was a single-center, prospective study (ClinicalTrials.gov identifier NCT01743872) that enrolled 23 patients (mean age 68±11 years; 14 men) undergoing endovascular intervention involving the superficial femoral artery. Four flush media (heparinized saline, dextran, carbon dioxide, and contrast) were used in succession in random order for each image pullback. Quality was defined as ≥270° visualization of vessel wall layers from each axial image. Mean proportions (± standard deviation) of image quality for each flush medium were assessed using 1-way analysis of variance and are reported with the 95% confidence intervals (CI). Results: Four OCT catheters failed, leaving 19 patients who completed the OCT imaging protocol; from this cohort, 51 highest quality runs were selected for analysis. Average vessel diameter was 3.99±1.01 mm. OCT imaging allowed 10- to 15-μm resolution of the lumen border, with diminishing quality as vessel diameter increased. Plaque characterization revealed fibrotic lesions. Mean proportions of image quality were dextran 87.2%±12% (95% CI 0.81 to 0.94), heparinized saline 74.3%±24.8% (95% CI 0.66 to 0.93), contrast 70.1%±30.5% (95% CI 0.52 to 0.88), and carbon dioxide 10.0%±10.4% (95% CI 0.00 to 0.26). Dextran, saline, and contrast provided better quality than carbon dioxide (p<0.001). Conclusion: OCT is feasible in peripheral vessels <5 mm in diameter. Dextran or saline flush media can allow lesion characterization, avoiding iodinated contrast. Carbon dioxide is inadequate for peripheral OCT imaging. Axial imaging may aid in enhancing durability of peripheral endovascular interventions.
Journal of Vascular Surgery | 2017
Andre F. Gosling; Daniel E. Kendrick; Ann H. Kim; Anil Nagavalli; Ethan S. Kimball; Nathaniel T. Liu; Vikram S. Kashyap; John C. Wang
Objective: Outcomes from carotid artery stenting (CAS) are related to experience and technical expertise of the operator. Simulation of CAS may enhance clinical proficiency. We interrogated the impact of endovascular simulation of CAS procedures in operators who are at various stages of training. Methods: Twelve trainees (students [n = 4]; junior surgery residents, postgraduate year [PGY] 1‐3 [n = 4]; and senior surgery residents or fellows, PGY 4‐7 [n = 4]) were apprised of characteristics of an endovascular simulator and CAS procedures. This was followed by four independent sessions that were assessed for objective measures including procedure and fluoroscopy times and contrast agent use. A qualitative analysis grading steps of CAS by two observers using a Likert scale was performed. One‐way analysis of variance and paired t‐tests were employed for data analysis. Results: For all participants (n = 12), procedure times (mean, 920 ± 279 seconds for the first session vs 454 ± 156 seconds for the fourth session; P < .01; confidence interval [CI], 315‐621) and fluoroscopy cumulative times (mean, 421 ± 230 seconds for the first session vs 222 ± 102 seconds for the fourth session; P < .01; CI, 78‐285) decreased with progression of cases. Students and PGY 1‐3 residents decreased their procedure times significantly in comparison of initial and final sessions (P < .05 and P < .01, respectively). For all groups, fluoroscopy cumulative times were reduced, and this decrement was significant in the PGY 1‐3 cohort (mean, 444 ± 8 seconds for the first session vs 265 ± 51 seconds for the fourth session; P < .01; CI, 81‐276). Initial CAS procedure times were significantly different between groups (P < .05), but this was observed to resolve by the final case at study completion. Qualitatively, the Likert scores of students and PGY 1‐3 residents significantly improved with case repetition, specifically in the following steps: (1) cannulation of common carotid artery and (2) sizing and deployment of embolic protection device. Senior operators (PGY 4‐7) demonstrated consistently better performance overall with minimal change in scoring with case repetition. Conclusions: Practice leads to improvements in endovascular simulator procedure and fluoroscopy times, especially for more novice trainees. Initial operator performance gaps can be approximated with a few sessions to expected proficiency. Incorporation of endovascular simulators in residency training may assist in shortening the learning curve in rarer endovascular procedures.
Journal of Vascular Surgery | 2017
Katherine L. Morrow; Ann H. Kim; Steven A. Plato; Andrew J. Shevitz; Jerry Goldstone; Henry Baele; Vikram S. Kashyap
Objective: Percutaneous mechanical thrombectomy (PMT) is regularly used in the treatment of both venous and arterial thrombosis. Although there has been no formal report, PMT has been linked to cases of reversible postoperative acute kidney injury (AKI). The purpose of this study is to evaluate the risk of renal dysfunction in patients undergoing PMT vs catheter‐directed thrombolysis (CDT) for treatment of an acute thrombus. Methods: This study is a retrospective review of all patients in a single institution with a Current Procedural Terminology code for PMT or CDT from January 2009 through December 2014. Each patient was grouped into one of the four following procedural categories: PMT only, PMT with tissue plasminogen activator (tPA) pulse‐spray, PMT with CDT, or CDT only. Preoperative and postoperative creatinine and glomerular filtration rate (GFR) values were obtained for each patient. The RIFLE (Risk, Injury, Failure, Loss, and End‐stage renal disease) criteria were used to categorize the extent of renal dysfunction. χ2 analysis, one‐way analysis of variance, and unpaired t‐test were used to assess significance. Results: A total of 227 patients were reviewed, of which 82 were excluded due to either existence of preoperative AKI, history of end‐stage renal disease, or lack of clinical data. Of the remaining 145 patients, 53 (37%) presented with arterial thrombosis (mean age, 62 years; 43% male) and 92 (63%) presented with venous thrombosis (mean age, 48 years; 45% male). The incidence of renal dysfunction was highest in the PMT/tPA pulse group (21%), followed by the PMT group (20%) and the PMT/CDT group (14%). CDT was not associated with renal dysfunction. PMT (P = .046), and PMT/tPA pulse (P = .033) were associated with higher rates of renal dysfunction than the CDT controls. The average preoperative GFR for the 22 patients who developed AKI was 53.7 ± 9.4 mL/min/1.73 m2. The minimum postoperative GFR within 48 hours was an average of 35 ± 16 mL/min/1.73 m2. Stratified by the RIFLE criteria, 13 (9%) patients progressed to the risk category, 6 (4%) progressed to the injury category, and 3 (2%) progressed to the failure category. None of the patients who developed renal dysfunction from PMT progressed to dialysis within the same admission period. Conclusions: The use of PMT as a treatment for vascular thrombosis is associated with renal dysfunction. Patients treated with PMT require postoperative vigilance and renal protective measures.
Journal of Vascular Surgery | 2017
Matthew Janko; Ryan Moore; Ann H. Kim; Andrew J. Shevitz; Katherine L. Morrow; David J. Johnson; Vikram S. Kashyap
Objective Asymptomatic internal carotid artery occlusion (CO) presents a clinical dilemma, and presently, the natural history, stroke risk, and optimal management remain ill defined. This study compared outcomes, including neurovascular events (NVEs) and health care costs, between patients with CO and patients with asymptomatic carotid artery stenosis (CS). Methods A prospectively maintained database was queried to identify patients with CO and CS with at least >50% carotid stenosis by duplex. We identified and reviewed 622 consecutive patients with asymptomatic carotid artery disease at one academic medical center between 2011 and 2013. Patients with CO (n = 97) were identified and propensity matched by age and gender in a 1:2 ratio with CS patients (n = 194) for further analyses. Univariate and multivariate models were used to analyze baseline characteristics, clinical variables, and 1‐year follow‐up data from the date of diagnosis. Multivariate analysis was performed by multiple linear regression modeling. Institutional Review Board approval was obtained. Results Follow‐up data were available for 99% of matched patients. CO patients were younger (72 vs 75 years; P < .01) and more likely male (67% vs 53%; P = .01) compared with CS patients. After propensity matching, baseline characteristics were similar between groups, with a trend toward higher use of statin therapy among patients with CO. Antiplatelet therapy was used in 79% of patients with CS and in 74% of patients with CO (P = .45). The rate of NVE among CO patients was higher than among CS patients at 1 year of follow‐up (14% vs 7%; P = .03). Among those with NVE, neither antiplatelet therapy (64% vs 77%; P = .49) nor statin therapy (86% vs 77%; P = .58) appeared to have a significant effect. Health care costs (
Journal of Vascular Surgery | 2017
Ann H. Kim; Andrew J. Shevitz; Katherine L. Morrow; Daniel E. Kendrick; Karem Harth; Henry Baele; Vikram S. Kashyap
14,361 vs
Journal of Vascular Surgery | 2015
Ann H. Kim; Daniel E. Kendrick; Vikram S. Kashyap
12,142; P = .44) and hospital admission rate (63% vs 71%; P = .18) were similar between groups. Not surprisingly, the rate of vascular procedures was higher in the CS group (55% vs 27%; P = .04). Conclusions Patients with asymptomatic CO experience more NVEs compared with similar patients with moderately severe CS. Further study of preventative strategies, including intensity of medical therapy, is warranted.
Journal of Vascular Surgery | 2016
Ann H. Kim; Daniel E. Kendrick; Pamela A. Moorehead; Anil Nagavalli; Claire P. Miller; Nathaniel T. Liu; John C. Wang; Vikram S. Kashyap
Objective Digital subtraction angiography (DSA) of the peripheral arterial vasculature provides lumenographic information but only a qualitative assessment of blood flow. The ability to quantify adequate tissue perfusion of the lower extremities would enable real‐time perfusion assessment during DSA of patients with peripheral arterial disease (PAD). In this study, we used a novel real‐time imaging software to delineate tissue perfusion parameters in the foot in PAD patients. Methods Between March 2015 and June 2016, patients (N = 31) underwent lower extremity angiography using a two‐dimensional perfusion (2DP) imaging protocol (Philips Healthcare, Andover, Mass). Of the 31 enrolled patients, 16 patients received preintervention and postintervention DSA images (18 angiograms), while contrast agent injection settings and the position of the foot, catheter, and C‐arm were kept constant. The region of interest for perfusion measurements was taken at the level of the medial malleolus. Perfusion parameters included arrival time (AT) of contrast material, wash‐in rate (WIR), time to peak (TTP) contrast intensity, and area under the curve (AUC). Results Patients (mean age, 67 years; male, 61%) undergoing 2DP had limbs classified as Rutherford class 3 (n = 9 limbs), class 4 (n = 11), and class 5 (n = 14) ischemia with a mean ankle‐brachial index of 0.63. For the whole cohort, median (interquartile range) AT measured 5.20 (3.10‐7.25) seconds; WIR, 61.95 (43.53‐86.43) signal intensity (SI)/s; TTP, 3.80 (2.88‐4.50) seconds; peak intensity, 725.00 (613.75‐1138.00) SI; and AUC, 12,084.00 (6742.80‐17,059.70) SI*s. A subset of patients had 2DP performed before and after intervention (n = 18 cases). A detectable improvement in SI and two‐dimensional flow parameters was seen after intervention. Average AT of contrast material to the region of interest shortened after intervention with percentage decrease of 30.1% ± 49.1%, corresponding decrease in TTP of 17.6% ± 24.7%, increase in WIR of 68.8% ± 94.2% and in AUC of 10.5% ± 37.6%, decrease in mean transit time of 18.7% ± 28.1%, and increase in peak of 34.4% ± 42.2%. Conclusions The 2DP imaging allows measurement of blood flow in real time as an adjunct to DSA. The AT may be the most sensitive marker of perfusion change in the lower extremity. Quantitative thresholds based on 2DP hold promise for immediate treatment effectiveness assessment in patients with PAD.