Sophia L Dellis

NYU Langone Medical Center, New York, New York, United States

Are you Sophia L Dellis?

Claim your profile

Publications (6)22.92 Total impact

  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: OBJECTIVES: Simulated mitral valve replacement may aid in the assessment of technical skills required for adequate performance in the operating room. We sought to design and assess a mitral valve replacement training station that is low-cost, nonperishable, portable, and reproducible as a first step in developing a mitral valve surgical skills curriculum. METHODS: Nineteen physicians (7 general surgery residents, 8 cardiothoracic surgery residents, and 4 attending cardiothoracic surgeons) underwent simulated mitral valve replacement testing. Simulated mitral valve replacement was performed on a training station consisting of a replaceable "mitral annulus" inside a restrictive "left atrium." Eight components of performance were graded on a 5-point scale. A composite score (100 point maximum) was calculated by weighting the grades by procedural time. The effect of training level was evaluated using analysis of variance and post hoc Tukey honestly significant difference. RESULTS: The speed of simulated mitral valve replacement varied among general surgery residents, cardiothoracic surgery residents, and attending cardiothoracic surgeons (52.9 ± 9.0 vs 32.8 ± 4.7 vs 28.0 ± 3.5 minutes, respectively; F = 25.3; P < .001). Level of training significantly affected all 8 evaluation components (P < .001). Composite scores increased with level of training (general surgery residents 32.9 ± 11.4, cardiothoracic surgery residents 65.1 ± 11.5, and attending cardiothoracic surgeons 88.3 ± 7.8 of a possible 100 points; F = 35.7; P < .001). Cardiothoracic surgery residents who reported having performed 10 to 50 mitral valve replacements as the primary surgeon had a composite score of 65.0 ± 2.8 (P < .01 compared with attending cardiothoracic surgeons). CONCLUSIONS: Simulated mitral valve replacement can be performed using this simple, affordable, portable setup. Performance scores correlate with level of training and experience, but residents who performed 10 to 50 mitral valve replacements still failed to reach attending-level proficiency. This training simulator may facilitate skills practice and evaluation of competency in cardiac surgery trainees.
    The Journal of thoracic and cardiovascular surgery 10/2012; 145(1). DOI:10.1016/j.jtcvs.2012.09.074 · 3.99 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: While it is known that band annuloplasty for functional mitral regurgitation (FMR) improves leaflet coaptation, the effect on regional coaptation geometry has not previously been well defined. We used three-dimensional transesophageal echocardiography (3D-TEE) to analyze the regional effects of semirigid band annuloplasty on annular geometry and leaflet coaptation zones of patients with FMR. Sixteen patients with severe FMR underwent a semirigid band annuloplasty. Intraoperative full volume 3D-TEE datasets were acquired pre valve and post valve repair. Offline analysis assessed annular dimensions and regional coaptation zone geometry. The regions were defined as R1 (A1-P1), R2 (A2-P2), and R3 (A3-P3); coaptation distance, coaptation depth, and coaptation length were measured in each region. Differences were analyzed with repeated measures within a general linear model. Band annuloplasty decreased mitral regurgitation grade from 3.7 to 0.1 (scale 0 to 4). Annular septolateral dimension (p<0.01) and coaptation distance (p<0.01) decreased significantly in all regions. Likewise, anterior and posterior leaflet coaptation lengths increased in all regions (p<0.01 and p=0.05, respectively), with region 2 showing the greatest increase (p=0.01). Changes in coaptation depth were not significant. Semirigid band annuloplasty for FMR produces significant regional remodeling of leaflet coaptation zones, with region 2 showing the greatest increase in leaflet coaptation length. This regional analysis of annular geometry and leaflet coaptation creates a framework to better understand the mechanisms of surgical success or failure of annuloplasty for FMR.
    The Annals of thoracic surgery 04/2012; 93(6):1876-80. DOI:10.1016/j.athoracsur.2012.02.066 · 3.65 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Perfusion strategies and operative techniques for minimally invasive mitral valve repair have evolved over time. During the past decade, our institution's approach has progressed from a port access platform with femoral perfusion to predominantly a central aortic cannulation through a right anterior minithoracotomy incision. We analyzed this institutional experience to evaluate the impact of approach on patient outcomes. Between 1995 and 2007, 1282 patients (mean age, 59.3 years; range, 18-99 years) underwent first-time, isolated mitral valve repair using a minimally invasive technique. Patient demographics included peripheral vascular disease (3.2%), chronic obstructive pulmonary disease (8.3%), atherosclerotic aorta (6.5%), cerebrovascular disease (4.3%), and ejection fraction less than 30% (4.3%). Retrograde perfusion was performed in 394 (30.7%) of all patients and endoaortic balloon occlusion in 373 (29.1%); the operative technique was a right anterior minithoracotomy in 1264 (98.6%) and left posterior minithoracotomy in 18 (1.4%). The etiology of mitral disease was degenerative in 73.2%, functional in 20.6%, and rheumatic in 2.4%. Data were collected prospectively using the New York State Cardiac Surgery Report System and a customized minimally invasive surgery data form. Logistic analysis was used to evaluate risk factors and outcomes; operative experience was divided into tertiles. Overall hospital mortality was 2.0% (25/1282). Mortality was 1.1% (10/939) for patients with degenerative etiology and 0.4% (3/693) for patients younger than 70 years of age with degenerative valve disease. Risk factors for death were advanced age (P = .007), functional etiology (P = .010; odds ratio [OR] = 3.3), chronic obstructive pulmonary disease (P = .013; OR = 3.4), peripheral vascular disease (P = .014; OR = 4.2), and atherosclerotic aorta (P = .03; OR = 2.8). Logistic risk factors for neurologic events were advanced age (P = .02), retrograde perfusion (P = .001; OR = 3.8), and emergency procedure (P = .01; OR = 66.6). Interaction modeling revealed that the only significant risk factor for neurologic event was the use of retrograde perfusion in high-risk patients with aortic disease (P = .04; OR = 8.5). Analysis of successive tertiles during this 12-year experience revealed a significant decrease in the use of retrograde arterial perfusion (89.6%, 10.4%, and 0.0%; P < .001) and endoaortic balloon occlusion (89.3%, 10.7%, and 0%; P < .001). The overall frequency of postoperative neurologic events was 2.3% (30/1282) and decreased from 4.7% in the first tertile to 1.2% in the second and third tertiles (P < .001). Central aortic cannulation through a right anterior minithoracotomy for mitral valve repair allows excellent outcomes in patients with a broad spectrum of comorbidities and has become our preferred approach for most patients undergoing mitral valve repair. Retrograde arterial perfusion is associated with an increased risk of stroke in patients with severe peripheral vascular disease and should be reserved for select patients without significant atherosclerosis.
    The Journal of thoracic and cardiovascular surgery 01/2012; 143(4 Suppl):S68-70. DOI:10.1016/j.jtcvs.2012.01.011 · 3.99 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: A Society of Thoracic Surgeons' publication recently associated "minimally invasive" approaches with increased neurologic complications; this proposed association was questionable due to imprecise definitions. To critically reevaluate this issue, we reviewed a large minimally invasive valve experience with robust definitions. From November 1995 to January 2007, 3,180 isolated, non-reoperative valve operations were performed; 1,452 (45.7%) were aortic replacements and 1,728 (54.3%) were mitral valve procedures. Surgical approach was standard sternotomy (28%) or minimally invasive technique (72%). Antegrade arterial perfusion was used in 2,646 (83.2%) patients and retrograde perfusion in 534 (16.8%). Aortic clamping was direct in 83.4%, with endoclamp in 16.4% and no clamp in 0.2%. Patients were prospectively followed in a proprietary database and the New York State Cardiac Surgery Reporting System (mandatory, government audited). A neurologic event was defined as a permanent deficit, a transient deficit greater than 24 hours, or a new lesion on cerebral imaging. Hospital mortality for aortic valve replacement was 4.0% (sternotomy [5.1%] versus minimally invasive [3.4%] p = 0.13); for mitral procedures it was 2.4% (sternotomy [4.8%] versus minimally invasive [1.8%] p = 0.001). Multivariate analysis revealed that age, female gender, renal disease, ejection fraction less than 0.30, chronic obstructive pulmonary disease, and emergent operation were risk factors for mortality. Stroke occurred in 71 patients (2.2%) (sternotomy [2.1%] versus minimally invasive [2.3%] p = 0.82). Multivariate analysis of neurologic events revealed that cerebrovascular disease, emergency procedure, no-clamp, and retrograde perfusion were risk factors. In patients 50 years old or younger (n = 662), retrograde perfusion had no significant impact on neurologic events (1.6% vs 1.1%, p = 0.57). A minimally invasive approach with antegrade perfusion does not result in increased neurologic complications. Retrograde perfusion, however, is associated with increased neurologic risk in older patients.
    The Annals of thoracic surgery 10/2011; 92(4):1346-9; discussion 1349-50. DOI:10.1016/j.athoracsur.2011.04.055 · 3.65 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Current video-assisted thoracoscopic surgery training models rely on animals or mannequins to teach procedural skills. These approaches lack inherent teaching/testing capability and are limited by cost, anatomic variations, and single use. In response, we hypothesized that video-assisted thoracoscopic surgery right upper lobe resection could be simulated in a virtual reality environment with commercial software. An anatomy explorer (Maya [Autodesk Inc, San Rafael, Calif] models of the chest and hilar structures) and simulation engine were adapted. Design goals included freedom of port placement, incorporation of well-known anatomic variants, teaching and testing modes, haptic feedback for the dissection, ability to perform the anatomic divisions, and a portable platform. Preexisting commercial models did not provide sufficient surgical detail, and extensive modeling modifications were required. Video-assisted thoracoscopic surgery right upper lobe resection simulation is initiated with a random vein and artery variation. The trainee proceeds in a teaching or testing mode. A knowledge database currently includes 13 anatomic identifications and 20 high-yield lung cancer learning points. The "patient" is presented in the left lateral decubitus position. After initial camera port placement, the endoscopic view is displayed and the thoracoscope is manipulated via the haptic device. The thoracoscope port can be relocated; additional ports are placed using an external "operating room" view. Unrestricted endoscopic exploration of the thorax is allowed. An endo-dissector tool allows for hilar dissection, and a virtual stapling device divides structures. The trainee's performance is reported. A virtual reality cognitive task simulation can overcome the deficiencies of existing training models. Performance scoring is being validated as we assess this simulator for cognitive and technical surgical education.
    The Journal of thoracic and cardiovascular surgery 01/2011; 141(1):249-55. DOI:10.1016/j.jtcvs.2010.09.014 · 3.99 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Functional mitral regurgitation (FMR) is associated with leaflet displacement and tethering. Little is known about regional coaptation zones, including variations in coaptation length (CL) and contributions of anterior and posterior leaflets. Regional coaptation zones were analyzed in patients with normal mitral valves and with FMR. Cardiac surgery patients underwent a three-dimensional transesophageal echocardiography. Four-dimensional volumetric datasets were acquired with Doppler interrogation. Offline analysis was performed. Orthogonal views were extracted in diastole and systole. Leaflet dimensions and coaptation distance and depth were examined for posterior and apical displacement of the coaptation zones. Twenty patients were analyzed (10 normal and 10 with 2 to 4+ FMR). Anterior leaflet CL was greater than posterior leaflet CL: 2.2+/-0.6 mm versus 0.9+/-0.3 mm in region 1, 3.2+/-0.7 mm versus 1.2+/-0.6 mm in region 2, and 1.8+/-0.4 mm versus 0.6+/-0.3 mm in region 3 (p<0.001). The FMR was associated with shorter leaflet CLs, with a mean anterior CL of 1.7+/-0.4 mm versus 3.1+/-0.4 mm (p=0.04), and a mean posterior CL of 0.7+/-0.3 mm versus 1.1+/-0.3 mm (p=0.03). The biggest difference in CLs was in A2-P2. Coaptation distance and depth were higher in the FMR group: 21.7+/-1.0 mm versus 17.9+/-1.0 mm (p=0.01), and 8.6+/-0.7 mm versus 5.0+/-0.7 mm (p<0.01). Mitral valve leaflet CL is asymmetric in normal valves, with anterior dominance. Functional mitral regurgitation is associated with a relocated coaptation zone, regional changes, and diminished coaptation. These data suggest an "anterior leaflet reserve." Posterior movement of the coaptation line compensates for annular dilation and presumed left ventricular enlargement in order to maintain competency until inadequate anterior leaflet CL occurs.
    The Annals of thoracic surgery 04/2010; 89(4):1158-61. DOI:10.1016/j.athoracsur.2009.12.061 · 3.65 Impact Factor