Athanasios Tsiouris

Simplified Frailty Index to Predict Adverse Outcomes and Mortality in Vascular Surgery Patients

BACKGROUND Frailty has been established as an important predictor of health-care outcomes. We hypothesized that the use of a modified frailty index would be a predictor of mortality and adverse occurrences in vascular surgery patients. METHODS Under the data use agreement of the American College of Surgeons, and with institutional review board (IRB) approval, the National Surgical Quality Improvement Program (NSQIP) Participant Utilization File was accessed for the years 2005-2008 for inpatient vascular surgery patients. Using the Canadian Study of Health and Aging Frailty Index (FI), 11 variables were matched to the NSQIP database. An increase in FI implies increased frailty. The outcomes assessed were mortality, wound infection, and any occurrence. We then compared the effect of FI, age, functional status, relative value units (RVU), American Society of Anesthesiology (ASA) score, and wound status on mortality. Statistical analysis was done using chi-square analysis and stepwise logistic regression. RESULTS A total of 67,308 patients were identified in the database, 3913 wound occurrences, 6691 infections, 12,847 occurrences of all kinds, and 2800 deaths. As the FI increased, postoperative wound infection, all occurrences, and mortality increased (P < 0.001). Stepwise logistic regression using the FI with the NSQIP variables of age, work RVU, ASA class, wound classification, emergency status, and functional status showed FI to have the highest odds ratio (OR) for mortality (OR = 2.058, P < 0.001). CONCLUSIONS A simplified FI can be obtained by easily identifiable patient characteristics, allowing for accurate prediction of postoperative morbidity and mortality in the vascular surgery population.

A modified frailty index to assess morbidity and mortality after lobectomy.

BACKGROUND Frailty has yet to be explored as a risk factor for thoracic surgery. We hypothesized that our modified frailty index (mFI) may be a predictor of morbidity and mortality following lobectomy. MATERIALS National Surgical Quality Improvement Program (NSQIP) participant use files were reviewed (2005-2010). Patients undergoing lobectomy were identified based on Current Procedural Terminology code 32480. We used an mFI with 11 variables, based on mapping the Canadian Study of Health and Aging Frailty Index to the NSQIP comorbidities. Data were analyzed using χ(2) test, independent sample t-test, Jonckheere-Terpstra test, and logistic regression. RESULTS Of 1940 open lobectomy patients identified, morbidity and mortality uniformly increased as the mFI increased; 14.9% of patients (75/504) with mFI of 0 had at least one complication, compared with 32% of patients (91/284) with mFI of 0.27 (P < 0.001). An mFI of 0 was associated with a mortality rate of 1% (5/504), compared with 5.6% (16/284) for mFI of 0.27 (P < 0001). Failure to wean from the ventilator, reintubation, surgical site infections, pneumonia, and Clavien 4 and above complications occurred in 1.8% (9/504), 2.6% (13/504), 2.2% (11/504), 5.4% (27/504), and 4.2% (21/504), respectively, in patients with an mFI of 0, compared with 7.4% (21/284), 7% (22/284), 3.2% (9/284), 10.9% (31/284), and 14.4% (41/284), respectively, in patients with mFI of 0.27. CONCLUSIONS This study demonstrates that the mFI may identify patients at higher risk for morbidity and mortality post-lobectomy. With the aging population, preoperative selection is important in minimizing morbidity and mortality and improving risk stratification for informed decision-making.

Significance of Postoperative Acute Renal Failure After Continuous-Flow Left Ventricular Assist Device Implantation

BACKGROUND Deteriorating renal function is common in patients with advanced heart failure and is associated with poor outcomes. The relationship between renal function and left ventricular assist device (LVAD) implantation is complex and has been explored in multiple studies with contradictory results. The aim of our study is to examine the significance of postoperative renal failure after implantation of a continuous-flow LVAD and its relationship to outcomes. METHODS From March 2006 to July 2011, 100 patients underwent implantation of a HeartMate II (Thoratec Corp, Pleasanton, CA) or HeartWare (Heart International, Inc, Framingham, MA) LVAD at our institution. Patients were stratified based on postoperative development of acute renal failure (ARF). Variables were compared using 2-sided t tests, χ(2) tests, Cox proportional hazards models, and log-rank tests to determine whether there was a difference between the 2 groups and whether postoperative renal failure was a significant independent predictor of outcome. RESULTS We identified 28 patients (28%) with postoperative ARF and 72 patients (72%) without postoperative ARF. The 2 groups were similar with regard to demographics and comorbidities. The patients with ARF were more likely to be intubated preoperatively (14.3% versus 1.4%; p = 0.021) and had higher preoperative central venous pressure (CVP) (14.3 mm Hg versus 10.7 mm Hg; p = 0.015). Postoperatively patients with ARF had a longer hospital stay (32.4 versus 18.7; p = 0.05), were more likely to experience right ventricular (RV) failure (25% versus 5.6%; p = 0.01) and ventilator-dependent respiratory failure (VDRF) (28.6% versus 6.9%; p = 0.007). There was a significant difference when comparing the ARF and non-ARF groups for 30-day (17.9% versus 0%; p < 0.001), 180-day (28.6% versus 2.8%; p < 0.001), and 360-day mortality (28.6% versus 6.9%; p = 0.012). CONCLUSIONS Patients in whom ARF developed after LVAD implantation had a higher rate of VDRF and RV failure and a longer length of stay (LOS). Postoperative ARF was associated with higher mortality at the 30-day, 180-day, and 360-day intervals. ARF after LVAD may be an early marker of poor outcome, particularly RV failure, and may be an opportunity for early intervention and rescue.

Factors determining post-operative readmissions after left ventricular assist device implantation

BACKGROUND In the current era of extensive healthcare reform, there has been a heightened focus on the frequency and cause for readmissions within 30 days of discharge given that readmissions are not reimbursed by most providers. The aim of our study was to determine the frequency, etiology and patterns of 30-day readmissions among recipients of continuous-flow left ventricular assist devices (LVADs) at our institution as well to determine whether there were any significant predictors of readmission. METHODS From March 2006 through June 2013, 150 patients underwent implantation of a continuous-flow LVAD at our institution. Patients were stratified into two groups based on their 30-day readmission status. A total of 12 patients died before discharge and were excluded from our analysis. Causes for 30-day readmissions and duration of hospital stay for the readmissions were recorded. Numerous pre-operative variables and post-operative complications were compared using 2-sided t-tests and chi-square tests between patients who were and were not readmitted within 30 days of their discharge after their LVAD implant. RESULTS The 30-day readmission rate was 26.1% (36 of 138), with approximately 70% of post-operative readmissions occurring within 10 days of the patients initial hospital discharge. Recurrent heart failure (12 of 36, 33.3%) and gastrointestinal bleeding (8 of 36, 22.2%) were the most common causes for 30-day readmission. The median length of stay (LOS) for readmission was 11.7 days. Thirty-day readmission did not affect short- or long-term survival. On univariate analysis, post-operative gastrointestinal bleeding (GIB) was a significant risk factor for 30-day readmissions (HR 1.4, 95% CI 0.19 to 0.99, p = 0.05), and overall length of stay was a significant factor in reducing 30-day readmission rates (HR 0.91, 95% CI 0.85 to 0.99, p = 0.02). CONCLUSIONS Our experience indicates that 30-day readmission rates after LVAD implantation remain relatively high, with most occurring within 10 days of discharge. Recurrent heart failure and GIB were the most common causes of post-operative rehospitalization. In addition, GIB during the index hospitalization was a significant predictor of 30-day readmission.

Preoperative risk stratification for thoracic surgery using the American College of Surgeons National Surgical Quality Improvement Program data set: functional status predicts morbidity and mortality

BACKGROUND Preoperative risk stratification for noncardiac thoracic surgery focuses on predicting postoperative lung capacity and cardiac risk. We hypothesized that preoperative functional status may be a predictor of morbidity and mortality after thoracic surgery. METHODS The National Surgical Quality Improvement Program Participant Use Files from 2005 to 2009 were accessed, and current procedural terminology codes for procedures involving the lung and pleura were used to identify thoracic surgery patients. Patients were grouped by independent or dependent preoperative functional status. Risks of infectious and noninfectious complications were evaluated. Chi-square, Fisher exact, and multivariate analyses with stepwise logistic regression were used. This study was approved by the Institutional Review Board. RESULTS Of 6,373 patients, 812 had a preoperative dependent functional status. Dependent patients had significantly higher rates of infection, other adverse events, and mortality. They were 9.3 times more likely (odds ratio [OR] 9.3) to have prolonged ventilation (P < 0.001) and 3.1 times more likely to be reintubated (P < 0.001). Postoperative pneumonia occurred in 10% (OR 2.7, P < 0.001). Postoperative mortality was 7.7 times higher (P < 0.001). Preoperative functional status, wound classification 3, and emergency procedures were independent predictors of both morbidity and mortality. Performing a thoracotomy was an independent risk factor for postoperative morbidity but not mortality (P < 0.001, OR 2 versus P = 0.415, OR 1.1). CONCLUSIONS Thoracic surgery patients, classified as functionally dependent preoperatively are at high risk for major morbidity and mortality. Although a limited observational study, results show that functional status is an essential component of preoperative assessment for thoracic surgery patients.

Distal pancreatectomy with or without splenectomy: comparison of postoperative outcomes and surrogates of splenic function

OBJECTIVES Published data on splenic preservation during distal pancreatectomy have been inconsistent. We hypothesized that patients undergoing spleen-preserving distal pancreatectomy (SPDP) would have fewer infectious and non-infectious complications than those undergoing en bloc distal pancreatectomy with splenectomy (DPS), and that their haematological parameters would be consistent with splenic function. METHODS Of 97 patients who underwent either SPDP using the Warshaw technique or en bloc DPS, 78 met our study inclusion criteria. Records were reviewed for data on age, gender, resection, indications for resection, operative time, blood loss, transfusion requirements, hospital stay, infectious complications, any other complications, postoperative white blood cell (WBC) and platelet counts. Data were analysed using the chi-squared test, the two-sided Mann-Whitney-Wilcoxon text, and simple and multiple logistic regression analyses. A P-value of <0.05 was considered significant. RESULTS Patients undergoing SPDP had a shorter length of stay and shorter operative time, were more likely to be completed laparoscopically, less likely to require re-operation, and had fewer infectious and non-infectious complications. However, these differences were not statistically significant. In multiple logistic regression analyses, patient age and length of hospital stay were both significant predictors of the occurrence of non-infectious complications (P= 0.04 and P= 0.006, respectively). Blood transfusion was a significant predictor of postoperative morbidity (P= 0.013 for infectious complications; P= 0.018 for non-infectious complications). White blood cell count was a statistically significant predictor of infectious (P= 0.02) and non-infectious (P= 0.04) complications, whereas platelet count was not. Patients who underwent DPS had statistically significantly higher WBC and platelet counts immediately postoperatively and at 6 months compared with SPDP patients. Postoperative mortality in both the SPDP and DPS groups was 0%. None of the 30 SPDP patients had evidence of splenic infarction. Pancreatic leaks occurred in 18% of patients in the SPDP group, compared with 8% in the DPS group (P < 0.05). CONCLUSIONS Spleen-preserving distal pancreatectomy using the Warshaw technique is associated with lower postoperative morbidity than DPS. Lower WBC and platelet counts suggest better splenic function in SPDP patients.

Clostridium difficile of the ileum following total abdominal colectomy, with or without proctectomy: Who is at risk?

BACKGROUND: Clostridium difficile enteritis is considered a rare entity, although recent data suggest a significant increase in prevalence and incidence. There is paucity of data evaluating risk factors of C difficile enteritis following total colectomy. OBJECTIVE: The aim of this study was to determine the incidence and risk factors of C difficile enteritis for patients who had undergone total abdominal colectomy with or without proctectomy. DESIGN: This study involves a retrospective chart review of 310 patients. Univariate analysis was performed on potential risk factors (p ⩽ 0.05) with the use of a logistic regression model, and a Fisher exact test was used for variables that had no occurrences of C difficile. These groups of variables were then examined in a multiple variate setting with stepwise logistic regression analysis. SETTINGS: This study was conducted at a tertiary referral center. PATIENTS: A data analysis was performed on patients who had undergone total abdominal colectomy with or without proctectomy who were tested for C difficile of the ileum. RESULTS: Twenty-two of 137 patients that were tested (16%) were positive for C difficile of the ileum. Univariate analysis of known risk factors for C difficile demonstrated that black race was a protective factor against C difficile (p = 0.016). The multivariate analysis demonstrated that emergency surgery (p = 0.035), race (p = 0.003), and increasing age by decade (p = 0.033) were risk factors for C difficile. LIMITATIONS: This study was limited by the small patient sample, and it was not a randomized trial. CONCLUSIONS: Black race is protective, and whites are 4 times more likely to acquire C difficile of the ileum after undergoing a total abdominal colectomy with or without proctectomy. The data also demonstrated that an increased age by a decade and emergency surgery are risk factors for C difficile enteritis, whereas the described risk factors of C difficile of the colon and type of colon surgery do not appear to influence the risk of C difficile of the ileum.

Impact of concomitant cardiac procedures performed during implantation of long-term left ventricular assist devices

BACKGROUND There is a paucity of data evaluating the effect of performing a concomitant cardiac procedure (CCP) on peri-operative survival in patients undergoing implantation of long-term left ventricular assist devices (LVADs). The objective of this study was to review our single-institutional experience with patients who underwent a CCP during implantation of a long-term continuous-flow LVAD. METHODS From March 2006 through June 2012, 130 patients underwent implantation of a LVAD. Of these, 76 (58.5%) were implanted as bridge-to-transplant and 54 (41.5%) as destination therapy. The LVAD implantation was isolated in 95 patients and with CCP in 35. This included 19 tricuspid valve repairs, 14 aortic valve repair/replacements, and 2 patent foramen ovale closures. The LVAD only and LVAD+CCP groups were compared regarding pre-operative demographics, peri-operative and midterm survival, and the incidence of post-operative complications. RESULTS Pre-operative central venous pressure (p = 0.047), moderate to severe tricuspid regurgitation (p = 0.011), cardiopulmonary bypass time (p < 0.0001), cross-clamp time (p < 0.0001), and right ventricular end diastolic diameter (p = 0.039) were higher in the CCP group. Body mass index (p = 0.01) and body surface area (p = 0.037) were higher in the LVAD-only group. Peri-operative and midterm survival at 30 days, 6 months, 1 year, and 2 years was 94%, 87%, 80%, and 73%, respectively, for isolated LVAD implants vs. 97%, 90%, 86%, and 86%, respectively, for LVAD+CCP (p = NS). Survival was similar for LVAD patients with tricuspid valve repairs, with aortic valve repair, and with patent foramen ovale repair (p = NS). Cox proportional hazard models showed a CCP was not an independent predictor of outcome (p = NS). CONCLUSIONS CCPs performed during implantation of a long-term continuous-flow LVADs does not increase peri-operative or mid-term mortality. In addition, unlike previous reports, there was no additive procedural risk for patients undergoing concomitant aortic valve repair or replacement.

Robot-Assisted Minimally Invasive Ivor Lewis Esophagectomy With Real-Time Perfusion Assessment

BACKGROUND Surgical resection is viewed as the most effective way to ensure both locoregional control and long-term survival in esophageal cancer. Although minimally invasive esophagectomy has been widely accepted as an alternative to open surgery, the role of robotic assistance has yet to be elucidated. We report our institutional experience with robotic-assisted Ivor Lewis esophagectomy using real-time perfusion assessment and demonstrate this as a safe and technically feasible alternative to traditional open Ivor Lewis esophagectomy. METHODS A retrospective chart review of all patients undergoing robotic-assisted Ivor Lewis esophagectomy at a single institution from 2011 to 2014 was performed. Operative and postoperative outcomes were recorded. RESULTS Fifty-four patients underwent robotic-assisted Ivor Lewis esophagectomy during the study period. Indication for surgery was cancer in 49 patients, 38 of whom underwent neoadjuvant chemoradiation therapy. The average operative time was 6 hours 2 minutes, and the average blood loss was 74 mL. There was 1 postoperative mortality (1.9%). Three (5.5%) patients experienced an anastomotic leak. The average number of lymph nodes harvested in cancer patients was 16.2 (range, 3 to 35). The average length of stay was 12.9 days. CONCLUSIONS Our study demonstrates that robotic-assisted Ivor Lewis esophagectomy using real-time perfusion assessment is a safe and technically feasible alternative to traditional open Ivor Lewis esophagectomy. It allows for R0 resection with adequate lymph node harvesting and a short hospital stay.

Lessons learned from 150 continuous-flow left ventricular assist devices: a single institutional 7 year experience.

Continuous-flow (CF) left ventricular assist devices (LVADs) have become the standard of care for patients with advanced heart failure refractory to optimal medical therapy. The goal of this study was to review our 7 year single institutional experience with CF LVADs. Mean age was 50.4 + 12.5 (17–69) years for bridge-to-transplantation (BTT) patients and 57.6 + 10.4 (31–81) years for destination therapy (DT) patients (p < 0.001). Overall, 38 patients (26%) were female and 58 (41%) were African American. Etiology of heart failure was ischemic in 54 patients (37%) and nonischemic in 93 patients (63%). Overall survival at 30 days, 6 months, 12 months, and 2 years was 93%, 89%, 84%, and 81%, respectively. Gastrointestinal bleeding (GIB) was the most common complication (24%), followed by stroke (18%), right ventricular (RV) failure (18%), ventilator-dependent respiratory failure (10%), reoperation for bleeding (10%), and driveline infection (9%). These data demonstrate excellent survival with low mortality for both BTT and DT patients on long-term LVAD support. However, for LVAD therapy to become the gold standard for long-term treatment of end-stage heart failure and a plausible alternative to heart transplantation, we need to continue to improve the incidence of frequent postoperative complications, such as RV failure, driveline infections, strokes, and GIB.