69.18 Predictors of Mortality in Patients Undergoing Pancreatic Resections in a Tertiary Healthcare Center

Posted on January 18, 2016

L. I. Amodu1, A. Levy1, M. Akerman1, M. Tiwari1, G. Georgiev1, M. Beltran Del Rio1, J. Nicastro1, G. F. Coppa1, E. P. Molmenti1, H. Rilo1 1North Shore University And Long Island Jewish Medical Center,Center For Diseases Of The Pancreas,Manhasset, NY, USA

Introduction: Surgical resection remains the mainstay of treatment for pancreatic cancer and several benign pancreatic conditions. While historically considered to be high-risk and complex procedures, morbidity and mortality from pancreatic resections have decreased dramatically. We carried out this study to determine independent predictors of mortality following pancreatic resection in a single tertiary healthcare institution.

Methods: Clinical data was collected retrospectively from the records of 111 patients who underwent pancreatic resections from 2004-2013. Mortality was ascertained by the use of hospital records, and the publicly accessible social security death master file (SS-DMF). A univariate screen of possible factors associated with mortality was performed using the Fisher’s exact test for categorical variables and the Mann-Whitney test for continuous variables; a multivariate analysis using logistic regression was performed to determine the independent effects of each factor. Area under the ROC curve (AUC) was used to evaluate the predictive accuracy of the chosen model.

Results: Mortality at 1, 3, and 5 years was 13.2%, 38.4%, and 45.9% respectively. Multivariate logistic regression found that a history of diabetes mellitus (OR = 6.48, p<0.0026), adenocarcinoma (OR = 8.17, p<0.0003), disease recurrence (loco-regional or distant) after initial resection (OR = 4.25, p<0.0235), and post-operative hypotension requiring vasopressors or inotropes within 30 days of surgery (OR = 12.76, p<0.0138) were associated with mortality (AUC=0.86).

Conclusion:Mortality in patients after pancreatic resection is associated with the primary diagnosis of pancreatic adenocarcinoma, disease recurrence after primary resection, and a history of DM. The only independent predictor not associated with primary pancreatic disease and comorbidity is hypotension requiring vasopressors or inotropes within 30 days after resection. The risk of mortality after pancreatic resection can be decreased by active measures to prevent the occurrence of post-operative hypotension and by rapid treatment when it does occur.

69.15 Contaminated Cases: A Dual Stage Approach to High Risk Ventral Hernia Repair

Posted on January 18, 2016

M. Bobbs1, N. Kugler1, T. Webb1, J. S. Paul1 1Medical College Of Wisconsin,Milwaukee, WI, USA

Introduction: The Modified Hernia Grade System (MHGS) was developed to risk stratify complex ventral hernia repairs. MHGS Grade 3 patients have mesh infections, dirty or contaminated fields, and/or violation of the alimentary tract. Reported surgical site infection rates (SSI) are over 40% following single-stage ventral hernia repair in contaminated fields. We developed a dual-stage (DSVHR) approach for MHGS Grade 3 patients to decrease the SSI and recurrence rates.

Methods: Retrospective review of adult general surgery patients undergoing DSVHR between January 2010 and June 2014. All patients were MHGS Grade 3. Primary endpoint was 30-day superficial and deep SSI. Secondary endpoints included other surgical site occurrences (SSO), 6-month recurrence, and mesh excision rates.

Results:Fifteen patients underwent DSVHR. Mean age was 56 years and median BMI of 38.3 kg/m2. Operative indication included ECF (n=6), ECF with infected mesh (n=2), infected mesh (n=2), and ventral hernia repair requiring bowel resection (n=5). Thirty-one operative procedures were performed with median of 2.5 days between procedures. Midline was re-established in 12 patients; five patients had underlay biologic mesh placement; seven underwent component separation with retrorectus mesh placement [synthetic (n=2), biologic (n=5)]. The remaining patients underwent bridging repair with biologic mesh. One patient developed recurrence after 6 months. One patient had a recurrence of their ECF. Four (27%) patients developed a SSI, four (27%) had a SSO. There were no post-operative mesh infections.

Conclusion: DSVHR in MHGS Grade 3 patients is associated with a lower surgical site infection rate than previously reported for those undergoing single stage repairs.

69.16 Single Institution Phasix Mesh Outcomes in a Population of Primarily Complicated/Recurrent Hernias

Posted on January 18, 2016

D. B. LePere1, M. P. Lundgren2, E. L. Rosato2, K. A. Chojnacki2, D. A. Ehrlich1, S. E. Copit1, P. J. Greaney1 1Thomas Jefferson University,Plastic Surgery,Philadelphia, PA, USA 2Thomas Jefferson University,General Surgery,Philadelphia, PA, USA

Introduction: Phasix is a biologic, fully absorbable polymer recently fashioned into a monofilament mesh scaffold being utilized for ventral hernia repair. The mesh has been shown, in a porcine model, to maintain support of a hernia repair site over a 52 week period. Little literature is available regarding outcomes using Phasix mesh for hernia repair. Early data available includes multi-center data outlining outcomes in a population consisting of a majority of primary hernia repairs. A study was undertaken to evaluate the performance of Phasix in a patient population consisting primarily of complex and recurrent hernias.

Methods: A single-institutional study was undertaken at Thomas Jefferson University Hospital in Philadelphia, PA to review outcomes of 52 patients between September 2014 and July 2015 in a population consisting primarily of complicated and recurrent hernias. Patient data extracted included demographic information, BMI, past medical history, past surgical history, mesh dimensions, operative and hospital course data, complications, and follow-up, including readmission.

Results: Fifty-two patients were identified upon review of our records (27 males, 25 females), all of which received overlay Phasix mesh for hernia repair after primary fascial closure. Thirty-four of the patients were over the age of 50 (65.4%), with an average age of 56.7 years. The average BMI of the population was 32.1, with 5 patients < 25 kg/m2 (9.6%), 16 patients 25 – 29.9 kg/m2 (30.7%), 15 patients 30 – 34.9 kg/m2 (28.8%) and 16 patients > 35 kg/m2 (30.7%). Forty-seven of the patients had ventral hernias (90.4%), 43 were incisional (82.7%), 5 were umbilical (9.6%), and 2 were inguinal (3.8%). Thirty-one of the patients were being treated for recurrent hernias (59.6%). The average surgery lasted 4.46 hours, and the average stay was 3.9 days, with 8 patients requiring readmission (15.4%). Post-operatively, 4 patients' recoveries were complicated by seroma (7.7%), 5 by hematoma (9.6%), 8 by infection (15.4%), 1 by ileus (1.9%), 3 required removal of mesh (5.8%), and 2 experienced recurrent hernias (3.8%), both secondary to post-operative infection.

Conclusion: The outcomes of this data series suggests Phasix is a reliable mesh for complex and recurrent hernia repair with a low recurrence and complication rate.

69.17 Combined Endoscopic Laparoscopic Surgery for Benign Colon Polyps: A Single Institution Cost Analysis

Posted on January 18, 2016

L. Tedesco1, J. Paolino1, L. Chen1, J. Yoo1, M. Kiely1 1Tufts Medical Center,Boston, MA, USA

Introduction: Endoscopic removal of benign colon polyps is not always possible, even with advanced endoscopic techniques. While segmental colectomy has been the traditional therapy for these polyps, this procedure is associated with an increased risk of complications and may be unnecessary given that fewer than 20% of these polyps harbor malignancy. Combined endoscopic laparoscopic surgery (CELS) has emerged as an alternative method to address these polyps. Here, endoscopic removal of complex polyps is facilitated by laparoscopic manipulation of the bowel wall, thereby avoiding a bowel resection and its associated risks. While feasibility, safety, and improved short term patient outcomes have been demonstrated, there has never been an evaluation of cost comparing these two approaches within a single institution.

Methods: In this observational cohort study, we examined five patients who underwent CELS for right colon polyps at a large tertiary care center (Tufts Medical Center, TMC) between April 2014 and July 2015. We compared the cost of CELS to that of ten patients who underwent a traditional laparoscopic right colectomy performed for a colon polyp or cancer during the study period. The cost analysis covered the perioperative period from operating room to hospital discharge. Microcosting estimates were obtained from the literature and applied to our specific operating room time and length of stay data for our cohorts. A modified societal perspective was used as the costing structure, with all cost estimates obtained for university hospital settings. The costing estimates were adjusted using the medical consumer price index to 2014 US dollars.

Results:A total of five patients were taken to the operating room with the intention of performing CELS for right colon polyps. Four out of the five patients successfully underwent CELS surgery. The median length of stay (LOS) for all successful CELS patients was 1 day. LOS for patients who underwent a laparoscopic right colectomy at TMC over the same time period was 4.5 days. The median OR time for successful CELS was 157.5 minutes, compared to 141.5 minutes for a laparoscopic right colectomy. The calculated total cost for a CELS patient was $7,513.10, compared to $14,600.50 for a laparoscopic right colectomy, a 49% cost-savings almost entirely attributable to a difference in length of stay.

Conclusion:In the perioperative period, successful CELS procedures are performed at a lower cost compared to traditional laparoscopic colectomy, with the most significant cost saver being shorter hospital length of stay. This is the first study to directly compare the cost of successful CELS to traditional laparoscopic colectomy in the surgical management of benign colon polyps within a single institution. Future studies will evaluate both short and long term costs using intention to treat analysis and a larger sample size.

69.10 Feeding Gastrostomy in Children with Complex Heart Disease: When is a Fundoplication Indicated?

Posted on January 18, 2016

J. L. Carpenter1, T. A. Soeken1, A. J. Correa1, I. J. Zamora1, S. C. Fallon1, M. J. Kissler1, C. D. Fraser2, D. E. Wesson1 1Texas Children’s Hospital,Division Of Pediatric Surgery, Department Of Surgery,Houston, TX, USA 2Texas Children’s Hospital,Division Of Congenital Heart Surgery, Department Of Surgery,Houston, TX, USA

Introduction: Malnutrition is common among children with complex heart disease (CHD) presumably due to a combination of sub-optimal intake and increased energy expenditure. Feeding gastrostomies are often used to improve the nutritional status of such patients. Many also have evidence of gastro-esophageal reflux (GER), which raises the question of the need for a fundoplication at the time of the gastrostomy insertion. When a gastrostomy is indicated, we generally use the open Stamm method to avoid abdominal CO2 insufflation that may affect venous return to the heart and cause hypercapnia and, because GER tends to resolve in infants and young children, we rarely perform a fundoplication at the time of gastrostomy preferring to use gastro-jejunal (GJ) feeding instead for patients with clinically significant GER. The purpose of this study was to evaluate the subsequent course of a cohort of infants and children with CHD following open Stamm gastrostomy without fundoplication.

Methods: We reviewed our experience with all CHD patients who underwent feeding gastrostomy placement from January 1, 2004 to April 7, 2015. We examined demographic data, cardiac diagnoses, operative details, post-operative complications, and the need for GJ feeding or fundoplication.

Results: An open Stamm gastrostomy was performed in 111 patients. Five additional patients had a fundoplication at the time of gastrostomy placement. Median age at surgery was 37 weeks (range 3 weeks to 13.7 years); average weight was 5.3 ±4.9 kg. Forty-eight patients (43%) had single ventricle physiology, 50 (45%) had biventricular physiology, and 13 (12%) had cardiomyopathy. Thirty-four (30%) patients experienced a total of 37 minor complications, which included tube dislodgement after maturation of the stoma (20), superficial surgical site infection (13), mechanical failure (3), and bleeding (1). Three (3%) patients experienced a major complication defined as a need for return to the OR or peri-operative death (<30 days). Only three (3%) patients required a subsequent fundoplication. At last follow-up (median 7.2 months, range 2 days-6 years), 56 of the 90 (62%) surviving patients continue to receive gastrostomy feeds, of those 7 (13%) patients continue to require GJ feeds.

Conclusion: Children with CHD tolerate an open Stamm gastrostomy well with minimal major complications. GER in CHD patients frequently resolves over time and patients seldom require a subsequent fundoplication. Our experience supports very selective use of fundoplication in infants and children with CHD who require a feeding gastrostomy.

69.11 Stage IV Rectal Cancer: Different Than Colon? A Population Study

Posted on January 18, 2016

M. Ostrowski1, M. Mora-Pinzon2, E. He2, J. Eberhardt2, T. Saclarides2, D. Hayden2 1Loyola University Chicago,Stritch College Of Medicine,Maywood, IL, USA 2Loyola University Medical Center,Department Of Surgery,Maywood, IL, USA

Purpose:
Besides the great difference in the treatment algorithm between colon and rectal cancers, some studies suggest that rectal cancer may affect different populations than colon cancer. We will examine the variation in demographic and socioeconomic factors using a population database.

Methods:
Secondary analysis of the 2011 Nationwide Inpatient Sample (NIS) was performed. Hospital discharges were identified using ICD-9 and ICD-9-CM codes for liver metastases, colon and rectal cancer and colon and rectal resections.

Results:
There were 3758 admissions for colorectal cancer metastatic to the liver in 2011. The mean age was 62 years (22-102) and the majority of patients were male (56.5%). Patients with stage IV colorectal cancer were 62.7% white, 15.5% black, 10.2% Hispanic, 2.4% Asian/Pacific Islander and 2.2% other. During these admissions, mean length of stay was 5.8 days (0-50). 31.5% (1182) of these admissions were for patients with a primary diagnosis of liver metastasis and a secondary diagnosis of rectal cancer. Patients with stage IV rectal cancer were younger than those with colon (mean age 58.9 vs. 63.6, p=0.000) and seemed more likely to be male, but the difference was not significant (58.2% vs. 55.9% male, p=0.182). Rectal cancer patients were much more likely to be white and less likely to be black than stage IV colon cancer patients (70.0% vs. 59.4%, and 8.8% vs. 18.5%, p=0.000, respectively). They were less likely to live in large metropolitan cities (p=0.000) and more likely to come from zip codes with higher income quartiles (p=0.001). Stage IV rectal cancer patients were also more likely to have private insurance and less likely to have Medicare or Medicaid when compared to colon cancer patients. Admissions were less likely to be urgent or emergent for metastatic rectal cancer (p=0.000) and the mean length of stay was slightly but significantly shorter (5.5 vs. 5.9 days, p=0.042). The majority of these patients did not have surgery during the admission (67.4%).

Conclusion:
Our study show that stage IV rectal cancer patients have varying demographics when compared to colon cancer patients. They are younger, more likely to be white, come from higher socioeconomic status and more likely to have private insurance. Their admissions are more likely to be elective. These findings suggest that patient characteristics and environmental factors may also differentiate colon and rectal cancer when stage IV disease is present and should be considered when establishing practice parameters.

69.12 Axillary Hidradenitis Suppurativa: Natural History and Factors Associated with Recurrence

Posted on January 18, 2016

N. Nweze1, R. Ahuja1, A. Parsikia1, L. Force1, L. Cetrulo1, J. Strain1, A. R. Joshi1,2 1Einstein Healthcare Network,Surgery,Philadelphia, PA, USA 2Jefferson Medical College,Surgery,Philadelphia, PA, USA

Introduction:

Hidradenitis suppurativa (HS) is a chronic, inflammatory cutaneous disease of the apocrine sweat glands, most commonly in the axilla. Surgical treatments for HS include incision and drainage or surgical excision with or without closure.

Methods:

We reviewed 214 patients (with 293 surgical procedures) with axillary HS who underwent surgical intervention. We divided the cohort into two groups based on recurrence, and then compared them.

Results:

The mean age of onset of HS was 32.5 years. Females comprised 84.1% of the cohort. Average BMI was 33.37. 55% of patients smoked. 19.2% of patients had diabetes. 1.9% of patients had a positive MRSA history. 35.5% of patients had bilateral axillary disease, and 9.3% had extra-axillary disease. The vast majority (93.5%) of interventions included excision of the diseased area, with a median area of excision of 78cm2. 45.4% of these wounds were closed, 48.8% were allowed to heal by secondary intention, and 4.8% were treated with a vacuum device. 22/214 (10.3%) of patients had recurrent disease. Closure vs. non-closure was not associated with a proclivity towards recurrence. Recurrence after surgical treatment was significantly associated with extra-axillary disease (p=0.039).

Conclusion:

Patients with HS were overwhelmingly obese and female, and many of them smoked. Surgical excision was the most common intervention, and closure was employed in roughly half of patients. Recurrence after excision, albeit uncommon, was associated with extra-axillary disease, but not closure after excision. These characteristics may be helpful in counseling patients about risk of recurrence and overall prognosis.

69.14 Analysis of Accessibility, Usability, Reliability, and Accuracy of Online Burn Treatment Resources

Posted on January 18, 2016

P. H. Chang1,2, M. Paz3, P. H. Chang1,2 1Shriners Hospitals For Children-Boston,Boston, MA, USA 2Massachusetts General Hospital,Boston, MA, USA 3Northeastern University,Boston, MA, USA

Introduction:
The 2014 National Burn Repository Annual Report states that there were 191,848 reported cases of burn injuries in the United States last year. When injured, patients frequently use online searches to find out the initial treatment for their condition. In this study, 40 websites for burn treatment on two major search engines were identified. The content was analyzed on the accessibility, usability, reliability, and accuracy of the information.

Methods:
The phrases ‘burn treatment’, ‘burn first aid’, ‘burn care’, and ‘how to treat a burn’ were searched on Google and Bing on 2/11/2015 and 4/29/2015, and the top 10 results of each search was recorded. Out of the 40 articles recorded, there are a total of 24 unique articles for Google and 19 for Bing that were analyzed using the LIDA Instrument v.1.2 (Minervation, Ltd.). This instrument uses a series of questions to score the accessibility, usability, and reliability of health websites and then produces an overall rating for that website. Accessibility is defined as whether people can access your website, usability is how easy it is for users to find out the information they want to know, and reliability is whether the website keeps up to date with the best current knowledge. This instrument defines a good website as one that scores a 90% or higher, a moderate website has a score of 50-89%, and a bad website has a score below 50%. To analyze the accuracy of the websites a checklist was created using ABA referral criteria and expert opinion by a trained burn surgeon on treating burns.

Results:

Of the total 24 unique burn treatment websites identified in Google searches, four were from an ABA verified burn center or the ABA itself, the rest of the websites were from medical personnel or made by non medical consumer reports. The average accessibility score was 86% (63-98%), usability was 75% (42-92%), reliability was 64% (20-87%), and the average total score was 78% (62-91%). The average accuracy score for these websites was 55% (3-97%) , and only one of the websites was considered good.

The Bing search engine provided 19 different burn treatment websites. Only one was from an ABA verified burn center. The average accessibility score was 86% (59-98%), usability was 71% (42-92%), reliability was 62% (7-87%), and the average total score was 76% (41-91%).

Between both search engines there was an overlap of 12 websites. In total there were only 31 unique websites, out of these four (12.9%) were from an ABA verified burn center or the ABA itself. Consistently the top two searches were WebMD and Mayo Clinic. From the Google search there were three websites that were considered good, or had a total score of 0.90 or higher; the Bing search produced one good website.

Conclusion:
The majority of burn treatment information is posted online by non-burn surgical specialists. While most websites appear accessible the usability and reliability varies considerably.

69.08 Appendectomy in Patients with HIV and AIDS: Not as Bad as We Once Thought

Posted on January 18, 2016

M. C. Smith1, P. J. Chung1, Y. C. Constable2, A. E. Alfonso1, G. Sugiyama1 1SUNY Downstate Medical Center,Department Of Surgery,Brooklyn, NY, USA 2SUNY Downstate College Of Medicine,Brooklyn, NY, USA

Introduction: Although the incidence of human immunodeficiency virus (HIV) and acquired immune deficiency syndrome (AIDS) is decreasing, due to the effectiveness of antiretroviral therapy, the prevalence of HIV is increasing. As a result, patients are living longer and presenting with common surgical conditions such as acute appendicitis. Previous studies have examined single institution experiences, which have shown an increase in complications. We sought to investigate the effect of HIV and AIDS on postoperative outcomes in patients who undergo appendectomy.

Methods: We selected patients from the Nationwide Inpatient Sample between 2005 and 2012 who underwent either laparoscopic or open appendectomy for acute appendicitis. We excluded cases of interval appendectomy. We used ICD-9 codes to identify patients with HIV and AIDS, and to identify complications. Data on patient demographics, length of stay (LOS), total hospital charges, and mortality during admission were also extracted. Using multivariate logistic regression, we created statistical models that controlled for age, gender, race, insurance type, socioeconomic status, number of Elixhauser comorbidities, and presence of perforation.

Results: There were 821 patients with HIV, 422 patients with AIDS and 338,425 patients served as controls. On univariate analysis comparing patients with HIV to controls, mean LOS was higher (3.8 days vs 3.0 days, p<0.001) and mean total charge was higher ($33,350 vs $30,714, p<0.0001). Comparing patients with AIDS to controls, those with AIDS had higher mean LOS (5.0 days vs 3.0 days, p<0.001), higher mean total charge ($44,486 vs $30,704, p<0.0001), more mechanical complications (2.4% vs 0.6%, p<0.0001), and more postoperative complications (6.6% vs 3.1%, p<0.0001). Multivariate analysis showed that HIV status was not an independent risk factor for mortality or intraoperative and postoperative complications, but was associated with increased LOS (OR 1.12, 95% CI 1.08-1.17, p<0.0001). AIDS was an independent risk factor for postoperative infections (OR 2.10, 95% CI 1.37-3.21, p=0.0007), digestive complications (OR 1.57, 95% CI 1.01-2.43, p=0.046), increased LOS (OR 1.23, 95% CI 1.18-1.30, p<0.0001), and increased total charge (OR 1.18, 95% CI 1.11-1.26, p<0.0001).

Conclusion: In this large, retrospective analysis, we found that in patients with acute appendicitis who undergo appendectomy, HIV and AIDS were not observed to be independent risk factors for mortality. Though AIDS is an independent risk factor for an increased risk of postoperative infection and digestive complications, appendectomy is a safe procedure in this patient population. Further studies to investigate how to minimize these risks in the population with AIDS are warranted.

69.09 Unplanned Reoperations following Abdominal Surgery: Incidence and Failure-to-Rescue by Procedure

Posted on January 18, 2016

H. S. Kazaure1, S. A. Roman2, J. A. Sosa2 1Stanford University,General Surgery,Palo Alto, CA, USA 2Duke University Medical Center,Surgery,Durham, NC, USA

Introduction: Data on unplanned reoperation (UR) after complex abdominal surgery are limited. The incidence of UR in ≤30 postoperative days and its association with other adverse outcomes were analyzed.

Methods: Patients who underwent 9 groups of abdominal procedures captured in ACS-NSQIP (2012) were abstracted. URs and their association with subsequent complications and failure-to-rescue (FTR: case fatality after ≥1 complication) were analyzed using bivariate and multivariate methods.

Results:There were 71445 patients; 82.7% underwent non-emergent surgery. The UR rate was 7.0 % (median time to UR: 8 days); approximately 84.5%, 11.4% and 4.1% had 1, 2 and ≥3 URs, respectively. URs were more likely following esophageal operations (9.7%), proctectomies (6.7%), and small bowel resections (6.2%). Common indications for UR were bowel compromise, wound complications, and bleeding. Patients who underwent URs were more likely to experience subsequent complications (64.7% vs. 28.8%, p<0.001) and overall mortality (11.4% vs. 3.5%, p<0.001). FTR varied widely by procedure: it was highest for pancreas resections (20.7% vs. 3.7, p<0.001, for UR vs. no UR) and non-significant after exploratory laparotomies (17.6 vs. 21.5%, p=0.231 for UR vs. no UR). After multivariate adjustment, UR was associated incrementally with mortality (adjusted odds ratio: 2.2, 95% CI: 1.9 – 2.5 for 1 UR and 3.2, 95% CI: 2.5 – 4.1 for ≥2 URs).

Conclusion:1 in 14 patients undergoing complex abdominal surgery undergoes an UR in ≤30 postoperative days. These patients are more likely to experience subsequent complications and FTR. Our results suggest that UR may be an indicator for quality of care for certain non-emergent abdominal procedures.

69.05 Minimally Invasive Segmental Colectomy: Impact of Hospital Type on Short-Term and Oncologic Outcomes

Posted on January 18, 2016

U. P. Nag1, M. Adam1, C. T. Ong1, Z. Sun1, J. Kim1, J. Migaly1, C. Mantyh1 1Duke University Medical Center,Surgery,Durham, NC, USA

Introduction: While a minimally invasive approach is associated with improved short-term outcomes and equivalent oncologic outcomes, this may not be generalizable across hospital types. This study is designed to evaluate impact of hospital type on differences in short-term and oncologic outcomes between minimally invasive segmental colectomy (MIS) and open surgery.

Methods: Adult patients undergoing segmental colectomy for non-metastatic colon adenocarcinoma were identified from the 2010-2012 National Cancer Data Base. Descriptive statistics were used to characterize patterns of MIS (laparoscopic or robotic) vs. open colectomy by hospital type. Multivariable models were used to examine the effect of hospital type on short-term and oncologic outcomes from MIS vs. open surgery while adjusting for patient, clinical, and tumor characteristics.

Results: A total of 97,620 patients underwent segmental colectomy for cancer: 44,027 (45%) underwent MIS, and 53.593 (55%) underwent open surgery. Overall, 25% of cases were performed at academic, 60% comprehensive community, and 15% community centers. MIS was more often utilized at comprehensive community (61%) and academic centers (27%) than community centers (12%). Compared to open surgery, the MIS group was more likely to have a complete tumor resection, shorter length of stay, lower 30-day mortality, and better overall survival regardless of institution type (all p < 0.01). Length of stay was significantly shorter by 1 day in the MIS vs. open group across all hospital types (all p<0.01), without increased 30-day readmissions. MIS was significantly associated with lower 30-day mortality, which was similar across hospital types: academic [OR 0.51 (CI 0.41-0.64), p<0.01], comprehensive community [OR 0.56 (CI 0.49-0.63), p<0.01], and community [OR 0.45 (CI 0.34-0.61), p<0.01]. Overall survival benefit was also comparable throughout hospital types: academic [hazard ratio (HR) 0.71 (CI 0.69-0.76, p<0.01], comprehensive community [HR 0.73 (CI 0.69-0.78), p<0.01], and community [HR 0.72 (CI 0.63-0.810, p<0.01].

Conclusion: Minimally invasive segmental colectomy is associated with superior perioperative outcomes and lower 30-day mortality. Although hospital type may impact outcomes of other procedures, the benefits of minimally invasive techniques were observed across hospital types. Utilization of laparoscopy and robotic surgery may be variable across hospitals, but wider dissemination of minimally invasive techniques should be emphasized for colon cancer.

69.06 The Impact of Frailty on Outcomes Following Paraesophageal Hernia Repair Using NSQIP Data

Posted on January 18, 2016

M. Chimukangara1, M. J. Frelich1, M. Bosler1, L. E. Reinb2, A. Szabo2, J. C. Gould1 1Medical College Of Wisconsin,General Surgery,Milwaukee, WI, USA 2Medical College Of Wisconsin,Biostatistics,Milwaukee, WI, USA

Introduction: Frailty is a standardized measure of physiologic reserve that has been used to predict morbidity and mortality following surgical procedures in the elderly. As a state of increased vulnerability to adverse outcomes, frailty is commonly associated with decreased reserves in multiple organ systems, such as declining cognition, physical ability, and health. We hypothesized that frailty, as assessed based on data derived from a large clinical database, would be associated with morbidity and mortality following paraesophageal hernia repair (PEH).

Methods: The National Surgical Quality Improvement Program (NSQIP) database was queried for ICD-9 and CPT codes associated with PEH repair. The NSQIP participant use files for the years 2011-2013 were utilized. Only subjects ≥ 60 years old were included. A previously described modified frailty index (mFI), based on 11 clinical variables available in the NSQIP dataset and derived from the model of cumulative deficits, was used to quantify frailty. Outcomes were 30-day mortality and the occurrence of post-operative 30-day complications. The Clavien-Dindo Classification system was used to characterize complication severity. Secondary outcomes were discharge destination and readmission. Multivariate logistic regression was used to determine the relationship between frailty, complications, and mortality.

Results: Of the 4434 PEH repairs performed in patients ≥ 60 years old in the study interval, 885 records were included in the final analysis (20%). Excluded patients were missing 1 or more variables in the 11-point mFI. The overall rate of complications that were Clavien-Dindo Grade ≥ 3 (Grade 3 = requiring endoscopic, radiographic, or surgical intervention) were 6.1%. Mortality was 0.9% (30-day). The overall readmission rate was 8.2% (30-day) and 10.9% of patients were discharged to a facility other than home. Relative to mFI scores of 0, 1, 2, and ≥3, the respective percentages for the four outcomes were as follows; Clavien-Dindo Grade ≥3 complication: 3.2%, 4.7%, 9.8%, and 23.3% (p <0.0001); mortality: 0.0%, 0.9%, 1.8%, and 2.3% (p 0.0974); discharge to facility other than home: 4.4%, 10.9%, 15.7%, and 31.7% (p <0.0001); and readmission: 8.9%, 6.8%, 8.5%, and 16.3% (p 0.1703). Grade ≥3 complications and discharge to a facility other than home were significantly correlated with mFI.

Conclusion: Frailty, as assessed by the mFI, is correlated with postoperative complications and discharge to a facility other than home following paraesophageal hernia repair. Due to many missing variables needed to calculate an 11-item mFI in the NSQIP dataset, significant portions of otherwise eligible patients were excluded from this analysis. The mFI may not be the ideal measure to assess frailty using the NSQIP dataset for this reason. Future investigation is needed to better quantify frailty based on the clinical variables contained in the NSQIP dataset.

69.07 Marginal Ulcer After Roux-en-Y Gastric Bypass: A Common Costly Problem

Posted on January 18, 2016

A. Furukawa1, D. T. Dempsey1, N. Williams1, C. Neylan1, K. Dumon1 1Hospital Of The University Of Pennsylvania,Department Of Surgery,Philadelphia, PA, USA

Introduction: Marginal ulcer (MU) is a known complication of Roux-en-Y gastric bypass (RYGBP) but few studies have looked at long-term incidence in this patient population. The purpose of this study is to better understand the prevalence, risk factors, complications, and cost of symptomatic MU in RYGBP patients in patients followed >2.5 years.

Methods: With IRB approval, all patients having RYGBP at our institution from 2003-2012 were reviewed. All patients who developed symptomatic marginal ulcers were analyzed (MU group=166). A randomly selected cohort of RYGBP patients without MU were selected and analyzed for comparison (non MU group=150) Charts were retrospectively reviewed for ulcer risk factors, ulcer complications, and charges (hospital and physician). Here, postoperative upper endoscopy (EGD) was used as a surrogate for utilization of healthcare resources. All statistical analysis including multivariate regression was performed Stata/IC 14.

Results: Over a 10 year span, there were 2,019 patients who underwent RYGBP. 166 developed MU (8% incidence of MU). Of those 166 patients, 135 (81%) were conservatively managed with medication and 31 (19%) needed ulcer operation. Of these 31 patients, 19 had emergent operations due to perforation and 12 had elective operations. Men were more likely to need emergency operation (p<0.05). On multivariate analysis, smoking and diabetes significantly increased the risk of MU (p<0.05), but hypertension, GERD, male gender, increased age, or increasing BMI did not. MU patients had a total of 437 EGDs after RYGBP, significantly more than patients without marginal ulcers who had a total of 41 (p<.001).

Conclusion: Symptomatic MU is common after RYGBP and leads to increased utilization of healthcare resources. Though smoking cessation is already recommended, increased MU incidence in smokers highlights the importance of adherence in RYGBP patients. To our knowledge, the relationship between MU and diabetes has not been looked at. Long term acid suppression after RYGBP should be considered, especially in men, diabetics, and in patients with smoking history.

69.03 Location of Ambulatory Laparoscopic Cholecystectomy in the Elderly Impacts Postoperative Acute Care

Posted on January 18, 2016

G. Molina1,2,3, B. Neville1, S. R. Lipsitz1, W. R. Berry1, A. B. Haynes1,2,3 1Ariadne Labs,Brigham And Women’s Hospital And The Harvard T.H. Chan School Of Public Health,Boston, MA, USA 2Massachusetts General Hospital,Department Of Surgery,Boston, MA, USA 3Harvard Medical School,Boston, MA, USA

Introduction: Laparoscopic cholecystectomy is increasingly being performed in the ambulatory setting, including hospital and freestanding ambulatory surgery centers (ASCs). Previous research has demonstrated the safety of ambulatory laparoscopic cholecystectomy in the elderly. These studies, however, did not distinguish between hospital and freestanding ASCs.

Methods: We used logistic regression, clustering by facility to evaluate if ambulatory setting (hospital ASC versus freestanding ASC) was an independent predictor of acute care (ED visits or inpatient admissions) within 1 and 7 days after a laparoscopic cholecystectomy among patients 65 years or older in South Carolina. Due to low number of events, we used a machine learning technique (supervised principal components regression) to adjust for all available patient characteristics (gender, race, insurance, median household income, and preoperative Charlson comorbidity index score). This technique is an alternative to using all variables when adjusting by generating one covariate from a combination of the original variables. This new covariate retains much of the original information as possible in a single covariate. A sensitivity analysis evaluating ASCs that performed at least 25 laparoscopic cholecystectomies annually among all adult patients was also performed in order to evaluate if these findings would persist after taking annual volume into account.

Results: There were 6,299 elderly patients who underwent a same-day laparoscopic cholecystectomy in the ambulatory setting in South Carolina from 2006 to 2013. Of these, 5,819 (92.4%) patients underwent a laparoscopic cholecystectomy at 58 hospital ASCs compared to 480 (7.6%) patients who underwent this operation at 16 freestanding ASCs. Elderly patients who had laparoscopic cholecystectomy performed at a freestanding ASC had a higher odds of being admitted to the hospital within 1 day when compared to hospital ASCs (Adjusted Odds Ratio (OR)=10.4, 95% Confidence Interval (CI) 3.9–27.5) but not within 7 days (OR=1.9, 95% CI 0.8–4.2). There was no significant difference between hospital ASCs and freestanding ASCs when evaluating subsequent ED visits (within 1 or 7 days). The sensitivity analysis confirmed these results.

Conclusion: Elderly patients who had same-day laparoscopic cholecystectomy at freestanding ASCs in South Carolina were more likely to be subsequently admitted to the hospital within 1 day when compared to hospital ASCs. As more surgical procedures are performed in the ambulatory setting, preoperative screening is critical to identify the safest and most efficient ambulatory surgery setting. Further research is needed on how to minimize acute care need following laparoscopic cholecystectomy at freestanding ASCs among the elderly.

69.04 Do Hospital Factors Impact Readmissions After Colorectal Resections At Minority-Serving Hospitals?

Posted on January 18, 2016

E. M. Hechenbleikner1,2,3, C. Zheng1,2,3, S. P. Lawrence1,2,3, Y. K. Hong1,2,3, L. B. Johnson1,2,3, W. B. Al-Refaie1,2,3 1Georgetown University Medical Center,Washington, DC, USA 2Georgetown University Medical Center, MH-SORC,Washington, DC, USA 3Georgetown University Medical Center, MGUH,Washington, DC, USA

Introduction:

Minority-serving hospitals (MSH) reportedly have higher readmission rates after surgical procedures including colectomy; however, little is known about the contribution of hospital characteristics to the elevated risk of readmission. This study sought to explore the extent to which hospital factors drive readmissions after colorectal resections performed at MSH in the context of patient- and procedure-related factors.

Methods:

Over 168,500 patients who underwent colon or rectal resections in 374 California hospitals between 2004 and 2011 were analyzed, utilizing records from the State Inpatient Database and American Hospital Association Hospital Survey. Stepwise logistic models were built to determine the associations between MSH and 30-day, 90-day, and repeated readmissions, adjusting for patient, procedure, and hospital factors gradually. MSH were defined as hospitals with highest proportion (top decile) of Black and Hispanic patients.

Results:

Among all patients, 88.5% underwent colon resections, 85.8% had open procedures, and 17.7% ostomies. Overall 30-day, 90-day, and repeated readmission rates were 11.2%, 16.9%, and 2.9%, respectively. Odds for 30-day, 90-day and repeated readmissions after colorectal resections were 22%, 21% and 38% more likely at MSH vs. non-MSH, respectively (all p <0.01). Patient factors accounted for up to 74% percent of the observed increase in odds for readmission after colorectal resections at MSH; in contrast, hospital-level factors contributed only 20% after controlling for patient and procedure factors (Table 1).

Conclusion:

Patient-level factors appeared to dominate the increased readmission risk following colorectal resections performed at MSH while hospital factors were less contributory. These findings need to be further validated to shape quality improvement interventions to decrease readmissions.

68.19 Routine Chest Radiographs in the Post-Operative Management of Pectus Excavatum Bar Removal

Posted on January 18, 2016

A. S. Poola1, S. W. Sharp1, S. D. St. Peter1 1Children’s Mercy Hospital – University Of Missouri Kansas City,Pediatric Surgery,Kansas City, MO, USA

Introduction:
Since its description, the Nuss procedure has been implemented as an effective minimally invasive repair for children and adolescents with pectus excavatum. While there has been much published literature on pectus bar placement, there are a limited number of reports studying the corresponding bar removal procedure. Even less reported is the post-operative management following bar removal. Common practices have included obtaining a post-operative chest radiograph (CXR) despite the minimal risk of intra-thoracic complications associated with this procedure. The aim of this study is to review our experience with the bar removal procedure and with not obtaining routine CXRs following this operation.

Methods:
A single institution retrospective chart review was performed from 2000 to 2012. Patients aged 8-35 who underwent a pectus bar removal procedure were included. We assessed operative timing of bar placement and removal, length of procedure and post-operative radiograph findings, specifically looking at rate of pneumothoraces.

Results:
335 patients were identified in this study. Of these, 80 percent were male. The mean age of bar placement was 14 years with a standard deviation of 3 years while the mean age of bar removal was 16 years with a standard deviation of 3.5 years. The average time between bar placement and bar removal was 33 months (range: 11-110 months). Most of our patients had one bar placed at initial procedure while 8 percent had two bars placed during repair. Operative times varied but on average bars were removed in 30 minutes (range: 10-96 min). Intra-operatively, 8 percent of patients had ossification of their bars although there was no noticeable effect of this finding on operative time. Of our sample, 143 patients obtained a post-operative radiograph and of these 139 films revealed a pneumothorax. Only 1 patient obtained a chest tube for management of their post-operative pneumothorax. Only three patients were re-admitted and zero patients were hospitalized following their procedures.

Conclusion:
Despite the detection of early post-operative pneumothoraces following bar removal, we have seldom found the need to clinically intervene on these findings. This suggests that not obtaining routine imaging following bar removal may be a safe practice.

68.20 Weight regain following sleeve gastrectomy – a systematic review with narrative analysis

Posted on January 18, 2016

M. Lauti1, M. Kularatna2, A. G. Hill1, A. D. MacCormick1 1University Of Auckland,Auckland, -, New Zealand 2Middlemore Hospital, University Of Auckland,Department Of Surgery,Auckland, Auckland, New Zealand

Introduction:

Weight regain is a recognised problem after bariatric surgery and is associated with recurrence of obesity-related co-morbidities. Sleeve gastrectomy is one of the most commonly performed bariatric procedures but the definition, incidence and cause of this problem is poorly reported and poorly understood. We performed a systematic review to clarify these issues in patients following sleeve gastrectomy.

Methods:

A systematic review was performed using four electronic databases to locate articles reporting the definition, rate or cause of weight regain in patients at least two years from sleeve gastrectomy. Papers were excluded if the series was of non-primary sleeves, had follow-up outcomes of less than two years or were not reporting primary research.

Results:

After abstract screening, 66 full text papers were reviewed of which thirteen met the inclusion criteria. Five papers reported a definition, five papers reported a rate and nine papers proposed a cause for weight regain following sleeve gastrectomy.

Definitions for weight regain reported in the literature included a gain of 10kg from nadir weight or an increase in BMI of 5kg/m2. Rates of regain following sleeve gastrectomy ranged from 10.2% at two years to 75.6% at six years. Proposed causes for the regain included initial sleeve size, sleeve dilation, increased ghrelin levels, inadequate follow-up support and maladaptive lifestyle behaviours.

Conclusion:

Weight regain appears to be a common complication following sleeve gastrectomy that is variably defined, described and reported in the literature. We make a number of recommendations to improve the reporting of clinical series so this problem can be better understood.

69.02 Blunt Traumatic Aortic Injury in Adolescents: Do Open and Endovascular Modalities of Repair Compare?

Posted on January 18, 2016

A. Shah2, A. Ashfaq2, S. R. Money2, C. K. Zogg1, J. Fraser4, V. J. Davila2, J. M. Chang2, T. Oyetunji3, R. J. Fowl2, W. Stone2, A. B. Chapital2, A. H. Haider1 1Brigham And Women’s Hospital,Center For Surgery And Public Health, Harvard Medical School, Harvard T H Chan School Of Public Health,BOSTON, MA, USA 2Mayo Clinic In Arizona,Department Of Surgery,Phoenix, AZ, USA 3Children’s Mercy Hospital And Clinics,Department Of Surgery,Kansas City, MO, USA 4Phoenix Children’s Hospital,Department Of Surgery,Phoenix, AZ, USA

Introduction:
Despite their rarity in the pediatric population, traumatic aortic injuries can prove to be highly lethal. Endovascular aortic repair (EVAR) has been successfully employed in the management of adult blunt aortic trauma. However, its efficacy has not been elucidated in the pediatric population. The objective of this study was to gauge the burden of traumatic aortic injuries using a nationally representative sample and compare outcomes associated with reparative modalities in a nationally representative population of adolescent patients.

Methods:
Four years (2000/2003/2006/2009) of data from the Kids Inpatient Database were queried for adolescent (10-19y) patients with diagnoses of blunt traumatic injuries to the abdominal/thoracic aorta. Included patients were divided into those who underwent EVAR and open aortic repair (OAR). Differences in associated outcomes — mortality, morbidity (acute kidney injury, stroke, myocardial infarction, cardiac arrest, ischemic colitis, paraplegia), length of stay (LOS), and total hospital cost of care — were compared using multivariable logistic/linear (family gamma; link log) regression. Propensity-score quintiles were used to account for differences in patient-level factors; models were further risk-adjusted to account for potential confounding associated with hospital region, teaching status, and children’s-hospital status. They were weighted to provide national estimates and account for clustering of patients within hospitals.

Results:

A total of 161 records were identified, weighted to represent 244 admissions nationwide. Average age on presentation was 17.0 (±2.0y) with a male preponderance (75.6%). EVAR was attempted in 27.4% (n=52) of cases. Patients who underwent EVAR had higher median ISS compared to the OAR group (38 vs 33, p=0.002). Risk-adjusted odds of death among EVAR patients were significantly lower relative to patients managed using OAR (p=0.031). No differences were found for complications. Marginally significant differences were reported for both LOS and total cost of care (table). Relative to OAR patients, use of EVAR was associated with a risk-adjusted predicted mean difference of 2.6 additional days and a corresponding increase in cost of approximately $20,200.

Conclusion:
Despite higher ISS among adolescent patients undergoing EVAR procedures, risk-adjusted odds of death were significantly lower relative to OAR. Combined with a lack of variation in morbidity, the results suggest that EVAR is comparatively safe may even convey a survival advantage among adolescent patients. Ongoing research is needed to consider longer-term outcomes in a larger group of patients.

68.17 Pattern of Calcium and PTH Normalization at 12 Months Follow Up after Parathyroid Surgery

Posted on January 18, 2016

O. A. Lavryk1, A. E. Siperstein1 1Cleveland Clinic,Endocrine Surgery,Cleveland, OH, USA

Introduction: At 12 month after parathyroid surgery we expect cured patients to have same biochemical characteristics as healthy individuals. The aim of the current study was to compare patients’ characteristics at 12 months after neck exploration for primary sporadic hyperparathyroidism (1?HP) with the healthy controls.

Methods: 484 patients were analyzed, who underwent parathyroid neck surgery for 1?HP from 2000-2014. 74 healthy subjects were enrolled as a control group. Calcium (Ca) and parathyroid hormone (PTH) were collected before and after surgery. To assess the biochemical profile of patients on the graphical plots of Ca vs PTH were used to compare the 95 % confidence area of healthy patients to those after parathyroid surgery. Patients were supplemented with Ca and vitamin D postoperatively.

Results:Preoperatively patients with 1?HP had a Ca of 10.9 ± 0.5 (mean ± standard deviation (SD) mg/dL and PTH 124.4 ± 68.5 pg/dL vs controls of 9.2 ± 0.3 mg/dL and 34.4±13.4 pg/dL, respectively. On plots of Ca vs PTH, all 1?HP patients preoperatively had values outside the normal zone. Postoperatively at 12 months, 335 (69%) of patients returned within the normal zone. 13 (2.7%) had absolute elevation of Ca and PTH, showing continued disease. 2 (0.4%) patients had hypoparathyroidism, both after undergoing subtotal parathyroidectomy. 149 (27.9%) had Ca and PTH values outside the normal zone, but not falling into the above categories. There was no marked difference in patients with simple adenoma vs multiple gland disease.

Conclusion:At 12 moths follow up postoperatively, many patients with 1?HP fail to have their Ca and PTH fall within the 95% confidence zone for normal individuals. Thus, 5% are attributed to the confidence interval chosen. Although some may reflect the persistent disease, many patients have unclear an pattern,despite Ca and vitamin D supplementation. Longer follow-up might be needed for patients after parathyroid surgery to reestablish stabilization of biochemical profiles.

68.18 Regional Variations of In-Patient Costs after Pancreaticoduodenectomy

Posted on January 18, 2016

F. Ali1, A. Dua1, S. Desai2 1Medical College Of Wisconsin,General Surgery,Milwaukee, WI, USA 2Southern Illinois University,Vascular Surgery,Springfield, IL, USA

Introduction: This study aims to investigate regional variations of outcomes including length of hospital stay, mortality and in-patient costs associated with patients undergoing pancreaticoduodenectomy in the United States.

Methods: A retrospective analysis was completed using the Nationwide Inpatient Sample to select patients with pancreatic cancer that underwent pancreaticoduodenectomy between 2007 and 2011 in the United States. ICD-9 codes were used to identify disease and procedure. Demographics, inpatient mortality, length of hospital stay and in-patient costs across different regions of the United States were compared. A multinomial regression was completed to compare the outcomes amongst the various regions.

Results: A total of 27,267 pancreaticoduodenectomies for pancreatic cancer were performed in the United States during the five-year study period. There were no regional variations in demographics but in terms of comorbidities, COPD was significantly increased in the Midwest as compared to the Northeast (P<0.05), while renal failure was higher in the West compared to the Northeast (P<0.05). In terms of outcomes, length of stay was significantly higher in the South and West compared to the rest of the USA. The cost of care was also significantly higher in the west. Mortality rates were highest in the South (Table I).

Conclusion: Length of stay and mortality were highest in the South for patients undergoing Whipple procedures for pancreatic cancer whereas cost of care was highest in the West. There are no demographic variations within patient populations based on regions however patients in the West did have a higher incidence of renal failure which may contribute to the higher overall cost of in-patient care.

« Previous Page
Next Page »