11.11 Assuring Survival of Safety-Net Surgical Patients

Posted on January 18, 2016

H. A. Pitt1,2, A. J. Goldberg2, A. S. Pathak2, J. A. Shinefeld1, S. M. Hinkle1, S. O. Rogers2, V. J. DiSesa1,2, L. R. Kaiser1,2 1Temple University,Health System,Philadelpha, PA, USA 2Temple University,Department Of Surgery,Philadelpha, PA, USA

Introduction: Survival of surgical inpatients is a key quality metric. Patient, surgeon and system factors all contribute to inpatient mortality, and sophisticated risk adjustment is required to assess outcomes. When mortality of general surgical patients was determined to be high at a safety-net hospital (53% Medicaid/Medical Assistance), a comprehensive approach was undertaken to improve patient survival. The aim of this analysis is to demonstrate that risk-adjusted mortality can be improved at an essential hospital through implementation of numerous best practices.

Methods: General surgical service line mortality was measured in the University HealthSystem Consortium (UHC) database from January 2013 through March 2015. Ten best practices were sequentially undertaken to reduce observed (O) and/or increase expected (E) mortality. These quality efforts included a) recruitment of new surgeons, b) participation in ACS-NSQIP, c) hardwiring Surviving Sepsis elements, d) increasing the number of diagnostic codes submitted, e) standardizing documentation in Preoperative Anesthesia Testing, f) expanding Palliative Care consultations, g) implementing an aspiration prevention protocol, h) initiating 100% mortality review, i) adopting an Early Warning System to detect sepsis and j) enhancing patient selection and preparation for surgery. UHC mortality rank, O, E and O/E ratios as well as early deaths were compared with control charts for the 27-month analysis. Statistical significance was set at the p<0.05 level. Case Mix Index, a financial metric which correlates positively with patient severity and procedure complexity, was monitored over the same time period.

Results:UHC general surgery mortality rank improved from the bottom decile to the top quartile among 102 Academic Medical Centers. Mortality data by quarter (Q) are presented in the Table.

During this time, Case Mix Index increased from 2.48 in Q1 2013 to 2.91 in Q1 2015.

Conclusion:Risk-adjusted mortality and early deaths decreased significantly over 27 months in general surgery patients. During this time, patient and procedure complexity increased by 17 percent as measured by Case Mix Index. Systematic implementation of quality best practices assured surgical patient survival at a safety-net medical center.

11.09 Outcomes of Appendectomy Performed on Weekend or on the Next Day of Admission

Posted on January 18, 2016

Z. Al-Qurayshi1, E. Kandil1 1Tulane University School Of Medicine,Surgery,New Orleans, LA, USA

Introduction: Time from hospital admission to operative intervention has been suggested to be a crucial risk factor for a number of surgical interventions. In this study, we aim to compare the post-appendectomy outcomes for operations performed on the next day of admission or on weekend to same day and weekday operations respectively.

Methods: A cross-sectional study utilizing the Nationwide Inpatient Sample (NIS) database for 2004-2009. ICD-9 codes were used to identify all patients who underwent appendectomy for acute appendicitis.

Results: 341,376 discharge records were included. 55,485 (16.3%) patients had appendectomy on the next day of admission, while 70,701 (24.7%) patients had the operation on weekends. Next day operations were more likely to be associated with postoperative complications [OR: 1.22, 95%CI (1.14, 1.30), p<0.001]. A hospital stay of more than 3 days was also more common for next day interventions (p<0.001). Appendectomies performed on weekends had a higher risk of complications compared to other days [OR: 1.09, 95%CI (1.02, 1.17), p=0.009]. Teaching, and urban hospitals were more likely to perform the appendectomy on the next day of admission (p<0.05). Older patients (>65 years), females, Blacks and Hispanics, and those on Medicaid , all were at higher risk of next day intervention (p<0.001 each). The average cost of next day operations was higher compared to same day operations ($9,422.10±138.25 vs. $8,278.00±84.75, p<0.001).

Conclusion: Appendectomies performed on next day of admission or on weekend are associated with disadvantageous outcomes. Demographic and economic factors, besides the hospital attributes, place certain subpopulations at higher risk of next day appendectomies.

11.10 Effect of a Preoperative Decontamination Protocol on Surgical Site Infections in Elective Surgery

Posted on January 18, 2016

E. T. Vo1,2, C. N. Robinson1,2, D. M. Green1,2, B. L. Ehni1,2, P. Kougias1,2, A. Lara-Smalling2, N. Logan2, S. S. Awad1,2 1Baylor College Of Medicine,Department Of Surgery,Houston, TX, USA 2Michael E. DeBakey Veterans Affairs Medical Center,Department Of Surgery,Houston, TX, USA

Introduction: Surgical site infections (SSIs) are associated with an increase in postoperative length of stay, cost, unplanned readmissions, and mortality. Despite adherence to Surgical Care Improvement Project (SCIP) criteria, the rate of SSIs remains high. Therefore, several efforts have been directed towards preoperative strategies to reduce SSIs. We have previously demonstrated that a decontamination protocol using chlorhexidine gluconate (CHG) washcloths and intranasal povidone-iodine (PI) is effective in decreasing SSIs in patients undergoing elective orthopedic surgery. Our objective was to determine the impact of this decontamination protocol on SSIs in patients undergoing elective surgery across four surgical services.

Methods: A retrospective review of a prospectively maintained database was used to identify patients undergoing elective surgery from 2013 to 2015. The preoperative decontamination protocol consists of patients watching an educational video on decontamination at the preoperative visit and applying the CHG washcloths and oral rinse the night before and the morning of surgery, and the intranasal PI the morning of surgery. Participating services included general surgery (GS), neurosurgery (NS), orthopedic surgery (OS), and vascular surgery (VS). Widespread implementation of this protocol at our center began in 10/2014. Rates of SSI were captured through the Veterans Affairs Surgical Quality Improvement Program from 10/2013 to 6/2014 during the pre-intervention period and from 10/2014 to 6/2015 during the post-intervention period. Outcomes were compared by wound class (clean vs. clean contaminated) and by surgical specialty. During the entire study period, there were no differences in patient management or SCIP compliance. Univariate analysis was performed using chi-square.

Results: A total of 4952 cases were evaluated (pre=2529, post=2423), of which 1682 were OS (pre=805, post=877), 1483 GS (pre=737, post=746), 941 VS (pre=534, post=407), and 846 NS (pre=453 post=393). Clean cases totaled 4194 (pre=2125, post=2069) and clean contaminated cases totaled 758 (pre=404, post=354). Overall, the SSI rate was significantly lower in the intervention group (pre=1.6% vs. post=0.9%; P=0.03). By surgical specialty, there was a significant decrease in SSIs in OS (pre=1.4% vs. post=0.3%; P=0.02) and a trend towards lower SSI rates in GS (pre=2.3% vs. post=1.9%; P=0.56), VS (pre=1.5% vs. post=0.5%; P=0.14), and NS (pre=1.1% vs. post=0.8%; P=0.57). By wound class, there was a significant decrease in SSIs in clean cases (pre=1.1% vs. post=0.4%; P=0.01), and a trend towards lower SSI rates in clean contaminated cases (pre=4.2% vs. post=3.7%; P=0.71).

Conclusion: Our data demonstrates that widespread implementation of a preoperative decontamination protocol decreases SSIs among patients undergoing elective surgery, specifically for surgeries with a clean wound class. This protocol may be a preventative strategy for SSIs and warrants further study.

11.07 Incidence of Central Venous Port Complications and Associated Factors

Posted on January 18, 2016

K. M. Babbitt1, C. S. Gunasekera1, P. P. Parikh1, R. J. Markert1, M. B. Roelle1, M. C. McCarthy1 1Wright State University,Department Of Surgery,Dayton, OH, USA

Introduction:
The requirement for reliable central venous access for chemotherapy administration and repeated phlebotomy has led to an increase in the placement of central venous ports. We investigated the incidence of complications associated with port placement and other factors.

Methods:
A retrospective chart review of ports placed at a large community hospital over a five-year period was performed. Demographics, BMI, preoperative diagnosis, port site, site of venous access, duration of port period, stage/nature of port complication, and use of steroids, TPN, and chemotherapy were collected. Chi square, Mann-Whitney, and Fisher’s Exact tests were used to analyze the data (SPSS Statistics 23.0, IBM, Armonk, NY).

Results:
Of 289 total patients, 68.2% were female. The mean age, BMI, and port duration were 57.6±13.5 years, 30.1±8.5 kg/m², and 291.7±192.5 days, respectively. The port sites were well balanced between the right and left chest wall (51.9% and 48.1%). Right chest wall placement had double the complication incidence compared to left chest wall placement (16.7% vs. 7.9%, p=0.024). Complications most commonly associated with right chest wall placement were port pocket infection and catheter related blood stream infection. Complications did not differ by venous access (internal jugular vein=13.8% vs. subclavian=11.1%, p=0.49). Complication incidence was nearly three times greater in steroid users vs. non-users (22.2% vs. 8.3%, p=0.016). Complication incidence did not differ between patients given or not given chemotherapy (p=0.24) or TPN (p=0.69). Moreover, there was no relationship between age, gender, port site, and venous access site and those complications that occurred five or more times (port pocket infection, catheter-related infection, and malfunctioning ports).

Conclusion:
These results suggest that the placement of ports in the right chest wall and steroid use could lead to a higher incidence of complications. The application of these findings could decrease complications related to port placement and improve both patient care and outcomes.

11.06 Are Lipomatous Tumors Metabolically Active? The Impact of Tumor Resection on Diabetes

Posted on January 18, 2016

F. Fallahian1,4, A. Ardestani1, C. Raut1,2,3, A. Tavakkoli1,2, E. Sheu1,2 1Brigham And Women’s Hospital,Boston, MA, USA 2Harvard Medical School,Boston, MA, USA 3Dana Farber Cancer Insititute,Boston, MA, USA 4University Of Missouri-Kansas City School Of Medicine,Kansas City, MO, USA

Introduction: The metabolic and immunologic properties of adipose tissue are linked to the pathogenesis of type 2 diabetes mellitus. Lipomatous tumors, such as liposarcomas, are rare but can reach significant size. We hypothesized that some lipomatous tumors are metabolically active and can alter systemic glucose homeostasis.

Methods: We performed a retrospective study of patients who underwent surgical excision of a lipomatous tumor at a tertiary cancer referral center (2004-2015). We divided patients into non-diabetics, well-controlled diabetics (HbA1c < 7), and poorly-controlled diabetics (HbA1c ≥ 7). We compared patient demographics, tumor characteristics, and measures of glycemic control among these groups both before and after tumor resection.

Results: 203 patients underwent 235 operations for lipomatous tumors. No differences were observed in tumor characteristics in patients with and without diabetes. However, tumor characteristics differed significantly between the well-controlled and poorly-controlled diabetics (Table 1). Patients with poorly-controlled diabetes had larger tumors that were more likely to be malignant and well-differentiated. Interestingly, we identified seven patients whose diabetes significantly improved with tumor resection. Overall, in the poorly-controlled diabetic group, there was a significant improvement in random blood glucose (109 mg/dL vs. 176 mg/dL, p < 0.05), without an associated change in BMI or number of diabetes medications, following tumor resection.

Conclusion: Development of a lipomatous tumor alone does not lead to diabetes. There was an association, however, between larger, malignant tumors and poorly-controlled diabetes. In a subset of patients, tumor resection improved glycemic control, suggesting that selected lipomatous tumors may be metabolically active.

11.05 Frailty Predicts Postoperative Morbidity and Mortality after Colectomy for Ulcerative Colitis

Posted on January 18, 2016

E. Telemi1, O. Trofymenko1, R. Venkat1, V. Nfonsam1 1University Of Arizona,Surgery,Tucson, AZ, USA

Introduction: The rates of ulcerative colitis, an inflammatory bowel disease, have been on the rise in U.S. for last several decades. Colectomy can be performed when other treatment options cannot provide reasonable quality of life to patients with ulcerative colitis or if dysplastic changes are identified on colonoscopy. Frailty has been used to assess the risks of colectomy in patients with various diagnoses.

Methods: NSQIP cross-institutional database was used for this study. The database contains records of more than 139 variables from multiple surgery types around United States from 2005 until 2012. 650 patient records (n = 650) with a primary diagnosis of ulcerative colitis were identified and used in the study. 34.0% underwent laparoscopic (n = 221) and 66.00% open (n = 429) colectomies. Modified frailty index (mFI) is a previously described and validated 11-variable frailty measure used with NSQIP to assess frailty. Outcome measures included serious morbidity, overall morbidity, Clavien IV (requiring ICU), and Clavien V (mortality) complications.

Results: Median age was 47 years, and BMI was 25.2Kg/m2. 52.8% of patients were males. 43.8% of patients were assigned ASA Class 3 or higher. The median mFI was 0 (0 – 0.54) and median. As mFI increased from 0 (non-frail) to 0.36 and above, the overall morbidity and increased from 26.2% to 68.8% and serious morbidity increased from 16.6% to 68.8%, respectively. The Clavien IV complications rate increased from 3.8% to 56.3%. Mortality rate has increased from 0.2% to 6.3%. All results were statistically significant at p<0.01. On a multivariate analysis mFI was independent predictor of serious morbidity (Adjusted Odd Ratio (AOR): 16.9, p<0.05) and Clavien IV complication rates (AOR: 117.5, p<0.01), independent of age, sex, BMI, ASA category, preoperative albumin, type of colectomy (laparoscopic or open), emergency status of surgery, and wound class.

Conclusion: Morbidity and mortality risks after colectomy in patients can be predicted using frailty. This can help physicians and patients with ulcerative colitis better stratify risks while considering surgical treatment options for these patients

11.04 Does Concomitant Thyroidectomy Increase the Perioperative Complications of Parathyroidectomy?

Posted on January 18, 2016

C. M. Kiernan1, C. Schlegel1, S. Kavalukas1, C. Isom1, M. F. Peters1, C. C. Solorzano1 1Vanderbilt University Medical Center,Nashville, TN, USA

Introduction:
Concomitant thyroid pathology has been reported in 17-84% of patients with primary sporadic hyperparathyroidism (HPT). However, it remains unclear whether the perioperative risks of concomitant thyroidectomy are greater than those of parathyroidectomy alone. This study examines the frequency of coexisting thyroid disease, concomitant thyroidectomy rates and complications of patients who underwent parathyroidectomy for HPT.

Methods:
A retrospective review of prospectively collected data on 709 patients who underwent parathyroidectomy for HPT over a 5-year period at a high volume center was performed. Patients who underwent parathyroidectomy were compared to patients who underwent parathyroidectomy with a concomitant thyroid procedure (total thyroidectomy or thyroid lobectomy). Patients who underwent previous parathyroid or thyroid operations were excluded. Chi-square, fisher’s exact, student’s t-test and Wilcoxon rank-sum test were utilized to compare cohorts.

Results:

641 patients met inclusion criteria. 49% of patients had thyroid disease on preoperative ultrasound and 20% of such patients were deemed to require a concomitant thyroid procedure. 574 patients (90%) underwent parathyroidectomy alone and 67 patients (10%) underwent parathyroidectomy with a concomitant thyroidectomy. There were no differences in age, gender, ASA class, preoperative calcium, PTH, or vitamin D levels between groups. When compared to parathyroidectomy alone, parathyroidectomy with a concomitant thyroid procedure was associated with longer operative times (median 57 vs. 91mins, p<0.01), increased rate of overnight stay (17% vs. 69%, p <0.01), and increased rate of transient hypocalcemia (3% vs. 15%, p<0.01). There were no differences in the rate of postoperative emergency department visits (3% vs. 6%, p=0.15), readmissions (1% vs. 3%, p=0.09) or permanent hypoparathyroidism (0.5% vs. 0%, p=0.55). Overall, there were 5 operative failures, all occurred in the parathyroidectomy alone group. There were no postoperative hematomas or recurrent laryngeal nerve injuries in either group.

Conclusion:
In this study, parathyroidectomy with a concomitant thyroid procedure was associated with longer operative times, increased rate of overnight stay and increased transient hypocalcemia. However, a concomitant thyroid procedure during parathyroidectomy for HPT did not increase the risk of recurrent laryngeal nerve injury, postoperative hematoma, permanent hypoparathyroidism or immediate operative failure.

11.03 Gallstone Pancreatitis and Choledocholithiasis: Bilirubin Levels and Trends

Posted on January 18, 2016

D. D. Yeh1, P. Fagenholz1, N. Chokengarmwong1, Y. Chang1, K. Butler1, H. Kaafarani1, D. R. King1, M. DeMoya1, G. Velmahos1 1Massachusetts General Hospital,Trauma, Emergency Surgery, And Surgical Critical Care/Department Of Surgery/Harvard Medical School,Boston, MA, USA 2Massachusetts General Hospital,Department Of Medicine,Boston, MA, USA

Introduction: Gallstones escaping the gallbladder into the common bile duct (CBD) can cause complications such as pancreatitis or biliary stasis with risk of cholangitis. Evaluation for and treatment of retained CBD stones may be performed using pre- or post-operative endoscopic retrograde cholangiopancreatography (ERCP) or intra-operative cholangiogram (IOC). Ideally, these investigations should be limited to only cases with high pre-test probability of retained stones, as routine investigation may result in unnecessary procedures and prolonged hospitalization. We hypothesized that initial levels and trends in serum bilirubin levels are predictive of diagnostic yield of CBD investigations in patients presenting to the Emergency Department with evidence of gallstone pancreatitis (GP) or choledocholithiasis (CDL).

Methods: We performed a retrospective analysis of a prospectively collected registry of all patients undergoing cholecystectomy for GP or CDL by the Acute Care Surgery team at our academic, tertiary hospital from June 2010 to January 2014. Demographic, laboratory, radiologic, and operative data were collected. Patients were divided into groups according to diagnosis and whether or not they had retained CBD stones on ERCP or IOC. Summary statistics were used to describe continuous variables and compared with Wilcoxon rank sum, while proportions were calculated for categorical variables and compared with chi square. Statistical significance was defined as two-sided p<0.05.

Results: 64 patients underwent cholecystectomy for GP and 49 patients underwent cholecystectomy for CDL. Overall rate of retained CBD stones was low for GP and high for CDL (22% vs. 81%, p<0.001). For GP, the maximum total bilirubin (TB) and max direct bilirubin (DB) values were significantly different between patients with and without retained stones (Table). For CDL, initial TB, subsequent TB, max TB, initial DB, subsequent DB, and max DB were significantly different between groups (Table). Hospital and post-op LOS were not significantly different.

Conclusion: For GP and CDL, initial levels and trends in serum bilirubin levels are significantly different between patients with and without retained stones on CBD investigation. These labs may be useful in predicting if a CBD stone has already passed and whether or not CBD investigation will be worthwhile time, effort, and complication risk.

11.02 Trends in Surgical Management and Postoperative Outcomes of Emergency Surgery for Diverticulitis.

Posted on January 18, 2016

L. Durbak1, G. D. Kennedy1, E. H. Carchman1 1University Of Wisconsin,School Of Medicine And Public Health,Madison, WI, USA

Introduction: The risk of developing diverticulosis by the age of 80 is virtually 100%. Up to 20% of individuals with diverticulosis will require hospitalization with 20-50% of these patients requiring operative intervention, resulting in an estimated annual cost of $2.6 billion. This study aims to examine the postoperative outcomes in the setting of emergent surgical management of diverticulitis, and then try to examine if these changes correlated with changes in surgical practice.

Methods: Data was obtained from the ACS NSQIP PUFs from 2005 to 2013. Inclusion criteria were patients undergoing emer-gency surgery with a post-operative diagnosis of diverticula of the colon or diverticulitis with or without mention of hemorrhage. Multivariate regression models were developed using trends in pre-, intra- and postoperative variables identified through univariate analyses. The impact of practice variables were analyzed using multivariate regression models, adjusting for biases in surgical approach due to preoperative patient factors by including propensity scores for each practice variable.

Results:Preoperative comorbidities dyspnea, dependent functional status, ascites and >10% weight loss significantly decreased over the study period. Intraoperative variables ASA class 2 and clean/contaminated wound class significantly decreased over the study period. ASA class 3 and the dirty/infected wound class increased. Of the postoperative variables considered, length of stay, ventilator dependency, and renal insufficiency decreased while organ space surgical site infection and sepsis increased. Multivariate analyses of pre- and intraoperative variable as explanatory factors of postoperative complications suggested several significant relationships, notably ascites and dyspnea with worse outcomes and ASA classes 2 and 3 with better outcomes. The odds ratios for laparoscopy and stoma creation over time were statistically significant but very close to 1 (95% CI between 1.00-1.09).Laparoscopy was significantly associated with decreased odds of several surgical complications, regardless of propensity score adjustment. The stoma creation variable was associated with increased odds of surgical complications, but these associations disappeared after propensity score adjustment. The drain placement variable was only significantly associated with an increased odds of sepsis.

Conclusion: In conclusion, this analysis suggests that a lack of substantial change in surgical practices for patients presenting with diverticulitis in the emergency setting is associated with a lack of consistent improvement in patient outcomes. This research adds to the existing evidence of the benefits of laparoscopic management of diverticulitis in the emergency setting.

11.01 A Nomogram to Predict Perioperative Blood Transfusion Among Patients Undergoing Abdominal Surgery

Posted on January 18, 2016

Y. Kim1, F. Gani1, F. Bagante1, G. A. Margonis1, D. Wagner1, L. Xu1, S. Buttner1, J. O. Wasey2, S. M. Frank2, T. M. Pawlik1 1Johns Hopkins University School Of Medicine,Surgery,Baltimore, MD, USA 2Johns Hopkins University School Of Medicine,Anesthesiology And Critical Care Medicine,Baltimore, MD, USA

Introduction: Stratifying a patient’s risk for perioperative packed red blood cell (PRBC) transfusion when planning major abdominal surgery is of interest to both patients and providers. We sought to identify preoperative factors associated with receipt of PRBC to create a nomogram that predicts an individual’s risk of transfusion with major abdominal surgery.

Methods: A nomogram to predict receipt of perioperative transfusion was constructed using a cohort of patients who underwent hepato-pancreatico-biliary (HPB)(n=2,792) and colorectal (n=2,171) surgery between 2009-2014. Discrimination and calibration of the nomogram was tested using area-under-the-curve (AUC) receiving operator curves and calibration plots.

Results: Among 4,963 patients undergoing either a HPB (56.3%) or colorectal (43.7%) procedure, 1,549 received ≥1 unit of PRBC for a perioperative transfusion rate of 33.1%. On multivariable analysis, age ≥65years (OR=1.5), race (Black: OR=1.6, Asian: OR=1.9), male sex (OR=1.1), preoperative Hb ≤8g/dL (vs. >12g/dL: OR=27.5), preoperative INR>1.2 (OR=2.6), Charlson score>3 (OR=1.9), and procedure type (colon surgery, referent: minor hepatectomy OR=1.1, rectal surgery OR=1.4, major hepatectomy OR=1.7, distal pancreatectomy OR=2.1, whipple procedure OR=2.7) were associated with risk of transfusion (all P<0.05). A nomogram was constructed to predict receipt of transfusion using these variables (Figure). Discrimination and calibration of the nomogram revealed good predictive abilities (AUC 0.76). Bootstrap validation of model accuracy revealed minimal evidence of model overfit.

Conclusion: Independent preoperative variables were used to create a nomogram to predict the likelihood of PRBC transfusion. This nomogram may be useful in stratifying a patient’s risk of needing a blood transfusion around the time of major abdominal surgery.

« Previous Page