34.06 The Effect of Surgical Approach on Oncologic Outcomes in Rectal Cancer Surgery

Posted on December 22, 2014

E. F. Midura2, D. J. Hanseman2, R. S. Hoehn2, B. R. Davis2, D. E. Abbott2, S. A. Shah2, I. M. Paquette2  2University Of Cincinnati,General Surgery,Cincinnati, OH, USA

Introduction:  The oncologic safety of minimally invasive surgery for colon cancer has been well established, however the role for a minimally invasive approach to rectal cancer has yet to be fully defined. Though current evidence to support the use of laparoscopic and robotic approaches is limited, these approaches are being adopted broadly into clinical practice. We sought to describe national practice patterns in different surgical approaches and operative outcomes for rectal cancer in the US. 

Methods:  The 2010 National Cancer Database (NCDB) was queried for surgical cases of rectal cancer. Surgical approach was classified as open, laparoscopic, or robotic. Patient, tumor, and hospital characteristics were examined for variation in approach. Oncologic efficacy was studied by examining whether harvest of ≥ 12 lymph nodes (controlling for radiation use) and negative surgical margins were achieved. We used propensity-score matching to compare laparoscopic or robotic surgery to open surgery, while controlling for case-mix differences. 

Results: We identified 9,253 patients, of which 68.6% had open, 26.4% laparoscopic, and 5.0% robotic surgery. Patients who underwent a minimally invasive approach were more likely to have private insurance, higher income, and be operated on in higher volume, urban hospitals. Patients who underwent open operations were more likely to have elevated CEA levels, higher histologic grade and more advanced pathologic stage. Patients who had robotic surgery were more likely to receive preoperative radiation compared to other approaches (p = 0.01). In unadjusted analysis, patients who had a minimally invasive approach had a lower incidence of positive resection margin, a shorter length of stay and a lower readmission rate compared to open surgery, however there were no differences in lymph nodes harvested or 30-day mortality (Table 1). After propensity score matching on age, gender, radiation use, tumor grade, and pathologic T and N stage, the laparoscopic approach was associated with a 2.6% decrease in the incidence of positive margin when compared to open surgery (p = 0.02), whereas the robotic approach was not associated with a difference in margin status when compared to open surgery. 

Conclusion: Minimally invasive approaches for rectal cancer resections are more commonly performed in high volume, urban, academic centers on privately insured patients. Patients with more advanced tumors are being resected by an open approach. Examination of a matched cohort of patients indicates that the laparoscopic approach may lead to improvements in resection margin status, though longer follow-up will be needed to determine whether this translates into better long-term survival.

 

34.07 Multimodality Therapy Improves Survival in Resected Early Stage (IB-II) Gastric Cancer

Posted on December 22, 2014

J. Datta1, M. T. McMillan1, L. Ruffolo1, R. Mamtani2, J. A. Drebin1, D. L. Fraker1, G. C. Karakousis1, R. Roses1  2University Of Pennsylvania,Medicine (Oncology),Philadelphia, PA, USA 1University Of Pennsylvania,Surgery,Philadelphia, PA, USA

Introduction:  Multimodality therapy (MT) is a recommended component of treatment for early stage gastric adenocarcinoma (ESGA). Compliance with these guidelines, and the impact of MT on survival in ESGA has not been extensively explored. We examined (1) temporal trends in sequencing of MT, (2) factors associated with MT use, and (3) effect of MT receipt on overall survival (OS) in resected ESGA.

Methods:  The National Cancer Data Base was queried for stage IB-II GA patients undergoing gastrectomy (1998-2011). Multivariate models were developed to identify factors associated with adjuvant chemoradiotherapy (ACRT) or perioperative chemotherapy (PC) receipt and to compare risk-adjusted OS by treatment group.

Results: Of 7,357 resected ESGA patients (median age 68 years, 69.1% male), 50.6%, 25.5%, and 23.9% received surgery only (SO), PC, and ACRT, respectively. Utilization of MT rose consistently between 1998 and 2011, increasing by 42.4% (p<0.001). While ACRT use increased only modestly (12.0%–23.5%, p=0.02), receipt of PC increased dramatically (8.0%–38.8%, p<0.001). Predictors of ACRT receipt were multifactorial, but most strongly associated with age<56 years (OR 3.31, 95% CI 2.62-4.17) and non-proximal tumor location (OR 2.78, 95% CI 2.42-3.19). Proximal tumor location (OR 3.79, 95% CI 3.26-4.41) and AJCC clinical stage IIB (OR 2.42, 95% CI 1.99-2.92) were the strongest predictors of PC use. Younger, white, higher-income, and less comorbid patients were also significantly more likely to receive PC (all p<0.01). Hospital-based selection of MT varied significantly by geographic region and academic affiliation (all p≤0.01). Survival analyses included 1,275 patients with a minimum follow-up of 5 years. Median, 1-yr, and 5-yr survival was 44.8 months, 75.0%, and 46.0% respectively. In this cohort, median survival was significantly longer for patients selected to receive MT (i.e. ACRT or PC) compared with those undergoing SO (47.1 vs. 43.3 months; p<0.001). R1 resection (HR 2.08, 95% CI 1.56-2.76), pathologic lymph node positivity (HR 1.91, 95% CI 1.65-2.22), and tumor T-classification 3/4 (HR 1.75, 95% CI 1.50-2.04) were strongly predictive of worse risk-adjusted OS. On stage-stratified Cox regression analysis, utilization of MT was independently associated with improved OS in both stage IB and II GA (IB: HR 0.65; p=0.002; II: HR 0.73; p=0.003).

Conclusion: Adoption of MT in ESGA is steadily increasing nationally, but remains incomplete. Patient-, tumor-, and hospital-related factors influence selection of MT sequence. In ESGA, MT receipt is independently associated with improved survival compared with undergoing SO.

 

34.08 Routine Somatostatin Analogue Use Decreases Pancreatic Fistulas After Whipple: A Meta-Analysis

Posted on December 22, 2014

K. Mahendraraj1, R. S. Chamberlain1,2,3  1Saint Barnabas Medical Center,Department Of Surgery,Livingston, NJ, USA 2New Jersey Medical School,Department Of Surgery,Newark, NJ, USA 3Saint George’s University,Department Of Surgery,Grenada, Grenada, Grenada

Introduction:
Pancreatic fistulas remain one of the most common and detrimental complications following the Whipple operation (pancreaticoduodenectomy). The prophylactic use of somatostatin analogues (SA) for the prevention of pancreatic fistulas is controversial. This meta-analysis aimed to assess the effectiveness of SA in preventing pancreatic fistulas among the conflicting data from published randomized controlled trials.

Methods:
A comprehensive search of  PUBMED, Embase and both the Cochrane and NIH Clinical Trial Registries was completed using the keywords ‘somatostatin’, ‘octreotide’, ‘fistula’, and ‘randomized controlled trial (RCT)’. Citations of relevant review articles were examined. Data on patient recruitment, intervention and outcome were extracted from the included trials and analyzed. 42 full-text articles were identified in this manner, and 22 of these were excluded for lack of randomization, inadequate blinding and incomplete outcome data. Only RCTs which were completed and had an endpoint of ‘pancreatic fistula’ were included. The risk ratio (RR) was calculated with 95% confidence intervals.

Results:
20 RCTs involving 2,596 patients who underwent pancreaticoduodenectomy were identified. 1,312 patients (50.5%) were randomized to receive SA while 1,284 patients (49.5%) received either a placebo or no intervention. The incidence of pancreatic fistula was 32% lower in the SA group (RR 0.68, 95% CI 0.54-0.86;p=0.001). The proportion of these fistulas that were clinically significant is not clear. On subgroup analysis, there was no significant difference in perioperative mortality between the two groups (RR 0.80, 95% CI 0.56-1.16; p=0.24), but there was a 30% lower incidence of overall complications in the SA group (RR 0.70, 95% CI 0.60-0.82; p=0.03). Significant heterogeneity was found among the identified trials with regard to the definition of fistula, dosage of octreotide, starting time and duration of treatment.

Conclusion:
The use of SA following pancreaticoduodenectomy significantly reduces both pancreatic fistula and overall complication rates, with no effect on perioperative mortality. Available data suggests a role for routine prophylactic SA use to improve patient outcomes following major pancreatic resection. Further clarification of the effects of these drugs is required through additional large, adequately powered randomized controlled trials with low risk of bias.

34.09 The Role of Intraoperative Pathologic Assessment in the Surgical Management of DCIS

Posted on December 22, 2014

M. R. Decker1, H. B. Neuman1, A. Trentham-Dietz3, N. K. LoConte4, M. A. Smith3, R. S. Punglia2, C. C. Greeberg1, L. G. Wilke1  1University Of Wisconsin Hospital & Clinics,Department Of Surgery,Madison, WI, USA 2Dana Farber Cancer Institute,Radiation Oncology,Boston, MA, USA 3University Of Wisconsin School Of Medicine & Public Health,Population Health Sciences,Madison, WI, USA 4University Of Wisconsin Hospital & Clinics,Carbone Cancer Center,Madison, WI, USA

Introduction: The elevated number of repeat operations for the treatment of ductal carcinoma in situ (DCIS) is costly for patients and the medical community, financially and psychologically.  Intraoperative pathologic assessment of DCIS may lead to reduction in these additional surgeries. This study examines the relationship between intraoperative pathologic assessment and subsequent operations after a diagnosis of DCIS.

Methods: SEER-Medicare patients diagnosed with DCIS from 1999 to 2007 who underwent lumpectomy without axillary surgery, as their initial surgical procedure, were eligible.  All subsequent breast surgical procedures were identified.  Use of intraoperative pathology (frozen section or touch preparation) during the initial surgery was assessed.  Multivariable logistic regression was used to describe the relationship between the use of intraoperative pathologic assessment and any subsequent mastectomy or lumpectomy within 90 days of the initial operation. 

Results: Of 8,259 DCIS patients who underwent lumpectomy without axillary surgery, 3,510(43%) underwent a subsequent mastectomy or lumpectomy. Claims for intraoperative pathologic assessment were present for 2,172 (26%) patients.  On univariate analysis, patients with intraoperative pathology during their initial surgery were more likely to have additional breast surgery than patients without intraoperative pathology (28% vs 25%, p=0.009). However, multivariable analysis demonstrated that intraoperative pathologic assessment had no statistically significant relationship with ocurrance of subsequent breast surgery (Adjusted OR 1.06 (95%CI: 0.93-1.19), p = 0.387). Only tumor size >2cm (AOR 2.28 (95%CI: 1.99 ­-2.60), p<0.001), poorly differentiated tumor grade (AOR 1.36 (95%CI: 1.13 -1.63), p<0.001 ), and patient residence in a rural area (AOR 1.20 (95%CI: 1.01 to 1.43), p=0.034) were associated with greater likelihood of subsequent surgery. 

Conclusion: The use of intraoperative pathologic assessment during lumpectomy from 1999-2007 was not associated with a reduction in subsequent breast operations in women with DCIS. Surgery in 2014 for DCIS has not changed from 2000, as there are no novel intraoperative tools that have been developed or standardization of margin assessment implemented.  These results highlight a need to identify cost-effective tools and strategies to facilitate surgical decision making and reduce the number of subsequent operations for women with intraductal disease.

34.10 Small Cell Gastric Carcinoma is a Lethal Diagnosis: A Clinical Outcomes Study from the SEER Database

Posted on December 22, 2014

B. L. Siracuse1,4, K. Mahendraraj1, R. S. Chamberlain1,2,3  1Saint Barnabas Medical Center,Department Of Surgery,Livingston, NJ, USA 2New Jersey Medical School,Department Of Surgery,Newark, NJ, USA 3St. George’s University School Of Medicine,St. George’s, St. George’s, Grenada 4Duke University,Durham, NC, USA

Introduction:  Small cell gastric carcinoma (SCGC) is an aggressive neuroendocrine malignancy that comprises less than 0.1% of all gastric carcinomas. No large patient series exist and clinical information regarding SCGC is derived from limited case studies. This study sought to examine the demographic and clinical factors in a large cohort of SCGC patients in order to compare clinical outcomes of SCGC to the more common gastric adenocarcinoma (GA).

Methods:  Demographic and clinical data on 71,607 patients with gastric cancer was abstracted from the SEER database (1973-2010). 207 SCGC and 71,400 GA patients formed the study populations. Abstracted data was analyzed using Chi square tests, t-tests, and multivariate analysis. Kaplan-Meier analysis was used to compare long-term survival between the groups.

Results: SCGC comprised 0.3% of all gastric cancers identified. The mean age of SCGC and GA patients was similar (68±12 vs.70±13 years, p=0.03) and both cancers were more common in males (64.7% and 66.0%, p<0.001) and Caucasians (70.4% and 65.6%, p<0.001). SCGC was more often undifferentiated (66.9%vs.2.9%, p<0.001) and had more lymph node positivity (59.6% vs. 55.9%, p=0.01) and metastatic disease than GA (64.5% vs. 46.1%, p<0.001). SCGC had lower mean survival times than GA (1.00±1.78 vs. 2.02±3.91 years, p=0.03). Mean survival time for SCGC patients treated with radiation was inferior to GA, but SCGC patients benefited more from surgery than GA patients (2.20±0.61 vs. 1.23±0.04 years, p<0.001). Multivariate analysis identified tumor size greater than 2 cm (OR=2.1, CI=1.9-2.4), regional (OR=2.8, CI=2.6-3.0) or distant disease (OR=2.1, CI=1.9-2.4), lymph node positivity (OR=1.6, CI=1.4-1.8), undifferentiated grade (OR=1.3, CI=1.1-1.4), Caucasian race (OR=2.0, CI=1.8-2.2), and male gender (OR=1.2, CI=1.1-1.3) as independently associated with increased mortality for SCGC, p<0.001. A survival advantage for SCGC was seen in patients treated with surgery alone (OR=0.5, CI=0.4-0.6) or in combination with radiation (OR=0.23, CI=0.2-0.3), p<0.001.

Conclusion: SCGC is a rare and often lethal gastric malignancy that presents most often in Caucasian males in their seventh decade of life, with larger tumor size, more undifferentiated histology, greater lymph node positivity, and higher rates of metastatic disease than GA. The majority of SCGC was untreated presumably due to advanced disease, but when treatment was employed, surgical resection resulted in a greater survival advantage than similarly treated GA. The combination of surgery and radiation was associated with the longest survival compared to other treatment modalities for SCGC, and should therefore be considered in patients with operable disease.

35.01 Is Incisional Negative Pressure Wound Therapy Associated with Decreased Surgical Site Infections?

Posted on December 22, 2014

K. Chopra1, N. N. Semsarzadeh1, K. K. Tadisina1, J. Maddox1, D. P. Singh1  1University Of Maryland School Of Medicine,Division Of Plastic Surgery,Baltimore, MARYLAND, USA

Introduction:  Negative pressure therapy has been increasingly used to treat open wounds over the past two decades. More recently, studies have reported the use of negative pressure therapy over closed incisions to decrease surgical site occurrences including infection and dehiscence. To assess cumulative status of reported findings, we conducted a meta-analysis of the current literature to evaluate the effectiveness of incisional negative pressure wound therapy (iNPWT) in lowering the incidence of surgical site infections (SSIs) as compared to standard dressings. 

Methods:  PubMed, the Cochrane Library, and Cumulative Index to Nursing and Allied Health Literature (CINAHL) were searched through August 2014 for publications comparing iNPWT to standard incisional care. A meta-analysis including 4631 participants from 14 published studies was performed. A fixed effects meta-analytical model was used to assess between-study heterogeneity and effect size, and funnel plots were used to assess publication bias.

Results: The overall rates of surgical site infection in the incisional negative pressure wound therapy group and control groups were 6.61% and 9.36%, respectively. Individual Odds Ratios or relative SSI likelihood rates by incision site location were 56% (p = 0.01) for the abdomen, 37% (p = 0.002) for the chest, 19% (p = 0.0001) for the groin, and 55% for the lower extremity (p = 0.022). The use of iNPWT was found to decrease SSI rate by 56% across all incision site regions considered together (p < 0.0001). However, a sensitivity analysis of heterogeneity (i.e., sub-group analysis) resulted in the three groin area studies being dropped and a final result that yielded an odds ratio of 0.504 (p = 0.0001), indicating a 50% reduction in SSI rate with iNPWT relative to standard care. Additionally, 9 of the 14 studies reported dehiscence rates among the two groups. Overall rates of dehiscence in the iNPWT and control groups were 5.32% and 10.68%, respectively. Heterogeniety was very high (I2 = 84%) and consequently data were not considered for further analysis. However, the effect size, Odds Ratio = 0.5 (CI 0.30 to 0.85), was significant, suggesting a correlation between iNPWT and lower dehiscence.

Conclusion: The results of the quantitative meta-analysis suggest that iNPWT is an effective method for reducing SSIs. It also appears that iNPWT may be associated with a decreased incidence of dehiscence although the evidence is still inconclusive. 

 

35.02 Fateful or Fruitful? ICP Monitoring in Elderly Patients with TBI is Associated with Worse Outcomes

Posted on December 22, 2014

Q. N. Dang2, J. Simon2, J. Catino1,3, I. Puente1,3, F. Habib1,3, L. Zucker1, M. Bukur1,3  1Delray Regional Medical Center,Trauma And Surgical Critical Care,Delray Beach, FL, USA 2Larkin Community Hospital,Surgery,South Miami, FL, USA 3Broward General Hospital,Trauma And Surgical Critical Care,Fort Lauderdale, FL, USA

Introduction:  In an expanding elderly population, Traumatic Brain Injury (TBI) remains a significant cause of death and disability. Guidelines for management of TBI, according to the Brain Trauma Foundation (BTF) include intracranial pressure (ICP) monitoring. Whether or not ICP monitoring contributes to outcomes in the elderly patients with TBI has not been explored. 

Methods:  This is a retrospective study extracted from the National Trauma Database 2007-2008 Research Datasets. Patients were included if age > 55 and they met BTF indications for ICP monitoring. Patients that had non-survivable injuries (any body AIS = 6), were dead on arrival, had withdrawal of care, or LOS < 48 hours were excluded. Outcomes were then stratified based upon ICP monitoring. The primary outcomes were in-hospital mortality as well as favorable discharge (to home or rehab facility). Logistic regression was used to analyze the effect of ICP monitoring on outcomes. 

Results: A total of 4437 patients were included with 11.1% having an ICP monitor placed. Patients requiring an ICP monitor were younger overall, more likely to present hypertensive, had higher injury severity, and more likely to require operative intervention. Median GCS (3) and Head AIS (4) were similar between groups. Of those patients with ICP monitoring, overall mortality was significantly higher (table) and they were less likely to have favorable discharge status. Craniotomy itself was not associated with increased mortality (p = 0.450)

Conclusion: Our findings suggest that the use of ICP monitoring according to BTF Guidelines in elderly TBI patients does not provide outcomes superior to treatment without monitoring. The ideal group to benefit from ICP monitor placement remains to be elucidated. 

 

35.03 Increased Age Predicts Failure to Rescue

Posted on December 22, 2014

G. Barmparas1, J. Murry1, M. Martin2, D. A. Wiegmann3, K. R. Catchpole1, B. L. Gewertz1, E. Ley1  1Cedars-Sinai Medical Center,Division Of Acute Care Surgery And Surgical Critical Care / Department Of Surgery,Los Angeles, CA, USA 2Madigan Army Medical Center,Department Of Surgery,Tacoma, WA, USA 3University Of Wisconsin,Madison College Of Engineering,Madison, WI, USA

Introduction:   Failure to rescue (FTR), defined as any death following the development of in-hospital complications, has become an important quality measure. The purpose of this investigation was to examine whether older patients are at higher risk of FTR following traumatic injuries.

Methods:   National Trauma Databank (NTDB) datasets 2007-2011 were queried. Patients ≥ 16 years admitted to centers reporting ≥ 80% of AIS and/or ≥ 20% of comorbidities, with ≥ 200 subjects in the NTDB and who had any reported complication were reviewed. Those who survived (non-FTR) were compared to those who did not (FTR) using a forward logistic regression model.

Results: Of 3,313,117 eligible patients, 218,986 (6.6%) met inclusion criteria and had at least one complication reported. Of these, 201,358 (91.2%) survived their complication (non-FTR) and 17,628 (8.8%) died (FTR). A forward logistic regression identified 22 variables as predictors of FTR. Of those, age 65 to 89 years was the strongest predictor (AOR [95% CI]: 6.58 [6.11, 7.08], p<0.001), followed by the need for mechanical ventilation (AOR [95% CI]: 2.99 [2.81, 3.17], p<0.001) and ICU admission (AOR [95% CI]: 2.61 [2.40, 2.84], p<0.001). Using age group 16 to 45 years as the reference group, the adjusted risk for FTR increased with increasing age in a stepwise fashion [AOR [95% CI]: 1.94 [1.80, 2.09] for age 46 to 65 years, 6.78 [6.19, 7.42] for age 66 to 89 years and 27.58 [21.81, 34.87] for age ≥ 90 years]. The adjusted risk of FTR also increased in a stepwise fashion with increasing number of complications, reaching AOR (95% CI) of 2.25 (2.07, 2.45), p<0.001 for ≥ 4 complications.

Conclusion: The risk of failure to rescue increases with age and number of complications.  Strategies which track this quality measure to encourage early recognition and treatment of complications in the elderly are necessary.

35.04 The Impact of Preexisting Opioid Use on Injury Mechanism, Type, and Outcome

Posted on December 22, 2014

W. Wilson1, S. O’Mara1,2, J. Opalek2, U. Pandya1,2  1Ohio University,Heritage College Of Osteopathic Medicine,Athens, OH, USA 2Grant Medical Center,Trauma Services,Columbus, OH, USA

Introduction: The prevalence of prescription narcotic use in the U.S. is on the rise. Opioid use and its impact on the management of trauma patients has yet to be thoroughly studied. The aim of this study is to determine the prevalence of pre-injury opioid use and its influence on specific outcomes amongst the trauma patient population.  

Methods: A retrospective review of all trauma patients presenting to a Level I Trauma Center was performed from January 1, 2010 to December 31, 2010.  Patients who died within 24 hours of presentation and those with incomplete medication data were excluded.  Electronic medical record review of history and physical documentation and urine drug screen records were used to determine pre-injury opioid status.  Pre-existing narcotic use, demographic data, injury mechanism and severity, injury type, and outcome variables were analyzed.

Results:A total of 3953 patients met inclusion criteria.  Among our sample, 644 (16.3%) were positive for pre-injury opioid use. Patients in the pre-injury opioid group were older (48 years vs. 41 years) and more likely to be female (37.9% vs. 30.6%).  The mechanism of injury was more often falls (32.8% vs. 22.0%).  Patients on narcotics were more likely to be admitted (82.6% vs. 77.4 %) despite having overall lower injury severity.  Analysis of less severely injured patients (ISS < 15) found a significantly increased length of stay (3.7 days vs. 2.9 days) in the narcotics group.  Evaluation of injury type revealed that head injury, abdominal injury and lower extremity/pelvic injuries were predictive of increased length of stay in these patients. 

Conclusion:There is a considerable prevalence of pre-injury opioid use in the trauma population.  These patients have unique characteristics and causes of injury.   Pre injury opioid use is predictive of increased hospital admission rate and increased length of stay, with important ramifications for patient care and health care costs.
 

35.05 A Restrictive Transfusion Strategy is Safe in Patients with Isolated Traumatic Brain Injury

Posted on December 22, 2014

A. Nguyen2, D. Plurad1, A. Kaji3, S. Bricker1, A. Neville1, F. Bongard1, B. Putnam1, D. Kim1  1Harbor-UCLA Medical Center,Division Of Trauma/Acute Care Surgery/Surgical Critical Care,Torrance, CA, USA 2Harbor-UCLA Medical Center,Department Of Surgery,Torrance, CA, USA 3Harbor-UCLA Medical Center,Department Of Emergency Medicine,Torrance, C, USA

Introduction: The optimal transfusion threshold for patients with traumatic brain injury (TBI) is not well defined. The purpose of this study was to examine the impact of a liberal versus restrictive transfusion strategy (RTS) on outcomes in patients with TBI. We hypothesized that a RTS is not associated with mortality in TBI patients.

Methods: We performed a retrospective cohort analysis of adult patients with TBI over a 40-month period. Patients with an Abbreviated Injury Scale (AIS) ≥3 in 2 or more regions outside of the head and those patients who did not undergo transfusion were excluded. Liberal transfusion strategy (LTS) patients received packed red blood cells for a hemoglobin ≤10 mg/dL whereas as RTS patients were not transfused until their Hb was ≤7 mg/dL. Multivariate logistic regression analysis was performed to identify independent predictors of mortality.

Results: Of 103 patients, 61 patients (59%) underwent a LTS and 42 patients (41%) underwent a RTS. Both groups were similar in age, gender, injury severity, and head AIS scores. There was no difference in the number of patients with severe TBI between the RTS and LTS groups (50% vs. 46%, p=0.7). On unadjusted analysis, there was no difference in mortality (31% vs. 38%, p=0.5) or complications (20% vs. 18%, p=0.8) between groups. On multivariate logistic regression analysis, age (OR=1.03; 95%CI=1.00-1.05, p=0.02), head AIS (OR=2.4; 95%CI=1.2-5.1, p=0.02), and subarachnoid hemorrhage (OR=3.6; 95%CI=1.3-10.3, p=0.02) were the only independent predictors of mortality.

Conclusion: A restrictive transfusion strategy may be safe in patients following isolated TBI. Prospective multicenter studies are required before any definitive recommendations regarding a restrictive transfusion strategy can be set forth.

 

35.06 Outcomes of Supracondylar/Intercondylar Humerus Fractures in Adults

Posted on December 22, 2014

W. K. Roache1, A. Harris2  1Howard University College Of Medicine,Washington, DC, USA 2University Of Florida,Jacksonville,Gainesville, FL, USA

Introduction:
Distal humerus fractures in adults are rare (0.5-2% of all fractures) and are approximately 30% of all humeral fractures. Supracondylar/intercondylar fractures are even less common, having only a 0.31% incidence. These injuries often occur from high-energy trauma, particularly in the young adult, and often present as open fractures with complex fracture patterns. These fractures in adults can be debilitating and difficult injuries to treat and frequently require operative management to create a stable platform to allow for early range of motion (ROM).

Methods:
A retrospective analysis was completed on a consecutive series of skeletally mature adults treated at a Level I Trauma Center with a radiologically visible supracondylar/intercondylar humerus fracture during the period from Feb. 2006 to May 2013. Exclusion criteria were patients with still maturing epiphyseal plates, supracondylar humerus fractures that were not inter-articular, and patients treated non-operatively. Postoperative data such as date of union, status of infection, range of motion (ROM), further complications, subsequent surgeries, status of physical therapy or occupational therapy (PT/OT), and comorbidities was gathered from postoperative and clinic visit notes. Postoperative elbow ROM was measured by tracking the arc of the forearm from full extension to flexion.

Results:
High-energy mechanisms accounted for injuries in 81% of the cases treated. Using the AO/OTA classification, 42% had a C2 and 39% had a C3 fracture pattern, with 53% of the cases being open fractures and 50% of the cases being polytrauma. Operative management, however, is not without risks. Complications were seen in 57% of the cases, with the major issues being elbow stiffness (54% of all complications) and infection (17%), often related to a compromised soft tissue envelope. In cases of infection, all were associated with open fractures. In cases of post union stiff elbow, 54% were associated with an open fracture, while only 41% of functional elbow cases involved open fractures. However, all cases of frozen elbow involved open fractures. Mean time from injury to operative fixation was 28.8 hours sooner in cases resulting in functional elbows than stiff elbows. Of the patients who regained full ROM, 91% started aggressive PT/OT immediately after surgery.

Conclusion:
The combination of soft tissue damage and comminution may lead to arthrofibrosis and the formation of heterotopic bone. It appears clear however that the ability to regain functional range of motion or better is associated with early operative intervention and more importantly, immediate participation in therapy driven modalities. Aggressive physical/occupational therapy was extremely important in restoring ROM, while nearly all patients who achieved full ROM performed immediate therapy postoperatively. Secondary interventions (manipulation; HO excision) appear to prove beneficial in restoring functional motion if stiffness does occur.

32.04 Prognostic Relevance of Lymph Node Ratio and Total Lymph Node Count for Small Bowel Adenocarcinoma

Posted on December 22, 2014

T. Tran1, M. Dua1, G. Poultsides1, J. Norton1, B. Visser1  1Stanford University School Of Medicine,Surgery,Stanford, CA, USA

Introduction:  Nodal metastasis is a known prognostic factor for small bowel adenocarcinoma (SBA). Like many gastrointestinal malignancies, inadequate lymph node evaluation may adversely influence survival and lead to understaging. The objective of this study is to evaluate the number of lymph node (LN) that should be retrieved and the impact of lymph node ratio (LNR) on survival.   

Methods:  The Surveillance, Epidemiology, and End Results database was queried to identify patients diagnosed with SBA and treated with curative surgical resection from 1988 to 2010. Patients who did not undergo lymphadenectomy (Nx) or had distant metastases (M1) were excluded from our analysis. The greatest survival difference for duodenal and jejunoileal tumors was determined using cut-point analysis and maximum log-rank test χ2 statistic. Survival was estimated using Kaplan-Meier method and compared using log-rank test. Multivariate cox proportional hazard model was utilized to identify independent predictors of survival. 

Results: A total of 2773 patients underwent surgical resection with lymphadenectomy for SBA from 1988 to 2010. Duodenal and jejunoileal adenocarcinomas each consisted of 50% of all small intestine tumors (n=1387 and n=1386, respectively). There were 1371 patients (49.4%) with negative nodal metastases (N0), whereas 928 (33.5%) and 474 patients (17.1%) had N1 and N2 metastases, respectively. Median LN examined for duodenal and jejunoileal adenocarcinomas were 9 and 8, respectively. Cut-point analysis demonstrated that harvesting at least 9 for jejunoileal and 5 for duodenal tumors resulted in the greatest survival difference. However, there was a significant survival difference for each additional LN examined up to 11 for duodenal and 20 for jejunoileal tumors. Increasing LNR was associated with decreased overall median survival (LNR=0, 71 months; LNR 0-0.2, 35 months; LNR 0.21-0.4, 25 months, and LNR > 0.4, 16 months; p<0.001).  This inverse pattern of survival in relation to LNR was observed in the entire cohort (see figure, p<0.001) and each subsite. Multivariate analysis revealed extent of lymph node retrieval, T-stage, and lymph node positivity were independent predictors of survival.  

Conclusion: LNR has a profound impact on survival in SBA. In order to achieve adequate staging, we recommend retrieving a minimum of 5 and 9 LN for duodenal and jejunoileal tumors.   

 

32.05 Wider Sentinel Lymph Node Diameter Is A Predictor Of Positivity For Cutaneous Melanoma

Posted on December 22, 2014

J. S. Merkow1, A. Paniccia1, E. Jones1, T. Jones1, M. Hodges1, J. Byers1, K. Lewis1, R. Gonzales1, W. Robinson1, N. Kounalakis1, R. Stovall1, C. Gajdos1, N. Pearlman1, M. McCarter1  1University Of Colorado Denver,Aurora, CO, USA

Introduction: The prognostic implication of a positive SLN is powerful. There are multiple studies that have identified factors such as Breslow thickness, presence of ulceration, regression, and age as predictors of a positive SLN.  However, there are no studies to date that have specifically looked at the size of a SLN and its relationship to positivity.  THE PURPOSE of this study was to determine if lymph node size was indeed associated with a positive SLN.

Methods: Retrospective review of an institutional prospectively maintained database of patients undergoing SLNB for cutaneous melanoma between February 1995 to January 2013. The diameters of the largest sentinel lymph nodes were measured at the time of histopathologic analysis and the widest of the three diameters (length, width, or height) was used for the analysis. A nodal diameter of 1.5 cm, included in two interquartile ranges of maximal nodal size in both positive and negative SLN, was used as the cut off value for a multivariate logistic regression model.

Results: A thousand and seventeen patients underwent SLN biopsy. Of these, 826 (81%) had complete information on SLN measurements and were included in the final analysis. Negative SLN biopsy was obtained in 677 patients (82%) vs. 149 (18%) with positive SLN biopsy. Patients with positive SLN biopsy were younger with median age of 50 years (38-59) vs. 53 years (43-61, p=0.032) and had a deeper primary lesion of 2 mm (1.3-3.2) vs. 1.4 mm (1-2.2, p<0.001). Gender and other tumor characteristics including ulceration, mitosis, regression, lymphovascular invasion, type of melanoma, and location were not statistically different between the two groups. The largest SLN diameter was 1.9 cm (1.5-2.2) in the positive SLN group and 1.6 cm (1.2-2.2) in the negative SLN group (1.2-2.2, p=0.030). A SLN with a diameter wider than 1.5 cm was found to have an 87% increased odds of being positive for nodal metastasis after adjusting for age, sex, and depth of primary lesion (p=0.005).

Conclusion: A sentinel lymph node biopsy is considered the standard of care for melanoma. Here we demonstrated that a wider sentinel lymph node size is associated with significant increased odds for a positive biopsy result that is independent from the depth of primary lesion. Sentinel lymph nodes wider than 1.5 cm are more likely to be positive than not.

 

32.06 The Clinical Impact of Lung Biopsy in Children for Non-Metastasis-Related Indications

Posted on December 22, 2014

I. J. Zamora1, S. C. Fallon1, T. J. Vece2, J. Rama2, D. L. Palazzi2, J. A. Coss Bu2, M. Hicks3, M. E. Lopez1, J. G. Nuchtern1,2, M. DeGuzman2, R. P. Guillerman4, F. Sheikh1, P. Lau1, D. L. Cass1,2  1Texas Children’s Hospital, Baylor College Of Medicine,Michael E. DeBakey Department Of Surgery,Houston, TX, USA 2Texas Children’s Hospital, Baylor College Of Medicine,Department Of Pediatrics,Houston, TX, USA 3Texas Children’s Hospital, Baylor College Of Medicine,Department Of Pathology,Houston, TX, USA 4Texas Children’s Hospital, Baylor College Of Medicine,Department Of Radiology,Houston, TX, USA

Introduction:

Pediatric surgeons are consulted for parenchymal lung biopsy in complicated patients with respiratory and immunologic illnesses, and occasionally compromised healing abilities.  Biopsy poses risks, and results may be non-diagnostic. The purpose of this study was to assess the risk-to-benefit ratio of lung biopsy by determining the operative complications and impact on clinical management.

 

Methods:

A retrospective chart review was performed for all children who underwent diagnostic lung biopsy from 1/2009-12/2013. Patients who had blebectomy for spontaneous pneumothorax, or biopsy post lung transplant or for suspected cancer metastasis, were excluded.  The indication for biopsy was classified as: those with new primary lung disease of unclear cause; known immunologic condition or history of organ transplant; known cancer; or those with bone marrow transplant.  Biopsy results were classified as: (1)-new diagnosis, (2)-helpful, diagnosis confirmed or disease ruled-out, (3)-unhelpful, or (4)-insufficient.  The primary outcome assessed was whether the biopsy led to a change in management.  Secondary outcomes included biopsy-related complications, number of chest tube days, and whether the specimen was adequate or diagnostic. Statistical analysis included chi-square and ANOVA.

 

Results:

Eighty-three patients were identified (53% male, mean age 8.3±6.7 years). The majority (64%) underwent thoracoscopic biopsy; 19% had thoracotomy, and 17% had image-guided needle-based procedures.  Comparisons of outcomes based on biopsy indication are presented in the Table.  Overall, biopsy results were classified as: (1)-33(40%), (2)-41(49%), (3)-5(6%), (4)-4(5%), and led to a change in management in 33/83 (40%) children.  Postoperatively, 16 (19%) patients were transferred to a higher level of care, and 27 (33%) experienced complications, including air-leak (n=20, 19%), need for reoperation (n=5; 4%), or bleeding (n=2, 2%). Eleven (13%) patients died, one directly attributable to surgical complications.  Fifty-eight children (70%) had a chest tube postoperatively; 5 (6%) required additional tubes, and the median duration of chest tube use was 4 (range, 1-19) days.

 

Conclusion:

The risk-benefit ratio of lung biopsy in children with complicated respiratory and immunologic conditions appears to be favorable, in that clinically useful information is identified in nearly 90% of cases.  Nevertheless, the procedure carries risks that must be considered in each patient by a multidisciplinary care team. 

 

32.07 When is it Safe to Forgo Abdominal CT Scan in Blunt Injured Children?

Posted on December 22, 2014

S. N. Acker1, C. L. Stewart1, G. E. Roosevelt3, D. A. Partrick1, D. D. Bensard1,2  1Children’s Hospital Colorado,Pediatric Surgery,Aurora, CO, USA 2Denver Health Medical Center,Department Of Surgery,Aurora, CO, USA 3Denver Health Medical Center,Pediatric Emergency Medicine,Aurora, CO, USA

Introduction: CT scan is the gold standard to diagnose solid organ injury following blunt trauma.  However, the radiation risks associated with abdominal CT scan include a 2-3:1000 risk of cancer, with younger age correlating to higher malignancy risk. We hypothesized that there are patient specific factors, including GCS on presentation, pediatric age adjusted shock index (heart rate/systolic blood pressure) (SIPA), and mechanism of injury, that can help identify those patients with low-grade injury who can be treated safely without need for a CT diagnosis.  We have previously shown that SIPA accurately identifies severely injured children following blunt trauma and hypothesized that SIPA would help to identify children with severe solid organ injury at high risk of requiring intervention.

Methods: We performed a retrospective review of all children admitted to two pediatric trauma centers following blunt trauma with any grade liver or spleen injury from 1/09-12/13.  Data collected include SIPA and GCS on presentation, mechanism of injury, injury severity score (ISS), need for interventions including red blood cell (PRBC) transfusion or laparotomy, and outcomes such as hospital length of stay (LOS) and discharge disposition.  The Low Risk Group was defined as GCS 15 with normal SIPA on presentation, and injury attributable to a single, non-motorized, blunt force to the abdomen.  The Non-Low Risk Group did not meet these criteria.  

Results:101 out of 206 children met the low risk criteria.  Patients in the Low Risk Group were older than those in Non-Low Risk Group (median age 11 years vs 9 years, p=0.01), were more likely to be male (75% vs 60%; p=0.02) and have a lower ISS (median 9 versus 17; p<0.001). 

Conclusion:Children, who present to the emergency department following blunt abdominal trauma by a non-motorized force with a normal GCS and SIPA, are unlikely to have a solid organ injury that will require intervention.  When treating patients who meet these criteria, clinicians can have an open dialogue with a child’s parents regarding the necessity of abdominal CT and the potential yield, particularly with regard to whether the CT scan will lead to a change in patient management.  Moreover, in children with normal SIPA and GCS on presentation, our pilot study suggests that this group may be managed with a period of observation, imaging only for changes in clinical parameters, and recommendations for a short period of activity restriction following discharge thus, obviating the need for abdominal CT reducing cost and radiation exposure.

 

32.08 Suction Rectal Biopsy is Diagnostically Equivalent to Full Thickness Rectal Biopsy in Children

Posted on December 22, 2014

E. D. Muise1, S. Hardee2, R. A. Morotti2, R. A. Cowles1  1Yale University School Of Medicine,Pediatric Surgery,New Haven, CT, USA 2Yale University School Of Medicine,Pathology,New Haven, CT, USA

Introduction: Rectal biopsy with evaluation by surgical pathology is considered the most accurate test when evaluating infants and children for Hirschsprung’s Disease (HD). The biopsy site and the percentage of submucosa in the biopsy sample are felt to be important for establishing a pathologic diagnosis. The ability for certain biopsy techniques to obtain adequate tissue, especially in older children, has been questioned. While both suction (SRB) and full-thickness (FTRB) rectal biopsy are commonly employed, no studies have described and compared the quality of the tissue samples obtained using these two techniques. We hypothesized that SRB and FTRB yield tissue specimens of different size, but that these specimens are equally diagnostic after pathologic analysis.   

 

Methods: Records of all children who underwent rectal biopsy to evaluate for HD between January 2007 and July 2014 were reviewed. Demographic data and details of the specimen and pathologic analysis were obtained. Volume of specimen, depth of submucosa, %submucosa, and diagnostic accuracy were compared between biopsy techniques and the effect of age on biopsy quality was assessed. Data were analyzed by mixed effects models with covariate adjustment for age at biopsy, and Fisher’s exact test.

 

Results: Forty-seven children, 28 male and 19 female, underwent a total of 58 rectal biopsies, 45 SRB and 13 FTRB. Thirty-two biopsies were performed after 12 months of age, 26 before 12 months (range 0-141months). Volume of SRB specimens was significantly smaller than FTRB across all ages (14.8mm3+/- 7.8 vs 121.3mm3 +/- 13.8, p=0.0001). Percent submucosa did not differ significantly between SRB and FTRB specimens across all ages (63.8% +/- 2.7 vs 66.5%+/-4.3, p=0.58). Volume of SRB specimens did not differ significantly when performed before or after 12 months of age (11.3mm3 +/- 2.5 vs 21.0mm3 +/-3.9, p=0.28). SRB compared with FTRB demonstrated no significant differences in volume when performed before 12 months  (12.2mm3 +/-3.6 vs 58.9mm3 +/- 11.2, p=0.16), and significant differences when performed after 12 months (18.7mm3 +/-20.0 vs 150.5mm3 +/-21.9, p=0.003). However, percent submucosa in SRB and FTRB performed both before 12 months of age (64.9%+/-3.3 vs 74.8%+/-6.7, p=0.36) and after 12 months of age (62.6%+/-5.1 vs 64.9%+/-5.3, p=0.70) was not significantly different. The number of inadequate biopsies was not significantly different when performed before or after 12 months (p=0.41). The diagnostic accuracy for both SRB and FTRB was 100% and there were no complications.

 

Conclusions: Tissue specimens obtained by SRB are smaller than those obtained by FTRB, especially in older children. SRB and FTRB appear equivalent in their ability to provide adequate submucosa for analysis and both provide highly diagnostic tissue for evaluation. Differences in cost and patient satisfaction between these two rectal biopsy techniques should be studied to further define the best overall technique.

32.09 Prenatal Counseling: Does it Alleviate Parental Mental Distress?

Posted on December 22, 2014

C. Sanner1, E. Lima1, K. H. Parker1, A. Lavery2, E. Perrone3, C. L. Neece1, E. P. Tagge3  1Loma Linda University,Behavioral Medicine,Loma Linda, CA, USA 2Loma Linda University Children’s Hospital,Pediatrics,Loma Linda, CA, USA 3Loma Linda University Children’s Hospital,Pediatric Surgery/Surgery,Loma Linda, CA, USA

Introduction:
Prenatal counseling attempts to educate families about diagnoses, likely outcomes and potential management options for their unborn child. An assumed additional benefit is to provide psychosocial support to parents, thus ameliorating emotional distress. Many authors have stated that prenatal counseling limits stress on both parent and baby. However objective documentation of such stress reduction is difficult to find. Taking advantage of our large prospective database on NICU parental mental health factors, we sought to examine the impact of prenatal counseling on a variety of psychology parameters. 

Methods:
Data were abstracted from a large prospective data set created to examine the relationship between parental mental health and NICU infant outcomes. Parents of our Level III NICU infants completed an hour-long questionnaire which included basic demographic questions as well as a variety of validated psychological instruments including 1) The Brief COPE Inventory; 2) Parental Stressor Scale: NICU (PSS: NICU); 3) The Psychological Well-Being Scale; 4) The Pediatric Inventory for Parenting (PIP); 5) Center for Epidemiologic Studies-Depression Scale (CES-D); 6) Stanford Acute Stress Questionnaire.

Results:
Completed questionnaires from parents of 77 infants managed in our Level III NICU were analyzed. The majority of the families were low income, less than half of the mothers were married and the majority of parents disclosed that their pregnancy was not planned. The mean CES-D depression score indicated that large portion of parents endorsed clinical levels of depression. A large percentage of parents met the clinical cut-offs for the hallmark symptoms of Acute Stress Disorder: dissociation (9.5%), anxiety (66.7%), avoidance (23.8%), and re-experiencing (28.6%). Parents endorsed a mean stress level score of 2.52 on the 4-point PSS: NICU scale, indicating moderate to high levels of stress. When patients were compared (one way anova) between prenatal counseling and no prenatal counseling, there was no significant difference in the results of the Stanford Acute Stress Questionnaire, the CES-D score, none of the 4 subscales of the PIP, coping mechanisms as evaluated in the Brief Cope Inventory, or overall stress as measured by the PSS:NICU.

Conclusion:
Significant mental health issues exist among parents of NICU babies, particularly depression and high stress levels. Prenatal counseling was not seen to impact any of a variety of psychology evaluations. This data suggests that perinatologists and pediatric surgeons should consider referring families for psychology support services if there is significant emotional distress noted at their prenatal counseling visit.
 

32.10 Early Closure of Gastroschisis Correlates with Earlier Initiation of and Progression to Full Feeds

Posted on December 22, 2014

J. C. Harris1, J. Poirier1, D. Selip2, S. Pillai1, A. N. Shah1, C. Jackson3, B. Chiu1  1Rush University Medical Center,General Surgery,Chicago, IL, USA 2Rush University Medical Center Fetal And Neonatal Medicine Center,Pediatrics,Chicago, IL, USA 3Tufts Medical Center,Pediatric Surgery,Boston, MA, USA

Introduction: Gastroschisis is a congenital anomaly affecting 1-6/10,000 live births. Return of bowel function after closure and time to initiation of enteral feeds are highly variable.  Previous studies show initiation of enteral feeds within 7 days after closure decreases hospital stay and predicts improved outcomes.  This study aims to identify factors associated with initiation of earlier enteral feeds. We hypothesize that post-natal management strategies that minimize intestinal irritation can lead to earlier enteral feed initiation.

Methods: A retrospective review of patients with gastroschisis treated between 2005-2014 at a single institution was performed.  Data points included: mother’s age, ethnicity, simple versus complex gastroschisis (defined by necrosis, atresia, or volvulus), gestational age, gender, ventilator days, length of stay, mortality, days to closure, time to first and full enteral feeds. The data were analyzed using both Spearman’s rho and the Kruskal-Wallis rank sum test where appropriate, and p value <0.05 was considered significant.

Results:  The charts of 43 patients (24 males, 19 females) born with gastroschisis were reviewed.  7 out of 43 were classified as complex gastroschisis.  Average gestational age at delivery was 35 weeks (28-39 weeks).  Overall survival rate was 88% (38/43).  The mean hospitalization was 88 days (0-498 days). Average maternal age was 19.5 years.  Maternal race was 35% Hispanic, 33% Caucasian, 30% African American, and 2% American Indian.   Average days on the ventilator were 12.9 days.  5 patients were closed on day of life (DOL) 0.  Mean days to closure were 7.3 (0 to 85) days.   First feeds, on average began DOL 20.5, and full feeds achieved on day 35.1.  If closed before DOL 5, first feed was intiated on average DOL 13; if closed after DOL 5, feeds began on DOL 28.

Average birth weight was 2620 grams.  This did not correlate with the time to definitive closure of the gastroschisis (p=0.6) nor was there an association with time to first feeding (p=0.52). Additionally, time to first feeds was not significantly different among ethnic groups (p=0.42), but there was a possible trend when analyzing time to definitive closure of the gastroschisis between ethnic groups (p=0.073).  However, early closure of gastroschisis was correlated with early initiation of feeds (p=0.0001) and shorter time to full feeds (p=0.04).

Conclusion:  Early closure of gastroschisis was associated with early feed initiation and shorter time to full feeds.  Returning bowel into the abdominal cavity sooner may minimize intestinal irritation, leading to earlier return of bowel function.  Based on these findings, early definitive closure may improve the outcome of patients with gastroschisis.

33.01 A population-based randomized controlled trial of breast cancer screening with alternate mammography and ultrasound for women aged 40 to 49 years in Taiwan

Posted on December 22, 2014

C. Huang1, C. Fann1, G. Hsu1, M. Ho1, S. Chen1, S. Chen1, C. Chen1, S. Sheen-Chen1, H. Chang1, D. Yeh1, D. Chen1, K. Chang1, S. Kon1, A. Ming-Fang1, Y.、L. Chen1, S. Chiu1, H. Chen1  1Breast Cancer Screening Group of Taiwan Breast Cancer Consortium

Background: The efficacy of mammography screening in young women aged 40-49 is not established. As the sensitivity of mammography is poorer in dense breasts, commonly seen in young women, and the performance of ultrasound is not affected by breast density, we conducted a randomized trial to investigate whether annual screening with alternating mammography and ultrasound could detect more breast cancers and earlier than no screening.

Methods:
A total of 79,690 women aged 40-49 years were invited from community in Taiwan since late 2003. These participants were randomly assigned to the study group (n=40127) to receive alternating mammography or ultrasound annually or to the control group (n=39563). The study group was further randomized to receive either mammography (the M-U group) or ultrasound (the U-M group) for the first year screen. Detection rate and sensitivity using one minus annual incidence rate of interval cancer as a percentage of the control group were compared between mammography and ultrasound based on two cross -over designed group. The excess of breast cancers detected by the combined study groups in comparison with the control group was estimated by the comparison of two cumulative incidence of breast cancer curves. The relative rate of reducing stageII+ breast cancers was also estimated using the Poisson regression model.

Results:
The attendance rate of the first -year screen was 59% (11921/20040) and 56% (11249/20087) for the M-U group and the U -M group, respectively. The repeated attendance rate of both groups was 85% in the second round and 91% in the third round. In the first round of screening, the detection rate of breast cancer for the mammography group (0.34%) was 1.5 -fold compared with the ultrasound group (0.22%). In the first three rounds, we found that the detection rates for invasive caner or all cancer by mammography were higher than their counterparts with ultrasound. The alternative screening tool design enables us to estimate the additional benefit of ultrasound, 28.6% for all breast cancer and 38.6% for invasive carcinoma. It was noted that in the fourth round, ultrasound can detect more invasive cancer with detection rate of 0.15% than 3 mammography (0.08%), as well as more early invasive cancers (stage I) (7 by ultrasound vs. 3 by mammography), although mammography consistently detects more in situ cancers than ultrasound (12 vs. 1). The combined study groups led to a marginally statistically significant 18% (Relative rate=0.82 (95% CI: 0.65 -1.02)) reduction of incidence in stageII+ breast cancers compared with the control group.

Conclusion:
The current randomized controlled trial not only demonstrated higher detection rate and performance using mammography but also indicated the complementary role of ultrasound applied to young Taiwanese women. These findings provide valuable information on mass screening modality for young women in Asian country.
 

33.02 Related Risk of Postoperative Myocardial Infarction and Blood Transfusion

Posted on December 22, 2014

R. H. Hollis1,2, J. T. McMurtrie1,2, L. A. Graham1,2, J. S. Richman1,2, T. M. Maddox4, K. M. Itani3, M. T. Hawn1,2  1Birmingham Veterans Administration Hospital,Center For Surgical, Medical Acute Care Research And Transitions (C-SMART),Birmingham, AL, USA 2University Of Alabama At Birmingham,Section Of Gastrointestinal Surgery, Department Of Surgery,Birmingham, AL, USA 3Boston University And Harvard Medical School,Department Of Surgery, VA Boston Health Care System,Boston, MA, USA 4University Of Colorado School Of Medicine,VA Eastern Colorado Health Care System,Denver, CO, USA

Introduction:   Patients with cardiac risk factors undergoing surgery are not only at increased risk of myocardial infarction (MI) but are often exposed to an increased risk of bleeding from perioperative antiplatelet therapy or from the need for therapeutic anticoagulation.  Patients who receive blood transfusions in the setting of MI are associated with worse outcomes.  However, the temporal relationship and related risk of postoperative myocardial infarction and blood transfusions are not well established.
 

Methods:   We matched patients with coronary stents who underwent inpatient non-cardiac surgery within two years of stent placement to two patients with similar cardiac risk and surgical procedures without coronary stents.  Our independent variable of interest was peri-operative transfusion and patients receiving preoperative transfusions were excluded.  Our outcome variables were MI and death at thirty-days post-operatively.  The relationship between timing of MI and transfusion was assessed.  MI time was determined by postoperative time to first troponin lab value meeting threshold 0.04 ng/ml. Time to transfusion was defined as time to receipt of first blood transfusion, with time zero assigned to intraoperative transfusions. GEE model was employed to adjust for risk factors for mortality and to determine predictors of MI.  Statistical analysis was performed by chi-square test and Wilcox t-test.
 

Results:  We identified16,807 patients with cardiac risk factors undergoing inpatient non-cardiac surgery, of which 327 (1.9%) experienced postop MI.  Of those patients with MI, 50% received a blood transfusion.  The overall median postoperative time to transfusion was sooner than MI (1 vs. 36 hrs).  Mortality was higher following MI compared to transfusion or no MI or transfusion (17.7% vs. 5.3% vs. 1.5%; P < 0.001).  Of patients with both outcomes of MI and transfusion, 57% (97/169) received a transfusion prior to MI and were at increased mortality compared to patients with first transfusion after MI on unadjusted (22.7 and 11.1%; P <0.05) and adjusted analysis(OR=2.4, 95% CI 0.9-6.2). Of all patients with a perioperative transfusion, only a history of a cardiac stent was significantly associated with MI (OR=1.8, 95% CI 1.4-2.5).
 

Conclusion:  The occurrence of a transfusion prior to MI is associated with increased mortality.  Cardiac stent placement is a predictor of MI in patients receiving a blood transfusion, suggesting that peri-operative bleeding may be in the causal pathway for post-operative MI in this patient population.

« Previous Page
Next Page »