52.01 Decision Making and Outcomes of Breast Reconstruction after Mastectomy: A Prospective, Longitudinal Study

Posted on January 18, 2016

C. Lee1,4, A. Deal4, P. A. Ubel2, R. Hugh1, L. Blizard3, K. R. Sepucha3, Y. Liu1, D. Ollila1,4, M. P. Pignone1,4 1University Of North Carolina At Chapel Hill,Chapel Hill, NC, USA 2Duke University,Fuqua School Of Business,Durham, NC, USA 3Harvard School Of Medicine,Brookline, MA, USA 4Lineberger Comprehensive Cancer Center,Chapel Hill, NC, USA

Introduction: Breast reconstruction after mastectomy has the potential to improve patient reported outcomes. Prior studies have been limited by retrospective, cross-sectional designs and lack of controls, and most have not assessed patient decision making. We evaluated outcomes of reconstruction in a prospective, longitudinal, controlled study and examined whether informed choice was associated with outcomes.

Methods: Adult women undergoing mastectomy for stage 0-III breast cancer or prophylaxis were enrolled at a single site and surveyed before surgery and at 12 months. Decision making was assessed with the Decision Quality Instrument, rating and ranking scales, and conjoint analysis. Making an ‘informed choice’ was defined as having at least 50% knowledge and treatment concordant with preferences. Satisfaction w/ breasts, psychosocial well-being (WB), sexual WB, and physical WB of the chest were measured with the BreastQ (range 0-100). Decisional outcomes were measured with the Satisfaction with Decisions and Decision Regret scales. Baseline scores were compared using t-tests. Preference concordance was assessed with a kappa score. Multivariable linear regression was used to examine differences (by treatment and by informed choice) in patient-reported outcomes at 12 months, adjusting for baseline values, demographics, and clinical characteristics.

Results: 126 patients enrolled (83% participation); 111 completed baseline and 12 mo surveys. The immediate reconstruction rate was 44% (15 delayed reconstruction cases were excluded to simplify analysis). Patients having reconstruction were more likely to be white (88 vs. 69%, p=0.05), partnered (81 vs. 49%, p=0.002), or college educated (95 vs. 76%, p=0.01), and less likely to have adjuvant therapy (17 vs. 65%, p<0.01). Mean age was 54 and did not differ by treatment. Mean knowledge was 57%. 63% of patients had treatment that was concordant with preferences. 43% of patients made an informed choice. Satisfaction with breasts, sexual WB, psychosocial WB, and physical WB-chest declined from baseline to 12 mos, regardless of reconstruction. On multivariable analysis and adjusting for baseline values, reconstruction was associated with higher satisfaction w/ breasts (beta 13.1, p=0.008) at 12 mos; it was not associated with change in sexual WB (p=0.07), psychosocial WB (p=0.16) or physical WB (p=0.61). Informed choice was not associated with higher psychosocial WB (p=0.20), decision satisfaction (p=0.57), or decision regret (p=0.77).

Conclusion: Patients having mastectomy-only are different at baseline from patients having reconstruction. Both groups experienced decline in all well-being scores after surgery. Reconstruction was associated with higher satisfaction with breasts at 12-mos, but not with higher sexual, psychosocial, or physical WB. Informed choice was not associated with better outcomes, and most patients did not make an informed choice.

51.05 Predictors of In-Hospital Postoperative Opioid Overdose after Major Elective Operations

Posted on January 18, 2016

C. E. Cauley1, G. Anderson1, M. E. Menendez3, B. Bateman2, K. Ladha2, A. B. Haynes1 1Massachusetts General Hospital,Department Of Surgery,Boston, MA, USA 2Massachusetts General Hospital,Department Of Anesthesiology,Boston, MA, USA 3Massachusetts General Hospital,Department Of Orthopedic Surgery,Boston, MA, USA

Introduction:
In 2000 the Joint Commission recommended using pain as a 5th vital sign, increasing focus on postoperative pain control. However, the benefits of pain management must be weighed against the potentially lethal risk of postoperative opioid overdose (POOD). This analysis aims to 1) describe national trends and outcomes of in-hospital POOD in patients undergoing surgery and 2) identify predictors of POOD.

Methods:
This is a retrospective multi-institutional cohort study of patients undergoing a major elective operation from 2002 to 2011 in the Nationwide Inpatient Sample. Patients with POOD were identified using ICD-9 codes for poisoning from or adverse effects of opioids. Patients with adverse effects from opioids required a concurrent code for respiratory failure, apnea, or hypoxia to be included in the POOD cohort. Multivariate logistic regression was used to identify predictors of POOD.

Results:
Among 13,982,557 patients, 11,669 (0.083%) had POOD with the rate increasing over the study period from 0.6 to 1.2 per 1,000 cases. Patients with POOD died more frequently during hospitalization (1.7% vs. 0.47%, p-value=<0.001). Substance abuse history was the strongest predictor of POOD (OR=14.9; 95% CI: 12.5-17.6) on multivariate analysis. Gender, age, race, income, geographic location, operation type, and comorbid diseases were significant predictors of POOD. Hospital variables including teaching status, size, and urban/rural location did not predict POOD.

Conclusion:
POOD is a rare, but potentially lethal complication, with increasing prevalence. Perioperative care providers should consider predictors of POOD to identify patients at greatest risk, while providing adequate pain control. Postoperative monitoring and treatment protocols should be thoughtfully employed to target high risk patients and avoid this potentially fatal complication.

51.06 Impact of Physician Specialization on Inpatient Hospital Outcomes after Thyroidectomy for Cancer

Posted on January 18, 2016

A. D. McDow1, J. D. Mellinger1, S. Ganai1, S. S. Desai2 1Southern Illinois University School Of Medicine,General Surgery,Springfield, IL, USA 2Southern Illinois University School Of Medicine,Vascular Surgery,Springfield, IL, USA

Introduction: General surgeons (GS) and otolaryngologists (ENT) commonly perform thyroid surgery. Each specialty differs in surgical training, including residency structure and breadth of operative experience. The aim of this study is to evaluate the impact of surgeon specialty and volume on outcomes following total thyroidectomy for cancer.

Methods: A retrospective cohort study was performed on patients undergoing total thyroidectomy for thyroid cancer between 2007 and 2011 as identified from the National Inpatient Sample (NIS). Physician identifiers were used to classify surgical specialty based on procedural complement. Propensity score matching for comorbidities was performed to allow for comparisons between GS and ENT for outcomes including complications, mortality, length of stay (LOS), and cost of care. Comparisons between provider volume and complication rates were also assessed.

Results: A total of 11,391 inpatient total thyroidectomies were identified in the NIS, of which 54% were performed by ENT and 46% by GS. Postoperative complications occurred in 3.4% of both groups. GS were less likely than ENT to have neurologic (p<0.01), respiratory (p<0.01), and hemorrhagic complications (p<0.001), and were more likely to have speech disturbances (p<0.05). There was no significant difference in inpatient mortality between providers; however, the cost of care and LOS were significantly lower for GS. After propensity score matching for comorbidities, cost ($8,304 vs. $11,530; p<0.001) and LOS (1.7 days vs. 2.3 days; p<0.001) remained significantly lower for GS in comparison to ENT. Surgeons, regardless of specialty, who performed less than 5 total thyroidectomies per year had a higher complication rate compared to those with higher volume.

Conclusion: While generalizability is limited based on exclusion of outpatient procedures, our analysis shows that physician specialty and case volume are significantly associated with cost, length of stay, and complication rate following total thyroidectomy. A more detailed evaluation of the differences in technique and postoperative management between the two specialties may inform changes in practice that may benefit patients who undergo total thyroidectomy by either specialty.

51.08 Identifying Injury And Fatality Risks In Aeromedical Transport: Making It Safer For The Life-Savers

Posted on January 18, 2016

H. H. Hon1, N. Barry1, U. MacBean1, J. P. Anagnostakos1, T. R. Wojda1, D. C. Evans2, C. Jones2, M. Portner1, B. A. Hoey1, W. S. Hoff1, P. Thomas1, S. P. Stawicki1 1St. Luke’s University Health Network,Department Of Surgery,Bethlehem, PA, USA 2Ohio State University,Department Of Surgery,Columbus, OH, USA

Introduction: Aeromedical transport (AMT) is well-established, safe, and reliable life-saving option for rapid patient transfers to healthcare delivery hubs. However, due to the very nature of AMT, fatal and non-fatal occurrences are bound to happen. This study reviews aeromedical incidents since 2003, aiming to provide additional insight into a wide range of factors potentially associated with fatal and non-fatal AMT incidents (AMTI). We hypothesized that weather/visual conditions, post-crash fire, and time of day are all correlated with the risk of AMTI with ‘injury or fatality’.

Methods: Various specialty databases were queried for AMTI between January 1, 2003 and June 30, 2015. Additional Internet-based resources were also utilized to find any additional AMTI (including non-US occurrences) in order to augment the event sample size available for analysis. Univariate analyses of the collected sample were then performed for association between ‘fatal crash or injury’ (FCOI) and weather/visual conditions, aircraft type/make, pilot error, equipment failure, post-incident fire, time of day (6am-7pm vs 7pm-6am), weekend (Fri-Sun) versus weekday (Mon-Thurs), season of the year, presence of patient on board. Variables reaching significance level of P<0.20 were included in multivariate analysis.

Results: A total of 58 AMTI were identified. Helicopters were involved in 51/58 AMTI, with 7/58 fixed-wing incidents. Comparing pre-2003 data with post-2003 data (Figure 1), we noted an overall decrease in AMTI per month (0.70 versus 0.38, respectively). However, the number of fatalities per year increased slightly (7.20 vs 8.08, respectively). There was no association between aircraft make/model and FCOI. In univariate analyses, weather/visual conditions, time of incident (7pm-6am), and post-incident fire all reached statistical significance sufficient for inclusion in multivariate analysis. Factors independently associated with FCOI included post-incident fire (O.R. 12.6, 95% CI 1.16-137.2) and time of incident between 7pm and 6am (O.R. 8.28, 95% CI 1.25-55.0). Weather conditions and impaired visibility were not independently associated with FCOI.

Conclusion: The current study supports previous observation that post-crash fire is independently associated with FCOI. However, our data do not corroborate published reports suggesting that weather conditions and impaired visibility are predictive of fatal AMTI. In addition, this study demonstrates that flights between the hours of 7pm-6am may be associated with greater odds of FCOI. Efforts directed at identification, remediation and active prevention of factors associated with AMTI and FCOI are warranted given the global increase in aeromedical transport.

51.02 No Catheter, No CAUTI: A Formal Process Improvement Project to Reduce Urinary Catheter Use in the OR

Posted on January 18, 2016

M. W. Wandling1,2,3,4, K. Schelling4, A. Mikolajczak4, W. Wilson4, C. M. Gonzalez4,5, C. Perry4, K. Y. Bilimoria1,2,3,4, A. D. Yang1,2,3,4 1Northwestern University Feinberg School Of Medicine,Department Of Surgery,Chicago, IL, USA 2Northwestern University Feinberg School Of Medicine,Surgical Outcomes & Quality Improvement Center,Chicago, IL, USA 3Northwestern University Feinberg School Of Medicine,Institute For Public Health And Medicine, Center For Healthcare Studies,Chicago, IL, USA 4Northwestern Medicine,Chicago, IL, USA 5Northwestern University Feinberg School Of Medicine,Department Of Urology,Chicago, IL, USA

Introduction:

Catheter-associated urinary tract infection (CAUTI) is a major, preventable source of hospital-acquired infection. Our hospital has performed poorly in regard to CAUTI rates in surgical patients, ranking in the bottom 10-20% of ACS NSQIP hospitals. Since CAUTI rate is a publicly reported measure linked to healthcare quality and financial reimbursement, we aimed to decrease CAUTI rates by focusing on both reducing catheter utilization in the operating room (OR) and removing them in the immediate post-operative period.

Methods:

A multi-disciplinary team of stakeholders involved in the peri-operative care of surgical patients used DMAIC methodology for this project. Institutional data were analyzed before and after our interventions. Goals were set to: 1) decrease the number of catheters inserted in the OR for cases less than three hours, 2) increase the number of catheters removed prior to transfer to the floor, 3) stimulate culture change regarding surgeons’ utilization of urinary catheters, and 4) decrease CAUTI rates attributable to catheters inserted in the OR. We implemented a formal pre- and post-operative discussion of catheter indication and necessity with the surgical team (pre-operative) and attending surgeon (post-operative) into the OR nursing procedural protocol. Documentation of compliance with each intervention was mandated to facilitate the audit and feedback of performance. Six months of post-intervention data regarding performance on each of our goals was collected and analyzed.

Results:

A mean of 437 catheters per month were inserted for cases lasting less than three hours prior to the intervention, compared to a mean of 391 catheters per month post-intervention (p=0.044). Utilization of catheters for cases shorter than three hours decreased from 23.7% pre-intervention to 20.9% post-intervention (p=0.001), representing an 11.8% relative reduction in overall catheter insertion. In addition to reducing overall utilization, the percent of catheters removed prior to the patient transferring to the floor increased from 18.3% to 21.0% (p=0.252). Lastly, the rate of surgical CAUTI decreased from 2.5 occurrences per month prior to intervention to 0.3 occurrences per month post-intervention (p=0.006), including zero occurrences in four out of the six months following intervention. This represents an 89.1% decrease in the rate of surgical CAUTI.

Conclusion:

By using rigorous, data-driven process improvement methodology, our multi-disciplinary team was able to identify and successfully implement novel interventions aimed at decreasing the rate of CAUTI among surgical patients at our institution. We successfully met our goals, most notably dramatically decreasing the rate of CAUTI in surgical patients. These interventions have led to institutional changes in practice regarding peri-operative catheter insertion that will continue to decrease CAUTI rates and improve the quality of surgical care.

51.03 Are NSQIP Hospitals Unique? A Description of Hospitals Participating in ACS NSQIP.

Posted on January 18, 2016

C. R. Sheils1,2, A. R. Dahlke1, A. Yang1, K. Bilimoria1 1Northwestern University,Department Of Surgery,Chicago, IL, USA 2University Of Rochester,School Of Medicine,Rochester, NY, USA

Introduction: The American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) is a well-recognized program for surgical quality measurement. Given the widespread use of ACS NSQIP in research and recent calls for it to become a platform for national public reporting and pay-for-performance initiatives, it is important to understand which types of hospitals elect to participate in the program. Our objective was to compare the characteristics of ACS NSQIP-participating hospitals to non-participating hospitals in the United States.

Methods: Using the 2013 American Hospital Association data on hospital characteristics, hospitals participating in ACS NSQIP were compared to non-participating hospitals. The 2013 Healthcare Cost Report Information System (HCRIS) dataset was used to calculate hospital operating margin as a measure of financial health. The CMS 2013 Inpatient Prospective Payment System (IPPS) Final Rule Impact File was used to abstract the Medicare and Medicaid Services Value Based Purchasing (VBP) and Disproportionate Share adjustment scores, which were used as proxies for hospital quality and patient population, respectively.

Results: Of 3,872 total U.S. general medical and surgical hospitals, 475 (12.3%) participated in ACS NSQIP. ACS NSQIP hospitals performed 29.0% of operations in the U.S, with a slightly greater share of inpatient operations (32.4%) and a smaller share of outpatient operations (27.1%). Compared to non-participating hospitals, ACS NSQIP hospitals had a higher mean annual inpatient surgical case volume (6,426 vs 1,874; p<0.001), a larger number of hospital beds (420 vs 167; p<0.001), were more often academic affiliates (35.2% vs 4.1%; p<0.001), were more often accredited by JCAHO and CoC (p<0.001), and had higher mean operating margins (p<0.05). ACS NSQIP hospitals were less likely to be designated as critical access hospitals (p<0.001). No significant differences in VBP or Disproportionate Share adjustment scores were found. States with the highest percentage of hospitals participating in ACS NSQIP were states with established surgical quality improvement collaboratives (Figure 1).

Conclusion: Hospitals that participate in ACS NSQIP represent 12% of all U.S. hospitals performing inpatient surgery, yet they perform nearly 30% of all surgeries done in the U.S. ACS NSQIP disproportionately includes larger, accredited, and academic-affiliated hospitals with more financial resources. These findings should be taken into account in research studies using ACS NSQIP, and more importantly, indicate that additional efforts are needed to address barriers to enrollment in order to facilitate participation in surgical quality improvement programs by all hospitals.

51.04 Venous Thromboembolism Rates Associated with Interrupted Prophylaxis and EMR Alerts

Posted on January 18, 2016

N. A. Lee1, R. Ramanathan1, Z. Gu2, E. M. Rensing2, R. Sampson2, M. B. Burrows2, S. M. Hartigan3, N. Nguyen2, T. G. Potter2, T. Trimmer2, A. C. Grover1 1Virginia Commonwealth University,Department Of Surgery,Richmond, VA, USA 2Virginia Commonwealth University Health System,Richmond, VA, USA 3Virginia Commonwealth University,Department Of Internal Medicine,Richmond, VA, USA

Introduction:
Venous thromboembolisms (VTE) are potentially preventable adverse events associated with significant morbidity and mortality. VTE rates are also a publicly tracked patient safety measure affecting institutional quality metrics and reimbursement. Electronic medical record (EMR) alerts have been widely adopted to improve VTE prophylaxis administration. This study examines the association of VTE prophylaxis and EMR alerts with VTE rates.

Methods:
At our urban academic medical center, 10318 adult surgical admissions between November 2013 and March 2015 were queried for VTE defined by Agency for Healthcare Research and Quality (AHRQ) Patient Safety Indicator 12 and Joint Commission Core Measure VTE-6. Daily pharmacologic prophylaxis patterns were prospectively collected for patients and categorized as continuous (initiated within 24h without interruptions), delayed (initiated >24h without interruptions), interrupted (interrupted for >24h with or without delay), none, or other (nonstandard drugs or dosing). Patients were also categorized by whether an EMR alert fired during their admission. Associations between VTE incidence, prophylaxis, length of stay (LOS), and demographics were explored.

Results:
There were 131 VTEs (12.7 VTE per 1000 patients) among the surgical admissions. Patients who developed a VTE were older (56.4 vs. 52.2 years, p<0.01) and had longer LOS (16.3 vs. 6.5 days, p<0.01). 45.7% of patients had an EMR alert to prescribe VTE prophylaxis. The VTE rate for patients who received an alert was significantly higher than patients without an alert (17.5 vs 9.0 per 1000 patients; p<0.01). 33.6% of patients received continuous prophylaxis, 18.3% experienced delays, 13.2% experienced interruptions, 25.7% received no prophylaxis and 9.3% had nonstandard prophylaxis regimens. Patients with interruptions had a significantly higher incidence of VTE than patients with continuous prophylaxis (8.8 vs 30.7 per 1000 patients, p<0.01) and all other prophylaxis groups. Increased LOS was associated with increased likelihood of interruptions and VTE. In multivariate logistic regression analysis, prophylaxis group and age were associated with VTE incidence independent of gender, race, and LOS. The highest incidences of VTE were among patients admitted to cardiac surgery, followed by trauma and otolaryngology. Interruptions were most common among patients admitted to trauma surgery, otolaryngology, and bariatric surgery.

Conclusion:
Despite continuous prophylaxis 8.8 per 1000 patients still developed a VTE, and interruptions conferred more risk of VTE than delays. This suggests patient safety measures should be reevaluated to emphasize minimizing interruptions in addition to delays and overall VTE rates. Increased efforts to investigate causes and develop multidisciplinary strategies to minimize interruptions should be undertaken. Optimization of EMR alerts may be a useful adjunct to identify patients at high risk for VTE.

50.09 Contemporary Outcomes Of Surgical Repair Of Thoracic Outlet Syndrome In A Nationally Validated Data

Posted on January 18, 2016

B. J. Nejim1, T. Obeid1, I. Arhuidese1, S. Wang1, K. Yin1, J. Canner1, M. Malas1 1Johns Hopkins University School Of Medicine,Surgery,Baltimore, MD, USA

Introduction: The aim of the study is to evaluate the occurrence of Thoracic Outlet Syndrome (TOS) surgery, clinical presentations, and 30-day post-operative outcomes using a nationally validated prospective database.

Methods: Patients who underwent cervical/first rib resection surgery (FRRS) between years 2005-2013 were identified in the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database. TOS types were examined using post-operative ICD-9 diagnosis codes. Chi-square and ANOVA tests were used to estimate the proportions and mean (±SD) values as appropriate.

Results: 1,180 patients underwent FRRS during the study period, 1,007 (85.3%) were of the neurogenic type (NTOS), 32 (2.7%) patients had arterial TOS (ATOS) and 141 (12.0%) patients had venous TOS (VTOS). The majority of patients were female (58.3%). Patients with ATOS were significantly older (ATOS: 48.9, ±13.8, VTOS: 35.2, ±11.9 and NTOS: 34.3, ±12.3 years). Overall, 19.3% of the cohort examined were current smokers and this proportion was not different across TOS types (p=0.065). However, among smokers, mean pack-years was higher for patients with ATOS (16.2, ±27.9) compared to NTOS (4.7, ±11.5) and VTOS (3.4, ±11.6) (p<0.001). Mean operation time was over one hour longer for patients with ATOS than for patients with NTOS or VTOS (p<0.001). Average hospital length of stay was longer for patients with ATOS (5.5 days, ± 6.7) (p<0.001). The majority of surgical site infections were superficial (1% of NTOS, 1.4% VTOS and none of the ATOS cases). Post-op pneumonia was proportionally higher in patients presenting with ATOS (6.25%, p<0.001). Post-op peripheral nerve injury was more likely to occur after FRRS for ATOS (p=0.019). Only one patient developed post-operative stroke after FRRS for ATOS. Return to OR within 30 days from the operation occurred in a total of 73 cases (5.6% in NTOS, 18.8% in ATOS and 7.8% in VTOS patients, p=0.007). For the years 2011-2013; all the re-operations were unplanned, with ATOS as the most likely to be re-operated on (p=0.013), readmissions occurred in 33 patients and didn’t differ across TOS types (p=0.525).

Conclusion: Thoracic Outlet Syndrome has been previously studied mainly in highly-experienced single-center contexts. We sought to use NSQIP to further explore its occurrence, types and post-op outcomes at a national level. To our knowledge, this is the first and largest study that represents outcomes of FRRS for TOS from different centers across the United States. Patients underwent FRRS due to ATOS had the highest rates of complications. Further studies are warranted to compare real-world outcomes to highly specialized centers.

50.10 Is Post-operative Atrial Fibrillation after CABG a Marker of an Underlying Process?

Posted on January 18, 2016

J. B. Grau1,2, C. K. Johnson1, C. E. Kuschner1, G. Ferrari1,2, R. E. Shaw1 1Valley Heart And Vascular Institute,Cardiothoracic Surgery,Ridgewood, NJ, USA 2University Of Pennsylvania,Surgery,Philadelphia, PA, USA

Introduction:

Post-operative atrial fibrillation (AF) after coronary artery bypass grafting (CABG) occurs frequently, affecting 20-35% of CABG patients. Post-CABG AF can result in longer hospital stay, and has been shown to decrease survival at 10 years. This study assesses survival and major adverse cardiovascular events (MACE) at 20 years in patients with new-onset post-operative AF after CABG.

Methods:
Patients undergoing CABG between 1994 and 2014 at the Valley Hospital were assessed and matched by surgical approach. In the resulting population of 1856 patients, 363 developed new-onset post-operative AF. Logistic regression was used to determine the significant factors associated with the development of post-op AF. Patients were followed for a mean of 12 ± 5 years (ranging 1 to 21.4 years). A univariate Kaplan Meier analysis was used to compare long-term survival and MACE in patients with and without new-onset Post-op AF. A multivariate Cox Proportional Hazards Regression analysis was used to determine which factors were independently associated with decreased survival.

Results:
Overall 363 patients (19.6%) developed AF post CABG. Logistical regression demonstrated that male gender, increased age, prior cerebrovascular accidents and on-pump CABG were associated with the development of post-op AF. Follow-up MACE in the 1856 patients included 30-day mortality (17 pts, 0.9%), late death (513 pts, 27.6%) late myocardial infarction (76 pts, 4.1%), late PCI (210 pts, 11.3%) repeat CABG (14 pts, 0.8%), any repeat revascularization (219 pts, 11.8%), any stroke (78 pts, 4.2%) and cerebrovascular accident (34 pts, 1.8%). Kaplan Meier analysis showed that patients with post op AF had poorer long-term survival (Log Rank Test p<0.0001) and increased occurrence of MACE (Log Rank Test p<0.0001). Multivariate Cox Proportional Hazards Regression showed that LVEF<35%, Increased age, diabetes, CHF, Active smoking, and History of Renal failure were independently associated with increased MACE and decreased survival. Presence of On-pump CABG, presence of cardiogenic shock, and use of LIMA-SVG over BIMA were independently solely associated with increased mortality. Presence of atrial fibrillation post CABG was independently associated with reduced survival (Adj HR: 1.440; 95%CI: 1.179-1.757; p<0.0001) and increased occurrence of MACE (Adj HR: 1.297; 95CI: 1.094-1.537; p=0.003).

Conclusion:

Our study demonstrates that new-onset AF after CABG is associated with poorer survival and increased MACE at 20 years follow-up. Unfortunately, we have no tools to know how many of the patients that developed AF actually had subclinical symptoms prior to CABG. It is unclear whether CABG causes AF or uncovers an underlying subset of patients conducive to poor outcomes, where AF is just an additional factor. This data suggests that AF is a marker of pathology that deserves close surveillance, not only in the post-operative period, but also in the long term follow up.

51.01 Transfusions and the Domino Theory of Complications: Failure to Rescue Begins In the Operating Room

Posted on January 18, 2016

C. D. Tzeng1, B. M. Evers1, J. B. Zwischenberger1, M. V. Williams1, B. R. Boulanger1, P. C. McGrath1, J. T. Martin1 1University Of Kentucky,Lexington, KY, USA

Introduction:
Failure to rescue (FTR: % patients who died after ≥1 complication) is an increasingly important hospital performance measure. This study sought to analyze 6 high-risk index operations to identify the most common complication cascades and their initial physiologic insults which could be targeted to prevent FTR.

Methods:
Elective patients with hematocrit ≥30% undergoing 6 index procedures were identified from the 2012-2013 ACS-NSQIP dataset. Association between number of morbidity events and FTR was analyzed with logistic regression. The most common complication cascade was analyzed for each procedure to identify the most common initial complication and to compare FTR rates.

Results:
34,556 patients underwent esophagectomy (morbidity/mortality/FTR-32.8%/2.1%/5.5%), lung resection (14.1%/1.7%/9.7%), gastrectomy (16.9%/1.7%/9.2%), pancreaticoduodenectomy (25.8%/2.3%/7.8%), hepatectomy (15.5%/1.6%/9.8%), and colectomy (15%/1.3%/6.5%). Patients who died suffered more sequential complications (Figure) and multiplicative odds of death with additional events. Bleeding requiring transfusion was the most common initial event in every index procedure (13.6%-overall, range 8.1%-lung, 21.5%-pancreaticoduodenectomy/hepatectomy). Among deaths, the most common cascade was bleeding (42%), respiratory failure (18%), septic shock (28%). Among non-transfused patients, the most common cascade was superficial infection, organ space infection, sepsis. Non-transfused patients were less likely to have multiple complications and had better mortality and FTR rates (mortality 0.99% vs. 4.84%; FTR 5.37% vs. 11.99%, p<0.001).

Conclusion:
Avoidance of perioperative transfusions mitigates the magnitude of complication cascades and improves FTR rates. Although bleeding requiring transfusion is commonly overlooked when measuring perioperative complications, it is the most common sentinel domino to fall. To augment surgical mortality reduction programs, hospitals should pursue aggressive programmatic transfusion avoidance strategies in the perioperative setting.

50.06 Factors Associated with the Academic Productivity of Cardiothoracic Surgical Divisions

Posted on January 18, 2016

C. Rosati1, N. P. Valsangkar1, P. N. Vardas1, L. Chabtini1, M. W. Turrentine1, J. W. Brown1, L. G. Koniaris1 1Indiana University School Of Medicine,Department Of Surgery,Indianapolis, IN, USA

Introduction: Several criteria are currently used to evaluate medical institutions and individual physicians as well. We aimed to investigate which tools might represent valid correlates of the academic productivity of cardiothoracic (CT) surgeons.

Methods: We collected data about 57 leading academic CT surgical groups in the US, including: seniority (i.e. time since graduation from medical school), H index and institutional role (i.e. being chairperson of the division) of each individual surgeon; hospital and medical school rankings according to the 2015-2016 US News & World Report; ranking by NIH funding (granted to either their medical school, department of surgery, or affiliated department of internal medicine), as provided by the 2014 report of the Blue Ridge Institute for Medical Research; and designation as either an independent CT department or as a division of a broader department of surgery.

Results: Among 663 CT surgeons, academic productivity, as measured by the median H index of each group, was significantly different among the 57 institutions (p < 0.001). Several hospital categories (‘Adult Cardiology and Heart Surgery’, ‘Adult Pulmonology’, ‘Adult Cancer’, and ‘Honor Roll’), as well as the medical school ranking for research (but not for primary care), from the US News & World Report were significantly associated with differences in academic productivity (see Graphic). Ranking based on NIH funding was not a significant correlate, either considering funding provided to the departments of surgery (top 10 institutions vs. other: median H index 14.5 vs. 14; p = 0.94), to medical schools (17 vs. 14; p = 0.26), or to the affiliated departments of internal medicine (15 vs. 14; p = 0.55), respectively. Designation as an independent department (median H index: 15) was not associated with a different academic productivity than designation as a division of a broader department of surgery (median H index: 14; p = 0.58). When the division chairperson had an individual H index of at least 50, the whole CT surgical group was overall more productive (median H index: 18) than those groups with a chairperson whose H index was less than 50 (median H index: 14; p = 0.005). Seniority was never found to be significantly different in any of the above mentioned comparisons.

Conclusion: The US News & World Report rankings provide a good representation of the academic productivity of cardiothoracic surgical groups, while NIH funding rankings do not. Designation as a stand-alone department or as a dependent division was found to be uninfluential. The individual productivity of the chairperson is strongly correlated with the overall productivity of the whole group.

50.07 Quality of Life (QOL) Impact of Surgical Resection And Adjuvant Chemotherapy in NSCLC Patients

Posted on January 18, 2016

O. A. Lucas1,2, W. Tan1, S. Yendamuri1,2, E. Dexter1,2, A. Picone1, M. Hennon1,2, M. Huang1,2, M. Reid1,3, A. Adjei1,2, T. Demmy1,2, C. Nwogu1,2 2University At Buffalo School Of Medicine And Biomedical Sciences,Buffalo, NY, USA 3University At Buffalo School Of Public Health,Buffalo, NY, USA 1Roswell Park Cancer Institute,Buffalo, NY, USA

Introduction: While others have examined post-operative QOL in NSCLC patients undergoing thoracotomy, similar reports in VATS patients, especially those who received adjuvant chemotherapy, are scarce

Methods: Patients with Stage I or II NSCLC anticipating surgical resection were enrolled in a study using PET probe guidance to detect lymph node metastasis. The Short Form 36 Health Survey (SF-36) assessed changes in patient QOL. The SF-36 was completed by participants pre-operatively and 1, 3, and 6 months post-operatively. A mixed analysis of variance compared scores longitudinally and between groups.

Results:100 patients were enrolled in the study. 88 patients completed the baseline survey and 75 completed at least one post-operative survey. 25 patients received adjuvant chemotherapy following VATS resection and lymph node mapping. Adjuvant chemotherapy started at 1.6 and ended at 3.7 months (medians) after surgery. The trend in overall QOL score between chemotherapy and non-chemotherapy patients was statistically different (p=.004). This difference peaked at 3 months post-operatively (57 vs. 48) and was greatest when comparing role limitations due to physical health (36 vs. 13, p=.028) [Figure 1]. The QOL dimensions significantly impaired by chemotherapy were physical functioning (p=.048), energy/fatigue (p=.001), emotional well-being (p=.031), and social functioning (p=.005). The entire surgical cohort (94 VATS and 6 Thoracotomy) had a drop in QOL scores at 1 month post-operatively (48 vs. 65 baseline score, p<.0001), with recovery by month 6.

Conclusion:Adjuvant chemotherapy patients have a longer QOL decline than those receiving no chemotherapy, but scores return to baseline by month 6. These data help physicians set more realistic QOL impact expectations during chemotherapy and surgery discussions with their patients.

50.08 Utility of Preoperative Aspirin for Patients Undergoing Combined Coronary and Valve Operations

Posted on January 18, 2016

R. Kashani1, S. Sareh1, M. Tamrat1, N. Satou1, R. Shemin1, P. Benharash1 1David Geffen School Of Medicine, University Of California At Los Angeles,Division Of Cardiac Surgery,Los Angeles, CA, USA

Introduction: Numerous studies have demonstrated the benefits of postoperative aspirin administration following coronary artery bypass grafting (CABG) and other cardiac operations. However, the role of preoperative aspirin administration in cardiac surgery still remains uncertain. While several studies have reported lower mortality rates associated with preoperative aspirin use in CABG patients, there is a paucity of data regarding patients undergoing CABG combined with valvular procedures. The present study aimed to determine the utility of preoperative aspirin administration on clinical outcomes in the setting of combined CABG with valve surgery.

Methods: A total of 565 patients who underwent combined CABG and valve surgery at our institution from 2008 to 2015 were identified for analysis. After exclusion of emergency cases, 536 patients were placed into two groups: those who received aspirin within 5 days prior to surgery (ASA group; n=328), and those who did not (N-ASA group; n=208). Baseline characteristics associated with aspirin use—including age, gender, race, body mass index, recent myocardial infarction, smoking, and use of anticoagulants—were compared between groups. Propensity score matching based on twenty-five preoperative factors was then used to account for intergroup differences. Logistic regression models were developed to compare the rate of postoperative complications, including renal failure, major adverse cardiovascular events (MACE), and 30-day mortality, between groups. All statistical analysis was performed using Stata 13.0 (StataCorp, College Station TX).

Results: After propensity score matching, the sample size was reduced to 517 patients (ASA; n=322, N-ASA; n=195). The ASA group had significantly higher rates of patients above age 75; patients taking anticoagulants, statins or angiotensin converting enzyme inhibitors; and patients with a prior history of peripheral vascular disease (P<0.05). ASA patients were not more likely to have mitral or aortic valve surgery. On logistic regression analysis, no significant differences were found in the rate of postoperative complications, including 30-day mortality. However, a trend towards a decreased incidence of MACE and an increased incidence of renal failure and intra/postoperative blood transfusion was evident [Table 1].

Conclusion: In a propensity-matched cohort, preoperative aspirin use was not associated with an increased incidence of adverse events in patients undergoing combined CABG and valve procedures. These patients often comprise a distinct group since coronary disease is frequently found incidentally in valve surgical candidates. The utility of aspirin administration in the preoperative period in such patients deserves further evaluation.

50.03 Upstaging and Survival after Robotic-Assisted Thoracoscopic Lobectomy for Non-Small Cell Lung Cancer

Posted on January 18, 2016

K. Toosi4, F. O. Velez-Cubian2, E. Ng4, C. C. Moodie1, J. Garrett1, J. P. Fontaine1,2,3, E. M. Toloza1,2,3 1Moffitt Cancer Center,Thoracic Oncology,Tampa, FL, USA 2University Of South Florida Morsani College Of Medicine,Surgery,Tampa, FL, USA 3University Of South Florida Morsani College Of Medicine,Oncologic Sciences,Tampa, FL, USA 4University Of South Florida Morsani College Of Medicine,Tampa, FL, USA

~~Introduction: We investigated efficacy of lymph node (LN) dissection, detection of occult LN metastasis, and survival of patients after robotic-assisted thoracoscopic pulmonary lobectomy for non-small cell lung cancer (NSCLC).

Methods: We retrospectively analyzed all patients who underwent robotic-assisted lobectomy for NSCLC by one surgeon over a 44-month period. Clinical stage was determined by history & physical, computerized tomography (CT), positron-emission tomography (PET), brain imaging studies (MRI), and endobronchial ultrasonography (EBUS). Pathologic stage was determined by intraoperative findings & final pathology. Patient survival was assessed through chart reviews, cancer registries, and national obituary searches. Kaplan Meier curves were generated for clinical and pathologic stages.

Results: Of 249 patients (mean age 67.8±0.6yr; range 39-87) who underwent robotic assisted pulmonary lobectomy for NSCLC, mean tumor size was 3.2±0.1cm (range 0.5-11.0cm), most commonly adenocarcinoma (62.7%), squamous cell carcinoma (21.7%), and neuroendocrine carcinoma (8.0%). Assessment of ≥3 mediastinal (N2) stations occurred in 245 or 98.4% of our cohort, with 218 (87.6%) of the patients having ≥3 N2 stations reported. Mean N2 stations assessed was 4.1±0.1 stations with a mean N2 stations reported of 3.6±0.1 stations. Our overall mean of LN stations (N1+N2) with actual lymph nodes retrieved was 5.5±0.1 stations. The mean individual mediastinal lymph nodes retrieved was 7.7±0.3 LNs, for a total of 13.9±0.4 N1+N2 LNs. There were 159 (63.9%) patients who were clinical stage 1 versus 134 (53.8%) who were pathologic stage 1, with 67 (26.9%) of patients upstaged (including 20 patients from N0 to N1, 17 patients from N0 to N2, and 4 patients from N1 to N2) and 37 (14.9%) downstaged. Using clinical stage, a statistically significant difference in survival only existed between Stage I and Stage IV patients; however, using pathologic stage shows a statistically significant difference in survival between Stage I and Stage III, Stage I and Stage IV, Stage II and Stage III, and Stage II and Stage IV patients. For pathological stage, 1-year and 3-year survival with 95% confidence intervals were as follows: Stage I 92% (87-97%) and 75% (63-87%), Stage II 83% (70-96%) and 73% (49-97%), Stage III 75% (63-87%) and 44% (26-62%), and Stage IV 67% (37-97%) and 0%. For clinical and pathologic stages, 1-year and 3-year survival improved the lower the stage.

Conclusion: Mediastinal LN dissection during robotic-assisted lobectomy results in more LNs and LN stations assessed with greater upstaging than during VATS or thoracotomy. Upstaging has a direct effect on patient survival.

50.04 Use of Digital Health Kits to Reduce Readmissions After Cardiac Surgery: Results of a Pilot Program

Posted on January 18, 2016

I. E. McElroy1, A. Zhu1, G. Miranda1, H. Wu1, M. Nguyen1, R. Shemin1, P. Benharash1 1University Of California – Los Angeles,Department Of Cardiac Surgery,Los Angeles, CA, USA

Introduction:
Unintended rehospitalizations after surgical procedures represent a large percentage of readmissions and have been associated with increased morbidity and cost of care. Beginning in 2017, Medicare will enforce provisions that subject hospitals to financial penalties for excess postoperative readmissions. Rehospitalizations following cardiac operations have been linked to rhythm disturbances and pulmonary complications, amongst others. Technologic advances in remote monitoring have led to the use of web-based digital health kits (DHK) aimed at reducing readmissions and improving postoperative outcomes. The present study was performed to determine DHK’s efficacy in preventing 30-day readmissions and changes in patient satisfaction following the use of these devices.

Methods:
This was a prospective study of all adult patients who underwent operations for valvular and coronary artery disease at our institution from 03/2014 to 06/2015. During the study period, 558 adult patients (Mean age: 64± 14, 33% female) were identified, 27 of whom received a DHK following discharge (531 control group). In addition to providing a live video link with a provider specializing in cardiac surgery, the kit allowed for automatic daily transmission of weight, oxygen saturation, heart rate and blood pressure. Patients also completed a daily health survey targeting heart failure symptoms, wound healing, ambulation and adherence to medications. Abnormal vitals or survey responses triggered automatic notifications to the healthcare team. Satisfaction surveys were administered to participants and members of the healthcare team (Scale 1-5, 5=highly-satisfied). Pearson’s chi-squared and student’s t-tests were used to assess statistical differences in baseline characteristics and outcome variables (STATA 13).

Results:
During the study period, use of DHK led to 1734 alerts and 138 interventions. The readmission rate for the DHK group was lower than the control group (7.4% vs. 10.9%, P=0.57). Reasons for readmission in the DHK group were amiodarone toxicity and syncope. Reasons for readmission in the control group included arrhythmias, heart failure, pericardial effusion, acute vascular complication, infection, and respiratory complications. Satisfaction surveys showed an overall satisfaction rating of 4.9 (± 0.53) for DHK patients and 4.9 (±0.20) for members of the care team.

Conclusion:
In our study, utilization of DHKs was associated with a non-significant decrease in 30-day readmission rates. Both patients and members of the healthcare team were highly satisfied with this technology. DHKs appear to extend care beyond the inpatient period and may provide much needed monitoring of surgical patients after discharge. However, their use leads to many provider alerts and interventions, making this modality resource-intensive. Further studies are warranted to evaluate the efficacy of such kits in reducing readmissions and costs of care.

50.05 Primary spontaneous pneumothorax in healthy menstruating females

Posted on January 18, 2016

C. Mehta1, B. Stanifer1, S. Fore-Kosterski1, A. Yeldandi1, C. Gillespie1, S. Meyerson1, D. Odell1, M. DeCamp1, A. Bharat1 1Northwestern Memorial Hospital,Chicago, IL, USA

Introduction: The etiology of primary spontaneous pneumothorax (PSP) in healthy normal-statured women differs from tall, thin men. Nevertheless, PSP is managed similarly in both men and women. We investigated the natural history of PSP in females and determined the incidence of clinical, morphological, and histological features of catamenial pneumothorax (CP).

Methods: Between 5/2009 – 6/2015, all healthy normal-statured, menstruating women without tobacco or marijuana smoking, lung disease, trauma or pregnancy were included in this study. Clinico-demographic variables were studied using a prospective database. Kaplan-Meier curves were constructed to determine PSP recurrence and analyzed using log-rank test.

Results:The mean age of the study cohort (n=33) was 33.4+/- 14.4 years. Nine (27%) had left-sided and 24 (73%) had right-sided PSP. Tube thoracostomy in all led to successful initial treatment. Overall recurrence rate during follow up (median 14 mo) was 64%. Right PSP had higher and earlier recurrence (70%; median follow up 7 mo) compared to left (56%; 16 mo, p=0.02). Four patients (12%) presented with a large recurrent tension pneumothorax within the first 6 months of the initial episode. Two patients refused further treatment after ipsilateral recurrence. The remaining 19 patients (58%) underwent surgery including apical wedge resection with mechanical pleurodesis (12/19, 63%) or subtotal pleurectomy (7/19, 37%). One patient had ipsilateral recurrence after surgery which was treated with talc poudrage. Histology showed apical blebs/fibrosis in 11 patients with mesothelial hyperplasia and reactive changes but no endometrial glands or stroma.

Eight (24%) patients had symptoms within 72 hours of onset of menses. Of these, 3 patients had known pelvic endometriosis. All 8 patients were placed on oral contraception, however all had ipsilateral recurrence within the first year of the initial episode. Six (75%) were right and 2 (25%) were left pneumothorax. All 8 patients underwent surgery at the second episode. Thoracoscopy revealed diaphragmatic fenestrations in 4 (50%) and nodular deposits in 3 patients (38%). No endometrial glands/stroma were seen in any specimens. Three (38%) patients had hemosiderin-laden macrophages at the site of fenestrations.

Conclusion:PSP in healthy, normal-statured menstruating women has high recurrence and can be life threatening. This may warrant surgical intervention at the initial episode. Further, the diagnosis of CP should be made clinically as pathologic findings of endometrial glands or stroma are rare. Women with clinical CP also have high recurrence despite hormonal suppression and thus should be considered for pleural symphysis at initial presentation.

50.01 Characteristics of Cardiothoracic Surgeons Practicing at the Top Ranked US Institutions for NIH Funding

Posted on January 18, 2016

C. Rosati1, N. P. Valsangkar1, P. N. Vardas1, L. Chabtini1, M. W. Turrentine1, J. W. Brown1, L. G. Koniaris1 1Indiana University School Of Medicine,Department Of Surgery,Indianapolis, IN, USA

Introduction: Academic success of individuals and research funding of their institutions are interrelated. We aimed to investigate which factors distinguish the cardiothoracic surgeons practicing at the top ranked departments from their peers.

Methods: Using online resources (2014 report of the Blue Ridge Institute for Medical Research, SCOPUS, institutional websites, CTSNet) we collected training information (place and timing of medical school and post-graduate education, attainment of a PhD) and academic metrics (publications, citations, H index) of the cardiothoracic (CT) surgeons practicing at the 50 departments of surgery with the highest NIH funding. We divided faculty members into Junior and Senior depending on the time (≤ 20 vs. >20 years, respectively) since graduation from medical school.

Results: Among 593 CT surgeons, 175 (30%) are practicing at the 10 departments (Top 10) with the highest NIH funding, while 418 (70%) at the remaining 40, with a median (range) number of CT surgeons per department of 15.5 (7-34) and 11 (1-24), respectively (p = 0.04), but no significant difference in seniority (junior: 41% of all at the Top 10 vs. 36% at the other; p NS). Senior CT surgeons of the Top 10 are more academically productive than their peers, while no such a difference is evident for junior faculty (see Graphic). CT surgeons of the Top 10 are more likely to have received at least part of their training at one of the Top 10 (66% of all, 72% of junior, and 62% of senior) than their peers (31%, 35%, and 29%, respectively; all p < 0.001). Any prior training at the same institution where they are currently faculty members is more prevalent in junior faculty of the Top 10 (58%) than their peers (40%; p = 0.001), with no such a difference for senior faculty (Top 10: 42%; other: 42%; p NS). No significant difference in the attainment of a PhD is found between the Top 10 (8% of all, 10% of junior, and 6% of senior) and the other institutions (10%, 13%, and 9%, respectively; all p NS), or in the prevalence of international medical graduates (Top 10: 19% of all, 17% of junior, 21% of senior; other: 21% of all, 19% of junior, 22% of senior; all p NS).

Conclusion: While senior cardiothoracic surgeons practicing at the top funded institutions are clearly more academically productive than their peers, no difference emerges for junior individuals. Faculty of the top funded institutions are more likely to have received their training at highly ranked institutions as well, with any prior training at that same institution particularly important in the recruitment of junior faculty. No difference is seen across institutions in the prevalence of surgeons with a PhD or who are international medical graduates.

50.02 Assessment on the Use of Mesenchymal Stem Cells with Transmyocardial Laser Revascularization

Posted on January 18, 2016

J. L. Chan1, Y. Zhou1, M. Li1, J. G. Miller1, S. Wang1, L. C. England2, D. Stroncek2, K. A. Horvath1 1National Institutes Of Health,National Heart, Lung And Blood Institute, Cardiothoracic Surgery Research Program,Bethesda, MD, USA 2National Institutes Of Health,Department Of Transfusion Medicine,Bethesda, MD, USA

Introduction:
Clinical application of stem cell therapy in cardiovascular disease has been performed via intracoronary or intravenous injection, as well as mobilization of bone marrow cells using granulocyte colony-stimulating factor (G-CSF). We report the intramyocardial injection of bone marrow-derived autologous mesenchymal stem cells (MSC) in combination with transmyocardial laser revascularization (TMR) for coronary artery disease patients with depressed left ventricular function.

Methods:
Between April 2012 and May 2015, six patients underwent concurrent TMR and MSC transplantation. To obtain autologous MSCs, bone marrow aspiration was performed three weeks prior to surgery, from which MSCs were subsequently isolated and expanded in vitro. Intraoperatively, MSCs were transplanted into ischemic areas by direct intramyocardial injection (1.5×108 cells) adjacent to TMR channels. Cardiac function was assessed by transthoracic echocardiogram measuring left ventricular ejection fraction (LVEF) and wall motion score index (WMSI) preoperatively as well as three, six, and twelve months postoperatively. Symptomatic improvement was evaluated with the Seattle Angina Questionnaire (SAQ) as well as the Canadian Cardiovascular Society Angina (CCSA) classification system, which were taken at the same time intervals as the echocardiograms.

Results:
Preoperatively, the mean LVEF was 41.26% (95% CI: 36.38%, 46.14%). At three months, LVEF significantly increased by 2.62% (p=0.04), and trended toward continued improvement at six months (+4.24%, p=0.08) and twelve months (+5.22%, p=0.23). Calculated mean WMSI demonstrated superior scores postoperatively at three months (-0.18, p=0.01) compared to baseline (2.16; 95% CI: 1.85, 2.46) and maintains this trend with time (six months: -0.26, p=0.07; twelve months: -0.18, p=0.09). SAQ surveys revealed higher quality of life at 6 months following treatment (+17.64, p=0.02). In comparison to baseline CCSA survey (mean: 3.33, 95% CI: 3.03, 3.70), postoperative CCSA scores indicated significant symptomatic recovery at all time periods (three months: 1.17, p=0.007, six months: 0.83, p=0.003; twelve months: 0.25, p<0.001). With one year follow-up, no major adverse cardiovascular events were observed.

Conclusion:
Intra-myocardial injection of MSCs combined with TMR appears to produce global left ventricular functional improvement as well as symptomatic relief. These findings support the need for additional studies evaluating the efficacy of MSCs during cardiac surgery.

49.10 Do Clinical Fellowships Impact a Department of Surgery’s Academic Productivity?

Posted on January 18, 2016

N. Valsangkar1, P. J. Martin1, J. S. Mayo1, D. V. Feliciano1, T. A. Zimmers1, L. G. Koniaris1 1Indiana University School Of Medicine,Department Of Surgery,Indianapolis, IN, USA

Introduction: The value of subspecialty surgical fellowships is unclear. Fellows might compete with residents for advanced clinical activities but potentially enhance research and clinical excellence within subspecialties. Evidence for the latter was unavailable. Herein we sought to determine the impact of post-residency fellowships on academic output including NIH funding rates in departments of surgery.

Methods: Fellowships offered at the top 55 NIH-funded departments of surgery were identified. Publications, citations and NIH funding history were determined for 4,015 faculty. Mann-Whitney-U test and t-test were used as appropriate.

Results: Cardiothoracic surgery fellowships (CT) were offered at all departments. Other surgical fellowships are offered in 52 of 55 departments (96.4%). Median department publications/citations (P/C) increased with the number of fellowships offered in addition to CT: no fellowship (27±93/437±2509), 1-3 fellowships (34±90/559±3046) and 4 or more (40±97/716±3200, p<0.05). No change in sectional P/C was noted for post-residency bariatric-MIS, colorectal, endocrine, plastic, trauma/CCS/ACS or vascular fellowships. Significant sectional improvements in P/C were observed for breast, HPB, oncology, pediatric, and transplantation (50%-75% increase in P/C versus institutions without fellowships, p<0.05). No differences in divisional or departmental NIH funding rates were observed regardless of the type or number of fellowships offered.

Conclusion: Select fellowships appear to improve the P/C for their respective subspecialty sections. Fellowships have no impact on NIH funding rates. These data suggest that better research is not occurring at many institutions that are producing subspecialists and that clinical training of fellows preempts academic productivity in many subspecialties.

49.11 Use of a Standardized Electronic Tool Improves Compliance Accuracy & Efficiency of Patient Handoffs.

Posted on January 18, 2016

C. Clarke1, S. Patel1, R. Day2, M. Ait Aiss1, G. Avaloa Monetes De Oca1, S. George1, C. Sweeney1, E. Grubbs1, B. Bednarski1, J. Lee1, J. Skibber1, T. Aloia1 1University Of Texas MD Anderson Cancer Center,Surgical Oncology, TX USA 2Mayo Clinic In Arizona,Surgery, AZ USA

Introduction: Following recent changes in work hour regulations, trainee-to-trainee patient care handoffs have become increasingly frequent. Each handoff creates a potential source for communication errors that can lead to near-miss and patient harm events, particularly after complex operations. The purpose of this project was to determine the utility and efficacy of a standardized electronic handoff system.

Methods: At a single cancer center, a task-force of attending surgeons, trainees, and quality improvement personnel was established to improve the trainee-to-trainee patient care handoff process. Root cause analysis and process mapping of the current state was performed to identify specific deficiencies and to document handoff compliance, accuracy, and efficiency. The I-PASS (Illness severity, Patient summary, Action List, Situational awareness and contingency planning, and Synthesis by receiver) methodology was then used to create a standardized electronic program within the REDCap (Research Electronic Data Capture) database frame work. Each on-service trainee was provided a daily email link to the database for structured entry of patient identifiers, acuity, ongoing issues, on-call tasks, and attending preferences. The system was programed to automatically aggregate these data into a single structured worklist for the on-call fellow. Patients identified as higher risk for complications required either a phone call (‘watcher’ designation) or a face-to-face sign out (‘unstable’ designation) in addition to the electronic patient handoff. Pre versus post-implementation handoff compliance (% of patients handed-off and % of night of surgery postoperative checks completed), accuracy (correct data transferred) and efficiency (on-call trainee time to complete worklist) were then compared.

Results: Baseline compliance with handoffs was 72% of patients. Errors in handoff data were found in 12% of unstructured handoffs and the mean on-call trainee time to compile the worklist was 15±11 minutes. During the pilot implementation of the structured electronic tool, 48 handoffs were performed on 49 patients admitted to high acuity surgical floors, increasing the compliance rate to 98%. Post-operative checks were documented in the chart on 27 of 29 eligible patients for 94% compliance. The data error rate dropped slightly to 10%, with most errors related to non-medical data that did not impair the trainee’s ability to care for patients. The mean on-call trainee time to compile the worklist reduced to only 5±3 minutes (p=0.023).

Conclusion: These data suggest that a standardized electronic tool can be used to improve compliance, accuracy and efficiency of handoff communication between surgical trainees. The electronic tool created allows for patient illness stratification and, when used with appropriate face-to-face or phone communication, has helped create a safer environment for patients undergoing complex surgical oncology procedures.

« Previous Page
Next Page »