75.08 Developing patient-centric discharge instructions to prevent readmissions after colorectal surgery

Posted on December 22, 2014

L. Li2, B. W. Trautner1,2, B. A. Campbell1, L. I. Herman1, V. Poppelaars1, D. H. Berger1,2, D. Anaya1,2, S. S. Awad1,2, A. Naik1,2  1Michael E. DeBakey VA Medical Center,Houston HSR&D COIN IQUEST,Houston, TX, USA 2Baylor College Of Medicine,Houston, TX, USA

Introduction: Approximately 16% of major bowel surgeries result in a re-hospitalization. Many of these re-hospitalizations are preventable if warning signs are brought to the attention of physicians at an early stage. A patient-friendly set of discharge instructions, or an after hospital care plan (AHCP), can provide a common language for patients and providers to recognize and communicate about the warning signs in a timely manner. Our purpose was to perform qualitative analysis of patient interviews concerning two components of our AHCP, the warning signs (WS; Figure 1A) and the everyday care instructions (ECI; Figure 1B) to determine whether our materials were communicating the intended information effectively.

Methods: We had previously developed the WS and ECI content with a panel of domain experts. An iterative design approach was used to develop the single-page color format for each document. We then conducted a series of semi-structured interviews with 7 patients who had undergone colorectal surgery within the 2 prior weeks. Six research personnel performed a thematic analysis of interview transcripts through card sorting.  Cluster analysis of themes used the “Jaccard Index” as a metric of distance. Patient response themes were derived from the “clusters” of patient responses (using a 0.33 distance threshold) identified by the research personnel.

Results:

Warning Signs: Patients first noticed the colors, warning signs, and heading. Patients understood that the green zone indicated “that everything is going well,” the yellow zone indicated “things you need to look out for,” and that the red zone indicated that a doctor needed to be seen right away. Patients rated the clarity of the information on this page, from 1 (lower) to 5 (higher), at an average of 4.7 (SD=0.5).

Everyday Care Instructions: The thematic analysis of patient responses to our open-ended questioning about the ECI page revealed that patients first noticed the instructions, and heading. Patients understood that the information on this page indicated that patients need to “Get moving” and that they needed to “Keep up with it.” Patients rated the clarity of the information on this page, from 1 (lower) to 5 (higher), at an average of 5.0.

Conclusion: The patient interviews and thematic analysis indicate that colorectal surgery patients understand the information provided and approve of the manner in which it is presented in the AHCP. Thus, the AHCP has the potential to help reduce preventable hospital readmissions following colorectal surgery through improving early recognition of and attention to early warning signs.

75.19 Fluorescence-based Methodology for Measuring Drug Accumulation in Normal Tissue Versus Tumor

Posted on December 22, 2014

J. Fletcher1, J. Warram1, Y. Hartman1, E. De Boer1, E. Rosenthal1  1University Of Alabama School Of Medicine,Otolaryngology,Birmingham, ALABAMA, USA

Introduction:  Fluorescent-based techniques are being introduced to guide surgical excision of cancer.  Antibodies can be covalently conjugated to near-infrared (NIR) dyes to permit real-time, optical localization of cancer in the surgical setting. Successful antibody-dye combinations are not identified by total tumor accumulation, but by the greatest difference between tumor and normal tissue. While this strategy has the potential to achieve complete resection, the accurate characterization of study drugs is essential to understanding binding kinetics of antibody-NIR dye candidates for this technique. 

Methods:  Using tissues obtained from a dose-escalation clinical trial assessing the safety of cetuximab conjugated to a NIR dye (cetuximab-IRDYe800) in patients with head and neck cancer, a novel methodology was explored to normalize for optical-based attenuation to accurately quantify drug uptake within tissues. Tumor and muscle (n=4) specimens were systematically homogenized and a SDS-PAGE assay was performed on cell lysate (40ug) from each sample. In addition, a serial dilution (0.02ug-0.1ng) of cetuximab-IRDye800 was run on a separate gel to serve as a standard. Using a specialized NIR fluorescent scanner designed to image IRDye800 (Odyssey, LICOR, Lincoln, NE), the gels were imaged and mean fluorescent intensity (MFI) from cetuximab-IRDye800 (150kd) bands were quantified in tumor and patient-matched muscle. MFI were compared to the standard curve to determine percentage of cetuximab-IRDye800 injected dose per gram of tissue (%ID/g).

Results: In the lowest cetuximab-IRDye800 dose group, average %ID/g for tumor (2.8×10-6 MFI) was found to be 8-fold greater than %ID/g for muscle (3.5×10-7 MFI). In the higher dose group (2.5-fold dose increase), average %ID/g for tumor (3.2×10-6 MFI) was found to be 3-fold greater than %ID/g for muscle (1.0×10-6 MFI).

Conclusion: Using this technique, the lower cetuximab-IRDye800 dose was shown to be optimal to provide the greatest difference between tumor and muscle tissue. 

 

74.18 Development of an Approach to Characterize the Complexity of Gastric Cancer Surgery

Posted on December 22, 2014

S. Mohanty1,2, J. Paruch1,3, K. Y. Bilimoria1,4, M. Cohen1, V. E. Strong5, S. M. Weber6  1American College Of Surgeons,Division Of Research And Optimal Patient Care,Chicago, IL, USA 2Henry Ford Hospital,Department Of Surgery,Detroit, MI, USA 3University Of Chicago Pritzker School Of Medicine,Department Of Surgery,Chicago, IL, USA 4Northwestern University Feinberg School Of Medicine,Department Of Surgery, Surgical Outcomes And Improvement Center,Chicago, IL, USA 5Memorial Sloan-Kettering Cancer Center,Department Of Surgery,New York, NY, USA 6University Of Wisconsin School Of Medicine And Public Health,Department Of Surgery,Madison, WI, USA

Introduction:

To allow fair comparisons of hospital quality, most risk adjustment approaches adjust for patient comorbidities and the primary procedure.  However, secondary procedures done at the same time as the index case may increase operative risk and merit inclusion in adjustment models. Including such information could also improve individual patient risk prediction. Our objectives were to evaluate the impact of complexity adjustment on (1)postoperative outcomes, (2)model performance and (3)hospital rankings in gastric cancer surgery. 

Methods:

Using 2007-2012 American College of Surgeons National Surgical Quality Improvement Program data, patients who underwent surgery for gastric adenocarcinoma were identified. Procedure complexity was characterized using secondary procedure CPT© codes and total work relative value units (RVUs).  Regression models were developed to evaluate the association between complexity variables and outcomes. The influence of procedure complexity on model performance and hospital comparisons was examined. 

Results:

Among 3,467 patients who underwent gastrectomy for adenocarcinoma, a secondary procedure was reported for 81.9% of total gastrectomies and 69.6% of partial gastrectomies. The presence of secondary procedures was associated with greater odds for adverse outcomes. For example, patients who underwent a synchronous bowel resection had a higher risk of mortality (OR=2.14, 95%CI: 1.07-4.29) and reoperation (OR=2.09, 95%CI: 1.26-3.47) (Table 1). Model performance was slightly better for nearly all outcomes with complexity adjustment (morbidity c-statistics: standard model, 0.690; RVU model, 0.694; secondary procedure model, 0.701). Hospital ranking did not change significantly after complexity adjustment (mortality, weighted κ 0.84).

Conclusion:

Surgical complexity adjustment improved individual risk prediction but did not appreciably affect hospital rankings. Inclusion of complexity variables into risk prediction tools, such as the ACS NSQIP Risk Calculator, should be considered. 

74.19 Epidemiology of Rectal Cancer Surgeries in the US: 2002-2011

Posted on December 22, 2014

H. Alturki1, S. Fang1,2, S. Selvarajah1, N. Nagarajan1, H. Alshaikh1, F. Gani1, C. K. Zogg1, A. Haider1, E. B. Schneider1  1Johns Hopkins University School Of Medicine,Center For Surgical Trials And Outcomes Research, Department Of Surgery,Baltimore, MD, USA 2Johns Hopkins University School Of Medicine,Department Of Surgery,Baltimore, MD, USA

Introduction: Minimally invasive procedures (laparoscopic and robotic) for the surgical treatment of rectal cancer and have been associated with fewer short-term complications. However, temporal changes in the use of minimally invasive procedures to treat rectal cancer have not been well reported. In this study, changes in the use of minimally invasive techniques to treat rectal cancer were examined across 10 years period. 

Methods:  The Nationwide Inpatient Sample (NIS) from 2002-2011 was examined to identify adults who were surgically treated for a primary diagnosis of rectal cancer using ICD-9 diagnosis and procedures codes. The data were weighted to produce national population-level estimates. Patient and hospital-level characteristics were described by surgical technique: open vs. minimally invasive which included laparoscopic and robotic techniques. Temporal trends for each type of surgical technique were examined. Factors associated with in-hospital mortality, and changes in mortality across the study period were assessed using multivariable regression, controlling for type of procedure as well as patient and hospital-level factors. 

Results: Of 257,994 in-patient who met study criteria, 177,911 (69%) patients underwent open low anterior resection (LAR), 62,889 (24%) open abdominoperineal resection (APR), 9,291 (2%) laparoscopic LAR, 5,227 (3.6%) laparoscopic APR, 2,101 robotic LAR (0.81%) and 575 (0.22%) robotic APR. There was a significant reduction in the proportion of open LAR and APR procedures over the study period, and a concomitant rise in the proportion rectal surgeries performed laparoscopically and robotically (Figure). Minimally invasive procedures were more likely to be performed in male patients, in teaching hospitals, and hospitals located in urban areas (p<.001, all). Over the entire study period, in-hospital mortality was 1.5%; however there was a significant decrease in mortality between 2002 and 2011 (2.0% vs.1.4% respectively, p: <0.05).    

Conclusion: There was a significant increase in the use of minimally invasive procedures over the study period. This is may be due to increasing evidence demonstrating reduced morbidity and mortality among patients undergoing minimally invasive vs. open procedures, as well as increasing surgeon experience with minimally invasive techniques and the development of newer instruments for minimally invasive techniques. However, case presentation, as well as variability in surgeon training and experience, may limit the universal adoption of minimally invasive procedures for the treatment of rectal cancer. 

 

72.11 Worse Outcomes In Patients Presenting With Primary Liver Cancer At Safety-Net Hospitals

Posted on December 22, 2014

A. Mokdad1, A. Singal1, J. Mansour1, G. Balch1, M. Choti1, A. Yopp1  1University Of Texas Southwestern Medical Center,Surgery Oncology,Dallas, TX, USA

Introduction:  Primary liver cancer is the fastest growing cause of cancer related deaths in the United States and affects patients disproportionately in lower socioeconomic classes.  The purpose of this study was to determine if patients with primary liver cancer who present at safety-net hospitals (SNHs) have a different presentation and prognosis compared to those presenting at non-SNHs.

Methods:  We conducted a retrospective analysis of patients with primary liver cancer identified in the Texas Cancer Registry between 1995 and 2010. SNH was defined as a hospital with a Disproportionate Share Hospital index greater than 0.15. Demographics, tumor characteristics, treatment regimens, and survival were compared between patients presenting at SNHs and non-SNHs. Univariate analyses were conducted using student’s t-test and log-rank test. Statistically significant variables were included in a Cox-regression model.

Results: Of the 20,259 patients diagnosed with primary liver cancer, 4,580 (24.3%) presented to SNHs. Patients presenting at a SNH were more likely to be Hispanic (61% vs. 29%, p<0.001) and have lower socioeconomic status (63% vs. 35%, p<0.001) compared to patients presenting at non-SNH. Tumor stage at presentation was similar in both hospital categories; 7,149 (51.9%) presented with local disease, 3,369 (24.5%) with regional spread, and 3,249 (23.6) with metastasis. Patients were 1.5 times (OR = 1.53, P < 0.001) more likely to have a procedure or receive chemotherapy and/or radiotherapy at a non-safety net hospital. The median survival for patients with liver cancer at a SNH was 33 days shorter compared that at non-safety net hospitals, 137 versus 170 days, respectively (P < 0.001). Survival was associated with age, sex, race, tumor stage, socioeconomic status, and treatment. After adjusting for patient demographics and tumor characteristics, patients in safety-net hospitals had an 8%  (HR = 1.08, P = 0.001) increased mortality compared to patients in non-safety net hospitals. This greater risk was rendered non-significant (HR = 1.01, P = 0.82) when reception of a procedure, radiation, and/or chemotherapy was accounted for. 

Conclusion: Patients with primary liver cancer who present to a SNH are more likely to be Hispanic and have a worse overall survival despite similar stage of tumor presentation.

 

72.12 Guideline Adherence Update in Stage II and III Patients Undergoing Colon Cancer Resection

Posted on December 22, 2014

R. L. Hoffman1, K. D. Simmons1, R. E. Roses1, N. N. Mahmoud1, R. R. Kelz1  1Hospital Of The University Of Pennsylvania,Philadelphia, PA, USA

Introduction:
Adherence to guideline-based care for colon cancer patients has been shown to result in decreased costs, shorter durations of inpatient stays, and improved survival. Surgeons play an integral role in the management of colon cancer. The aim of this study was to examine adherence to NCCN guidelines amongst colon resection patients by demographic and clinical characteristics.

Methods:
Patients aged 65-84 years diagnosed with AJCC stage II and III colon cancer who underwent a colon resection were identified in the SEER-Medicare database (2005-2009). High risk (HR) stage II disease was defined as those with a T4 tumor, poor differentiation and <12 lymph nodes examined. Adherence was classified as undertreatment (UT), overtreatment (OT) or concordant care (CC) using stage- and grade-specific NCCN guidelines in combination with chemotherapy and radiation codes from Medicare claims files. Comorbidities were determined using ICD-9 codes from AHRQ predefined category buckets based on diagnoses present in Medicare Outpatient claims at or before the time of colon cancer diagnosis. Descriptive statistics were computed to determine adherence patterns.

Results:
A total of 13017 stage II and III patients who underwent colectomy were identified; 3618 (27.8%) were categorized as stage IIA/B low risk (LR), 3314 (35.5%) as stage IIA/B HR, and 6085 (46.7%) as stage IIIA/B/C. There were a total of 2287 patients ages 65-69, 3,206 ages 70-74, 3,807 ages 75-79 and 3717 ages 80-84 years. Males constituted 43.2% of the cohort (5619). White patients made up 82.5% (10743) and black 9.2% (1200). A total of 8322 (63.9%) patients had one or more comorbidities, with hypertension and iron deficiency anemia being the most common. CC was noted in 6348 cases (48.8%), UT in 5837 (44.8%) and OT in 832 (6.4%). CC was the most likely treatment for stage IIA/B LR patients (87.2%, 3153), patients under the age of 80 (52.3%, 4860), white patients (49.2%, 5280), black patients (47.6%, 571), Asian patients (48.8%, 238) and patients with 1+ comorbidities (52.8%, 4393). Location of residence was not associated with the likelihood of CC.  UT was most common amongst stage IIA/B HR patients (80.1%, 2671), stage III patients (54.2%; 3166), patients ages 80-84 (35.9%; 2098), Hispanics (48.9%; 112), and patients without documented comorbidity (53.1%, 2492). OT was less frequent than UT. Patients over 79 years of age (35.9%) and women (59.3%) were over-represented in the UT group; men (50.8%) and patients ages 70-74 (30.7%) were over-represented in the OT group.

Conclusion:
Stage IIA/B low-risk patients were most likely to receive concordant care. The lower rate of CC among stage IIA/B high-risk and stage III is likely attributable to multiple factors, including clinical judgment based on the strength of evidence presented in the guidelines. Understanding the nuances of treatment adherence across different groups will allow a more targeted focus of efforts to decrease disparities in care.

72.13 Outpatient Thyroidectomy: Current Practice and Utilization Trends in california

Posted on December 22, 2014

J. B. Hamner1, P. Ituarte1, L. Goldstein1, L. Kruper1, S. Chen1, J. Yim1  1City Of Hope National Medical Center,Division Of Surgical Oncology,Duarte, CA, USA

Introduction:   Multiple factors including cost and patient motivation have resulted in an increase in the number of thyroidectomies performed in ambulatory settings.  Information on practice patterns and rates of outpatient thyroidectomy are limited, however.

Methods: The Healthcare Cost and Utilization Project (HCUP) captures inpatient hospital and ambulatory surgery center data. We evaluated HCUP data for 5 years in California to identify all patients undergoing thyroidectomy. Influence of patient type, hospital, disease and surgery (total vs. partial thyroidectomy) characteristics by year and visit type (inpatient vs. outpatient) were determined by bivariate analysis using Chi-Square test and non-parametric test for trend. A multivariate logistic regression model was used to examine predictors of outpatient thyroidectomy.

Results: 37,188 thyroidectomies were identified with 9,319 (25%) conducted in outpatient settings.  Outpatient thyroidectomy accounted for 19.3% of thyroidectomies in 2005, increasing to 30.6% by 2009.  Outpatient thyroidectomy was associated with a lower rate of post-operative complications vs. inpatient (1.17% vs. 4.22%, p<0.01).  Predictors of outpatient thyroidectomy included younger age and fewer comorbidities.  Racial minorities [p-value<0.01] and Medicaid recipients [OR=0.55, p<0.01]  had lower odds of outpatient thyroidectomy.  Hospital volume influenced outpatient thyroidectomy with intermediate volume centers having higher odds of outpatient thyroidectomy than low or high volume centers (OR=1.18 vs. 0.69 and 0.72). 

Conclusion: Outpatient thyroidectomy rates have increased in California, and have fewer recorded complications than inpatient thyroidectomies.  This supports the practice of outpatient thyroidectomy in appropriately selected patients.  Interestingly, racial minorities and Medicaid recipients had lower odds of outpatient care.  These findings suggest that there are multiple systemic factors beyond socioeconomic factors in the selection of patients for outpatient thyroidectomy. 

 

72.14 Geospatial Travel Patterns of Major Cancer Surgery Patients within a Regionalized Health System

Posted on December 22, 2014

A. K. Smith1, N. Shara2,4, A. Zeymo2, R. Estes2, K. Harris1,2, L. Johnson1,3, W. Al-Refaie1,3  1MedStar Georgetown Surgical Outcomes Research Center,Surgery,Washington, DC, USA 2MedStar Health Research Institute,Biostatistics,Hyatsville, MARYLAND, USA 3Georgetown University Medical Center,Lombardi Comprehensive Cancer Center,Washington, DC, USA 4Georgetown University Medical Center,Georgetown-Howard Universities Center For Clinical And Translational Sciences,Washington, DC, USA

Introduction: Regionalization of complex surgeries has led patients to travel longer distances for surgical care. This may be burdensome to vulnerable populations, including older adults and ethnic/racial minorities, who may lack the means to travel long distances. To date, little is known about travel patterns of patients undergoing major cancer surgery in a regionalized care setting. To inform this issue, we sought to map travel patterns among these vulnerable populations who received major cancer surgery within a large, regionalized healthcare system.

Methods: We identified 6,120 patients who underwent lung, esophageal, gastric, liver, pancreatic and colorectal resections from 2002 to 2014 within our large and diverse multi-hospital healthcare system. Patients’ age, race/ethnicity and insurance status were extracted from our electronic health records. We then used Geographic Information System (GIS) software in R to map the distribution of patients’ addresses based on cancer surgery type and vulnerability characteristics. We used visual inspection to assess the distribution and magnitude of travel distances between cancer surgery and each patient characteristic.

Results: 48.2% were non-white, 49.9% were >65 years old and 54.9% of patients had private insurance. Results from the maps showed that for all six oncologic resections, patients over 50 years and whites tend to travel further than younger patients and other racial/ethnic groups for surgery (see figure).Conversely, mapping patients by insurance status did not demonstrate similar geospatial patterns.  

Conclusion: These maps offer a preliminary understanding into variations of geospatial travel patterns to receive major cancer surgery in a regionalized setting. Future research should focus on quantifying differences in travel distances and its impact on timely delivery of surgical care.

 

72.15 Delay in Breast Cancer Diagnosis and Treatment: a Retrospective Review to Identify Risk Factors

Posted on December 22, 2014

N. M. Zaremba1, M. Martin1, A. T. Davis1, P. Haan1, H. L. Bumpers1  1Michigan State University,Department Of Surgery,Lansing, MI, USA

Introduction: It is well known that breast cancer mortality is higher in African American (AA) and young women than in White and older women. This disparity has been attributed to delayed diagnosis, treatment, and follow-up of breast abnormalities. Neither age nor race have been shown to be independent risks for delayed diagnosis or treatment, but socioeconomic status supplants race in some studies. AA women present younger and with later stage disease compared to Whites. Precise factors accounting for these delays have not been defined. This is Phase I of a multi-institutional study to identify risk factors for delays in diagnosis and treatment. Phase II will compare with a large metropolitan hospital caring for the underserved. 

Methods: A retrospective chart review of 107 consecutive patients with a diagnosis of DCIS or invasive breast cancer was performed at a single Midwestern institution where the majority of patients are White and middle class. Delayed diagnosis was defined as 60 days or greater from sign/symptom onset to biopsy and delayed treatment was 60 days or more from tissue diagnosis to first treatment (surgery or neoadjuvant therapy). Income estimates were based on zip code and 2010 US Census. Univariate analysis was performed. This study was approved by the Michigan State University Institutional Review Board. 

Results: Age ranged from 24 to 91 years. Fifteen (14.0%) patients had delayed diagnosis and 10 (9.3%) had delayed treatment. Patients without a family history of cancer had higher risk for diagnostic delay, RR 5.4 (95% CI 1.9-15.8), than patients with a family history of cancer, p =0.001. Similarly, patients without a family history of breast cancer had higher risk for delayed diagnosis, RR 6.9 (95% CI 1.6-29.0) than patients with a family history of breast cancer, p =0.001. As expected, palpable masses and large tumors (T3-4) were associated with delayed diagnosis, RR 4.7 (95% CI 1.4-15.8), p = 0.004 and RR 4.4 (95% CI 1.8-10.7), p =0.014, respectively. Advanced stages (III-IV) were also associated with diagnostic delays, RR 3.1 (95% CI 1.2-8.1), p =0.049. Those with a lower estimated income had significantly delayed treatment, RR 3.8 (95% CI 1.03-13.7), p =0.042. On univariate analysis, none of the following parameters were significantly associated with a delay in diagnosis or treatment: age, race, insurance status, marital status, employment, smoking, menopausal status, or HRT. Diagnostic delays were also due to patients (fear of physicians, misconceptions about cure with healthy lifestyle, personal issues) and physicians (low suspicion, inconclusive imaging, need for medical intervention prior to treatment). 

Conclusion: This group of patients had little delay in diagnosis and treatment. The individual parameters significantly associated with delayed diagnosis and treatment were lack of family history of breast cancer or any cancer, and lower estimated household income. Phase II of this study will be done for comparisons.

72.16 Insurance Disparities in the Treatment and Outcome of Colon Cancer Patients

Posted on December 22, 2014

C. M. Kiernan1, K. Idrees2, N. B. Merchant2, A. A. Parikh2  1Vanderbilt University Medical Center,General Surgery,Nashville, TN, USA 2Vanderbilt University Medical Center,Surgical Oncology,Nashville, TN, USA

Introduction:
Prior studies have demonstrated that uninsured or underinsured cancer patients are more likely to receive substandard treatment and to have worse outcomes. The purpose of this study was to evaluate the impact of type of health insurance on treatment and survival in colon cancer patients, utilizing a large population database.

Methods:
Using the National Cancer Database, we identified 702,892 patients diagnosed with colon cancer from 2003-2011. Patients were stratified into 5 cohorts: Private, Medicare, Military, Medicaid, and Uninsured to test the association of health insurance type with receipt of adjuvant chemotherapy (2003-2011) in patients with stage 3 disease by multivariable logistic regression as well as overall survival in patients with stage 1, 2 and 3 disease by Cox-Proportional regression.  Patients with metastatic disease were excluded.

Results:
Within the cohorts of the Uninsured and Medicaid-insured, a higher proportion of patients were African American (AA) or Hispanic.  Patients in these cohorts were more likely to present with higher stage tumors, were more often treated at academic centers, and were less likely to undergo surgical resection (p<0.001). By multivariable analysis, stage III patients who were uninsured [OR 0.55, 0.51-0.60, p<0.001], on Medicaid [OR 0.51, CI 0.51-0.60, p<0.001], or AA [OR 0.82, CI 0.79-0.85, p<0.001] were less likely to receive adjuvant chemotherapy. By Cox-Proportional regression, lack of insurance [HR 1.64, CI 1.54-1.75, p <0.001], Medicaid [HR 1.80, CI 1.71-1.90, p<0.001], Medicare [HR 1.08, CI 1.05-1.10, p<0.001], military insurance [HR 1.20, CI 1.05-1.37], AA race [HR 1.24, CI 1.20-1.27, p<0.001], as well as stage 2 [HR 1.47, CI 1.43-1.50, p<0.001] and stage 3 [HR 2.35, CI 2.29-2.41, p<0.001] disease were independently associated with worse overall survival.

Conclusion:
In addition to racial disparities, insurance disparities contribute to inequality in the utilization and distribution of health care.  Uninsured and Medicaid patients present with advanced stage disease, appear less likely to undergo surgical resection or receive adjuvant chemotherapy, and have worse overall survival.  Insurance disparities should be further investigated and followed, independent of race, as the healthcare climate in the US changes and evolves.
 

72.17 Feasibility of a Web-based Intervention in Breast Cancer Patients

Posted on December 22, 2014

J. G. Bruce1, N. Steffens3, J. Tucholka3, H. B. Neuman1,2,3  1University Of Wisconsin,School Of Medicine And Public Health,Madison, WI, USA 2University Of Wisconsin,Carbone Cancer Center,Madison, WI, USA 3University Of Wisconsin,Wisconsin Surgical Outcomes Research Program, Department Of Surgery, School Of Medicine And Public Health,Madison, WI, USA

Introduction:  The Internet is a frequently used resource for breast cancer patients seeking information about their diagnosis and treatment options. While the Internet is a convenient information source, access and literacy vary widely. Our objective was to evaluate the feasibility of delivering online information to breast cancer patients in the setting of a clinical trial and assess for disparities in care.

Methods:  We evaluated breast cancer patients (Stage 0-3) participating in a clinical trial investigating the delivery of Internet information prior to surgical consultation. Following their diagnosis phone call, women were emailed links to web-based information. A validated questionnaire was completed prior to their first surgical consult assessing breast cancer knowledge, goals for treatment, time spent reviewing the links, education history, and baseline Internet use. Descriptive statistics were performed.

Results:  To date, 46 women have been approached to participate and 35 have enrolled (76%). Five of the approached women were found to be ineligible and an additional 7 declined due to issues related to access to or comfort with the Internet (n=5) and emotional distress (n=2). Non-participants citing Internet access/comfort issues had a median age of 73 [63-81] years.

Median participant age was 57 (29-78) years. The majority had Stage 1 breast cancer (64% Stage I, 15% Stage 2, 6% Stage 3, 15% DCIS). Two patients did not complete survey data. Of the remaining 33, 65% had at least a college degree, 29% had some college, and 6% had high school or less. Internet use was described as multiple times daily (68%), once daily (23%) or a couple of times per week (9%).

The median time between study email and surgeon consult was 4 (1-13) days; 7 women had ~24 hours prior to the consult. 26 women (79%) reviewed the emailed material (62% for >1 hour, 31% for 15-60 minutes, 7% for <15 minutes). Cited reasons for no review included no access to Internet following receipt of emailed links (n=2), not receiving study email (n=1), insufficient time before surgeon visit (n=1), finding it unnecessary (n=1) and unknown (n=1). No relationship was observed between likelihood to review websites and age, education or Internet use. 

Conclusion:  We determined that delivery of Internet breast cancer information via email is feasible. Disparities in Internet access and literacy are concerns for web-based interventions. However, we were able to successfully enroll 76% of women approached.  Specific challenges included spam filters blocking study emails and short intervals between diagnosis and surgeon consult. Additionally, older patients were more likely to decline participation due to lack of comfort with or access to the Internet, and may require alternative interventions. Final study analysis will determine the impact delivery of this online information has on women’s experience with breast cancer surgery decision-making.
 

72.18 Race/Ethnicity Has No Impact on Outcome for Stage III Breast Cancer

Posted on December 22, 2014

B. P. Townsend1, K. Miller1, Q. Chu1  1LSU Health,Surgery,Shreveport, LA, USA

Introduction: African-American (AA) women have a higher breast cancer mortality rate than Caucasian (C) women, despite having a lower incidence rate. There is a paucity of data examining whether such a disparity exists among patients with stage III breast cancer who were treated at an open access, academic, charity hospital.

Methods: 135 patients with stage III breast cancer were identified from a prospective breast cancer database. All had standardized treatments. Primary endpoint was death from any cause. Overall survival (OS) was calculated from date of diagnosis to date of death or date of last follow-up.  Statistical analyses included log-rank test, Kaplan-Meier survival analysis, and Cox Proportional Hazard Model. A p ≤ 0.05 was deemed statistically significant.

Results: 83 patients (61%) were AA. Comparing the two groups, there were no differences in t-stage distribution (p=0.18), nodal distribution (p=0.73), tumor grade (p=0.43), hormone receptor statuses (p=0.06), and HER-2 status (p=0.81).  The 5-year OS for AA and C was 41% and 44%, respectively (p=0.34). On multivariate analysis, only HER-2 was a predictor of OS (p=0.04).

Conclusion: Race/ethnicity has no impact on outcome of patients with stage III breast cancer who were treated at an open access, academic cancer center.

 

72.19 Disparities in Colorectal Cancer among Different Races in the State of Arizona

Posted on December 22, 2014

M. R. Torres1, H. Aziz1, V. Nfonsam1  1University Of Arizona,Tucson, AZ, USA

Introduction:

Colorectal cancer (CRC) is the third most occurring diagnosed cancer and the third leading cause of cancer related deaths. Despite the efforts to improve preventive methods and treatments, disparities still remain in CRC incidence among different races, particularly focusing on stage at presentation and anatomic location of the cancer. This study assesses the disparity in incidence of CRC among different races in the state of Arizona.

Methods:

 A retrospective analysis of CRC data from the Arizona Cancer Registry (1995-2010) was performed. Races identified from the database were: White, African American, Hispanic, Native Americans, Asian, and Pacific Islander (PI). Early stage colon cancer was defined as stage I and II, while advance stage was defined as III and IV. Using logistic regression modeling, differences in tumor characteristics were assessed among different races in Arizona.

Results:

There were 39,623 reported incident cases of colorectal cancer during the study period. There was a significant decrease in overall incidence of CRC among different races. The Hispanic population presents an increase in incidence of early stage Colon Cancer (3.8/100,000 to 4.7/100,000) and a decrease in the advance stage Colon cancer (10.6/100.000 to 5.6/100,000). The African American population has an increase in the early stage (10.3/100,000 to 11.1/100,000) and a decrease in the advance stage (17.4/100,000 to 15.4/100,000). The Native American population shows an increase in the early stage (4.6/100,000 to 8.7/100,000) and an increase in the advance stage colon cancer (10/100,000 to 10.4/100,000). Asian/Pacific Islanders have an increase in the early stage incidence (2.8/100,000 to 6.3/100,000) and a decline in the advance stage Colon Cancer (10/100,000 to 9.5/100,000).

Conclusions:

Overall CRC incidence is trending down for Whites, Hispanics, African Americans, and Asian/PIs, except for the Native American group.

 

72.10 Insurance Status Impacts Treatment and Survival in Early Stage Pancreatic Adenocarcinoma

Posted on December 22, 2014

E. A. Boevers1, A. M. Button1, B. McDowell1, C. F. Lynch1, S. Bhatia1, J. J. Mezhir1  1University Of Iowa,Surgical Oncology And Endocrine Surgery,Iowa City, IA, USA

Introduction: Previous population studies have shown that insurance status is a significant predictor of survival time for various cancers. This has not been studied in early stage pancreatic ductal adenocarcinoma (PDA), where due to the short survival times, diagnosis at an early stage offers the best chance of survival.  The objective of this study was to determine the impact of insurance status on 1) receipt of multimodality therapy including radiation and surgery and 2) overall survival in patients with early stage PDA.

Methods: Surveillance, Epidemiology and End Results Program data were evaluated for patients diagnosed in the years 2007-2011 with Stages I and II PDA.  Data were analyzed in a multivariate logistic regression model to examine variables associated with receiving either radiation or surgery, and for overall survival in patients with resectable PDA. 

Results:  Patients with Stage I (n = 2,104, 22%), IIA (n = 3,323, 34%), and IIB (n = 4,311, 44%) PDA were evaluated. Most patients (n = 8,231, 85%) were insured, while 1,257 (13%) patients were covered by a Medicaid program and 250 (3%) patients were uninsured. Overall, 32% received radiation therapy and 49% were treated with pancreatectomy.  After adjusting for age and stratifying by stage, a significant difference in the modes of treatment received by patients with differing insurance was found.  Medicaid patients with Stage I PDA were 47% less likely to receive radiation compared to insured patients (p=0.0002),and insured patients were 4.1 times more likely to receive radiation compared to uninsured patients (p=0.004).  A similar pattern was seen in patients with Stage II disease.  Medicaid patients with Stages I and II PDA were significantly less likely to undergo pancreatectomy compared to insured patients, and insured patients with Stage II PDA were 2.2 times as likely to receive surgery than uninsured patients (p<0.0001).

Median survival across types of insurance was significantly different (Medicaid=8 months, Insured=11 months, uninsured=12 months, p<0.0001).  Multivariate analysis with control for age, gender, race, radiation, and surgery showed the risk of death was 1.24 times greater (95% CI 1.2-1.3, p<0.0001) in Medicaid patients versus insured patients.

Conclusion: Treatment with multimodality therapy, including radiation and pancreatic resection, remain remarkably low in patients with early stage PDA.  In this study, insurance status had a measurable influence on the treatment patients received and on overall survival. These data provide initial evidence that certain patients and health care providers may require targeted education and improved access to multidisciplinary care.  More inquiry is needed to determine why this is affecting the Medicaid population, and how these disparities can be remedied in patients with this disease.

70.16 Variations in Metastatic Pattern Among Female Breast Cancer Patients: A US Population Based Study

Posted on December 22, 2014

R. Riccardi1,3, S. Patil1, R. S. Chamberlain1,2,3  1Saint Barnabas Medical Center,General Surgery,Livingston, NJ, USA 2University Of Medicine And Dentistry Of New Jersey,Newark, NJ, USA 3Saint George’s University,Grenada, Grenada, Grenada

Introduction: Worldwide statistics reveal that approximately 30% to 70% of female primary breast cancer patients will ultimately develop recurrence and succumb to metastases. The site of recurrence is intimately linked to survival, as visceral metastases patients generally succumb to disease early, while bone metastases patients can live years if properly treated.  Advances in chemotherapy therapy drugs and regimens including the addition of targeted and immunotherapy therapy has yielded dramatic results in regards to prolonging survival for metastases to certain metastatic sites.  This study aimed to patterns of metastatic spread among different age and ethnic groups and to more precisely determine metastatic site specific survival in metastatic breast cancer from a recent large nationwide cohort.

Methods: Data on all patients with breast cancer were abstracted from the Nationwide Inpatient Sample (NIS) database (2002-2009). Patients with MBC were identified using ICD-9 codes 1960-19889 (Table 1). Four ethnic groups (Caucasian, African American, Hispanic and Others), Two age groups (<50 years and >50 years) and two outcome groups (Alive and Dead) were compared for differences in metastatic patterns. Standard statistical methods were used for data analysis. 

Results: 708,423 patients were analyzed for incidence of metastatic breast cancer. Overall metastases to axillary lymph nodes occurred in 24.3%, bone in 4.6%, and respiratory tract in 2.8%. Both lymph node and extralymphatic metastases were highest among African Americans compared to other ethnic groups. Similarly Hispanics had the highest percent of lymph node and extralymphatic metastases compared to Caucasians except for skin, peritoneum and adrenal metastases. Extra lymphatic metastases both to solid and hollow viscera were more common in patients < 50 years except for liver, brain and ovary. Mortality occurred in 32.9% of patients with metastases to the large intestine, 23.8% to the respiratory tract, 23.9% to the kidney, and 23.7% to the GIT NEC. On multivariate analysis metastases to inguinal lymph nodes, large intestine, liver, respiratory tract, brain, bone, peritoneum, skin, gastrointestinal tract NEC and to the thoracic lymph nodes was independently associated with increased mortality.

Conclusion: Metastatic breast cancer, especially extralymphatic metastases is more common in the younger population. Metastases to atypical sites like the inguinal LN and large intestine is associated with the highest mortality. Increased knowledge of metastatic site specific survival, should be useful to improve breast cancer staging (e.g., perhaps with the addition of M1bone, M1viscera, M1brain), and to better stratify clinical trial enrollment. 

 

70.17 Robotic Assisted Esophagectomy in the Obese Patients

Posted on December 22, 2014

A. Salem1, M. Thau2, K. Meredith1  1University Of Wisconsin,Section Of Surgical Oncology – Division Of General Surgery – ‏The University Of Wisconsin School Of Medicine And Public Health,Madison, WI, USA 2University Of South Florida College Of Medicine,Tampa, FL, USA

Introduction: The impact of body weight on outcomes after robotic-assisted esophageal surgery for cancer has not been studied. We thus examined short-term operative outcomes in patients according to their body mass index (BMI) following robotic-assisted Ivor-Lewis esophagogastrostomy (RAIL) at a high-volume tertiary-care referral cancer center and evaluated the safety of robotic surgery in patients with an elevated BMI.

Methods: A retrospective review of all patients who underwent RAIL for pathologically confirmed distal esophageal cancer was conducted. Patient demographics, clinicopathologic data, and operative outcomes were collected. Current guidelines from the US Centers for Disease Control and Prevention and from the World Health Organization define underweight BMI range as below 18.5 kg/m².  Normal range is defined as a BMI range of 18.5 to 24.9 kg/m². Overweight is defined as a BMI range of 25.0 to 29.9 kg/m² and obesity is defined as a BMI of 30 kg/m² and above. Statistics were calculated using Wilcoxon Rank-Sum and Spearman Coefficient tests with a p-value of 0.05 for significance.

Results:134 patients (106 men, 28 women) with an average age of 66 years were included. The majority of patients, 76% (N=102) received neoadjuvant therapy. When stratified by BMI, five patients were underweight, twenty-nine were normal weight, fifty-eight were overweight and forty-two were obese. All patients had R0 resection. Median operating room (OR) time was 421 minutes and median estimated blood loss was 160 cc. BMI significantly correlated with longer OR time (coefficient= 0.23; p=0.01) and higher EBL (coefficient=0.17; p=0.05).

Conclusion:This retrospective study shows that patients with distal esophageal cancer and an elevated BMI undergoing RAIL have increased operative times and EBL during the procedure.
 

70.18 Should We Operate for an Intra-abdominal Emergency in the Setting of Disseminated Cancer?

Posted on December 22, 2014

C. L. Scaife1, K. C. Hewitt1, X. Sheng2, K. W. Russell1, M. C. Mone1  1University Of Utah,General Surgery / Surgery,Salt Lake City, UT, USA 2University Of Utah,Biostatistics / Pediatrics,Salt Lake City, UTAH, USA

Introduction:
Patients with an advanced cancer diagnosis, who develop an intra-abdominal surgical emergency, pose a medical decision making dilemma. While patient survival is already limited due to advanced malignancy, a surgical intervention may be futile and costly.  Patients, families, and medical providers need more accurate data to assist with these difficult treatment decisions. This study was designed to establish morbidity and mortality rates, and to attempt to identify preoperative risk factors which may predict a poorer outcome.   

Methods:

The national NSQIP database was queried for patients with disseminated cancer undergoing emergent abdominal surgery (2005-2012).  Preoperative NSQIP variables were used for prediction models for 30-day major morbidity and mortality. Example NSQIP variables that were used included patient functional status, ASA class, weight loss, renal function, age, albumin, sepsis, HCT, and pre-op cardiopulmonary comorbidities.  A tree model and logistic regression were employed to find factors associated with these outcomes. The analysis was carried out on a training dataset; model performance (misclassification rate) was then evaluated on a validation dataset.  

Results:

In patients with an abdominal surgical emergency and disseminated or Stage IV cancer, there was overall a major surgical morbidity rate of 47% and a surgical mortality rate of 26%. The tree model for morbidity showed that sepsis or a hematocrit <29.6 was predictive for a major morbidity (error rate 36%). The tree model for mortality showed an ASA score of 4 or 5 with a totally dependent functional status to be predictive of mortality (error rate 24%).

Conclusion:

The decision to operate for an intra-abdominal emergency in the setting of disseminated cancer is difficult when considering the morbidity to the patient, the overall expected cancer specific and post-operative expected survival, and cost. Our study confirms that the risk for surgical morbidity and peri-operative death in this population is very high. Preoperative patient factors, including sepsis, ASA class, anemia, and patient functional dependence, all strongly predict poor patient outcomes. We have also developed further logistic regression models, derived from this database, to provide detail to help with decisions related to care.

70.19 Sarcopenia as a Predictor of Outcomes of Palliative Surgery

Posted on December 22, 2014

A. M. Blakely1, S. Brown2, D. J. Grand2, T. J. Miner1  2Brown University School Of Medicine,Department of Radiology,Providence, RI, USA 1Brown University School Of Medicine,Department Of Surgery,Providence, RI, USA

Introduction:

Sarcopenia, defined as the degenerative loss of muscle quality with age, has been shown to be associated with worse outcomes following resection of various cancer types. Sarcopenia is usually assessed by evaluating psoas muscle density and/or cross-sectional area at the level of the third or fourth lumbar vertebra. To date, sarcopenia has not been evaluated as a predictor of outcomes following palliative surgery.

 

Methods:

Retrospective analysis of a prospective database of all palliative surgery patients receiving an operation from January 2004 to December 2012 was performed. Those patients with a pre-operative abdominopelvic computed tomography (CT) and follow-up of at least 6 months were selected for analysis. CT scans were evaluated by a resident and attending radiologist for mean total psoas muscle cross-sectional area and density. Time from CT to operation, primary tumor type, and post-operative complications were recorded.

 

Results:

From January 2004 to December 2012, 57 patients were identified as having undergone a palliative procedure with prior abdominopelvic CT scan available for evaluation.  Of 57 patients, 27 (47.4%) were male, with median age of 63 years. Median time from CT scan to operation was 10 days. Regarding primary tumor type, 15 (26.3%) were pancreas, 9 (15.8%) colon, 7 (12.3%) stomach, 4 (7.0%) lung, 3 (5.3%) ovary, 3 (5.3%) cholangiocarcinoma, and 16 (28.1%) other. Thirty-day post-operative morbidity was 31.6% and mortality was 5.3%. Mean total psoas area was 1622.8 mm2 (range 785.3 to 3641.3 mm2, standard deviation (SD) 586.6), and the mean psoas density was 49.8 Hounsfield units (HU) (range 27.1 to 69.8 HU, SD 9.9). Increasing age was associated with decreased total psoas area (p=0.0027) and decreased psoas density (p=0.0005). However, neither total psoas area nor density were associated with the development of complications (p=0.88 and p=0.48, respectively).

 

Conclusions:

Sarcopenia, assessed by either psoas muscle area or density, was not associated with complications following palliative surgery. Patient selection for palliative surgery continues to be challenging. Sarcopenia may retain some value in the pre-operative assessment of advanced cancer patients as a component of a frailty index.

70.20 Secular Trends in Morbidity and Mortality Among Surgical Patients with Disseminated Malignancy.

Posted on December 22, 2014

S. Bateni1, F. J. Meyers3, R. J. Bold2, R. J. Canter2  1University Of California – Davis,General Surgery,Sacramento, CA, USA 2University Of California – Davis,Surgical Oncology,Sacramento, CA, USA 3University Of California – Davis,School Of Medicine,Sacramento, CA, USA

Introduction: There is substantial risk of acute morbidity and mortality following surgical intervention for patients with disseminated malignancy. However, concerns also exist regarding the risks of untreated surgical disease among these patients. We sought to characterize temporal trends in morbidity and mortality among disseminated malignancy patients, hypothesizing that surgical intervention would remain a prevalent modality among these patients.

Methods: We queried the American College of Surgeons National Surgical Quality Improvement Program from 2006 to 2010 for patients with disseminated malignancy. Excluding patients undergoing a primary hepatic operation, we identified 21,755 patients. Parametric and non-parametric statistics were used to evaluate the association of patient characteristics and surgical interventions to 30-day morbidity and mortality outcomes. Logistic regression analysis was performed to identify independent predictors of 30-day morbidity and mortality.

Results: The prevalence of surgical intervention for disseminated malignancy was stable at 1.9 – 2.2% of all NSQIP procedures per year. Among disseminated malignancy patients, the most frequent operations performed were bowel resections (24.7%), varied gastrointestinal procedures (22.0%), and multivisceral resections (13.7%). The rates of bowel resection, celiotomy/lysis of adhesions and appendectomy/cholecystectomy, showed small, but statistically significant, decreases over time (26.1 vs 22.6%, 8.4 vs 5.8%, 6.6 vs 2.9% respectively, p<0.001). The rate of emergency operations also decreased over the study period  (17.4 vs 15.0%, p<0.005). In contrast, the rate of preoperative independent functional status rose (82.3 vs 86.1%, p<0.001), while the rate of preoperative weight loss (14.4 vs 12.8%) and sepsis (20.6 vs 15.7%) decreased (p<0.005). Rates of 30-day morbidity (30.1 vs 23.5%), serious morbidity (16.1 vs 10.6%), and mortality (10.4 vs 9.3%) all decreased over the study period (p<0.05). On multivariate analysis, male sex, BMI, age, impaired functional status, weight loss, pre-operative sepsis, leukocytosis, elevated creatinine, anemia, and hypoalbuminemia all predicted worse 30-day morbidity and mortality. A lack of DNR status was associated with greater morbidity, while a present DNR status was associated with higher mortality.

Conclusion: Although 30-day morbidity, serious morbidity, and mortality have decreased for patients with disseminated malignancy undergoing surgical intervention, they remain elevated, and surgical intervention remains prevalent in patients with incurable malignancy. These data highlight the importance of careful patient selection and an evaluation of the goals of therapy among this patient population.

70.06 Hypertension Risk among Cancer Patients Treated with Sunitib: a Systematic Review and Meta-analysis

Posted on December 22, 2014

S. Lew1,3, R. S. Chamberlain1,2,3  1Saint Barnabas Medical Center,Department Of Surgery,Livingston, NJ, USA 2New Jersey Medical School, Rutgers University,Department Of Surgery,Newark, NJ, USA 3Saint George’s University,School Of Medicine,True Blue, , Grenada

Introduction:  Sunitinib is a multi-targeted tyrosine kinase inhibitor widely used in cancer therapy, which has been linked to varying degrees of treatment related hypertension (HTN). Current publications contain wide variations in the incidence and severity of sunitinib-related HTN. In addition, the HTN risk uniquely associated with two current mainstay sunitinib regimens, 37.5mg continuous dosing and 50mg 4 weeks on, 2 weeks off dosing, has not been well established.  This study sought to determine the incidence of sunitinib-associated severe HTN, to stratify HTN risk based on the type of malignancy treated, and to investigate the dosage related risk of severe HTN.

Methods:  A comprehensive literature search of PubMed, Google Scholar and the Cochrane Central Registry of Controlled Trials was completed. Keywords searched were ‘sunitinib’, ‘sutent’, ‘SU11248’, ‘hypertension’, and ‘clinical trial’. All clinical trials were analyzed for patient recruitment, intervention, and outcomes. Incidence and risk ratio (RR) were calculated with 95% confidence intervals.

Results: 61 single or double arm, phase II/III clinical trials involving 6,813 patients treated with sunitinib were identified. The incidence of sunitinib associated severe (grade 3-4) HTN was 6.5% (95% CI [4.9-8.6]; p<0.0001) among 4,311 patients in 50 clinical trials in which sunitinib monotherapy was used in a single arm trial or RCT.  Among 4,433 Sunitinib treated in 18 RCTs, the risk of severe HTN in the treatment group was 5% compared to 2% in the control (RR 2.82, 95% CI [1.63-4.88]; p=0.0002). The risk of sunitinib associated severe HTN in RCC patients was not higher than the control (RR 1.38, 95%CI [0.64-2.96]; p=0.41) but it was significantly different from all other malignancies (p=0.05). The incidence and risk of severe sunitinib-associated HTN were not significantly different among breast cancer (BC) and non-BC, Gastrointestinal stromal tumor (GIST) and non-GIST, between 37.5mg continuous and 50mg intermittent dosage regimens, or when monotherapy and/or concomitant chemotherapy were used. Significant heterogeneity was found among identified trials with regards to underlying malignancies, dosage of the treatment, and duration of the treatment.

Conclusions: Sunitinib treatment is associated with a significantly increased risk of all- and high-grade HTN. The differences in dosing regimens or combinational chemotherapy with sunitinib did not yield significant hypertension risk reduction. Although adequately powered large studies are needed to further investigate the contributing HTN risk factors and ideal management or prevention, blood pressure should be carefully monitored in patients treated with sunitinib to prevent cardiovascular complications.

Next Page »