17.14 An Emerging Public Health Concern: Dialysis Patients Returning to the OR After Common Procedures

Posted on December 22, 2014

K. Brakoniecki1, S. Tam1, P. Chung2, A. Alfonso2, G. Sugiyama2  1SUNY Downstate College Of Medicine,Brooklyn, NY, USA 2SUNY Downstate Medical Center,Surgery,Brooklyn, NY, USA

Introduction:
Improvements in care of patients with end-stage renal disease (ESRD) have resulted in decreased mortality rates since 2001.  A greater number of those on dialysis are living longer and undergo common general surgery procedures. The few studies which have investigated surgical outcomes in patients with ESRD have shown rates of return to the operating room (OR) nearly 3 times of those not on dialysis.  Subsequently returning to the OR is associated with postoperative mortality and morbidity. Our objective was to assess the morbidity and mortality of patients with ESRD undergoing the most common general surgery procedures, and determine the risk factors for returning to the OR.

Methods:
Data was extracted from the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database. Patients that underwent appendectomy, cholecystectomy, ventral hernia repair, and colectomy from 2005-2010 were selected by CPT codes and then separated based on dialysis status. Matched cohorts based on age, gender and procedure type, were created randomly using SPSS. Multivariate logistic regression was used to analyze the effect of dialysis on mortality, morbidity, and returning to the OR. The dialysis group was further stratified into “return to OR” and “no return to OR”, and compared in terms of pre-operative and surgical profiles. Multivariate logistic regression analysis was performed to analyze the effect of returning to the OR on postoperative mortality and morbidity in dialysis patients.

Results:
A total of 195,585 patients underwent the described procedures, of which 1,163 dialysis patients were identified, and matched with non-dialysis patients.  Dialysis was associated with a higher risk of returning to the OR, (odds ratio [OdR] 2.97 [1.99-4.46]), mortality (OdR 9.05 [4.09-20.00]), and morbidity (OdR 1.66 [1.29-2.13]). Of the dialysis patients, 94 (8%) returned to the OR, and return to the OR was associated with increased mortality (OdR 4.35 [2.11-8.99]) and morbidity (OdR 7.62 [4.68-12.41]). Those with post-operative infection complication were over 8 times (OdR 8.23 [4.92-13.75]) more likely to return to the OR.

Conclusion:
Using the ACS-NSQIP database we describe increased rates of return to the OR in dialysis patients, for the most common general surgery procedures. Furthermore, within this population, returning to the OR is associated with significantly increased morbidity and mortality. Surgeons should be aware of the risk of returning to the OR, and subsequently the high risk of mortality and morbidity in dialysis patients. Further study is needed to elucidate preventable risk factors to protect patients from this emerging public health issue.
 

17.15 Developing a Tiered Method to Link Health Information Technology Attributes with Patient Outcomes

Posted on December 22, 2014

M. A. Zapf1, A. Kothari1, P. Kuo1, G. Gupta1, P. Wai1, J. Driver1, T. Markossian2  1Loyola University Chicago Stritch School Of Medicine,Surgery,Maywood, IL, USA 2Loyola University Chicago Stritch School Of Medicine,Public Health Sciences,Maywood, IL, USA

Introduction: Linkage of large administrative datasets containing information at the hospital and patient levels offers the opportunity to conduct timely research on the impact of organizational characteristics including health information technology (HIT) on surgical outcomes. We aimed to develop a methodology to create a tiered dataset to study the impact of HIT on surgical outcomes.

Methods: One patient-level dataset was linked with two hospital-level datasets for the years 2007-2009. These were: 1) the Healthcare Cost and Utilization Project – State Inpatient Database (HCUP-SID) from Florida containing patient-level hospitalizations including surgical outcomes, 2) the Dorenfest Institute for Health Information Technology (HIMSS) database that contains data about hospitals’ HIT adoption and use, and 3) the American Hospital Association (AHA) Annual Survey which contains detailed hospital information. The goal was to maximize the number of hospitalizations in HCUP-SID having hospital IT information. Hospital unique identifiers were used to link the HCUP-SID to the AHA database. The AHA database was linked to the HIMSS data via hospital Medicare number. After primary linkage, manual matching with ZIP code and hospital name increased the number of hospitalizations in HCUP by including observations that did not initially match with the HIMSS.

Results:Hospitalizations were generated from 247 (2007) and 246 (2009) hospitals in the HCUP-SID. Exactly 196 (79.3%, 2007) and 206 (83.7%, 2009) hospitals were directly matched. After manual matching, the numbers increased to 211 (85.4%, 2007) and 220 hospitals (89.4%, 2009). In the final dataset, hospital-level IT characteristics from 2,486,167 of 2,563,383 (97.0%) and 2,502,342 of 2,606,165 (96.0%) hospitalizations were identified. Of these, manual merging was responsible for linking 36,880 (1.5%, 2007) and 33,282 observations (1.3%, 2009). Manually merged hospitals had a smaller number of hospitalizations per hospital compared to directly matched (2,459 v 12,890 in 2007 & 2,377 v 11,985 in 2009, both p<0.05). In the final database, the number of common general surgery operations (appendectomy, hemorrhoidectomy, cholecystectomy, inguinal hernia, and thyroidectomy) tallied 64,213 and 65,942, while complex operations (colorectal resection, gastrectomy, esophagectomy, and kidney/liver transplant) were 25,812 and 25,917 in 2007 and 2009. In total, the linked databases contain over 100 patient-level variables and 1,288 possibly clinically associated hospital-level characteristics.

Conclusion:We demonstrated the feasibility of creating a tiered database using the HCUP-SID, HIMSS, and AHA Annual Survey datasets with a high match rate and minimal lost patient encounters. Manual merging was essential for capturing lower volume hospitals. Using the same approach, additional datasets at the hospital or area levels could be appended to our dataset with the goal of expanding our analytical scope.

17.16 Feasibility of a Symptom Tracking Smartphone Application

Posted on December 22, 2014

A. R. Scott1,2, G. K. Low1, A. D. Naik1,2, D. H. Berger1,2, J. W. Suliburk1  1Baylor College Of Medicine,Houston, TX, USA 2VA Center For Innovations In Quality, Effectiveness And Safety,Houston, TX, USA

Introduction:
Limited communication and care coordination following discharge may contribute to surgical complications. Smartphone applications (“apps”) offer a new mechanism for communicating with patients and directing their care. It is unclear, however, whether or not patients are willing and able to use apps as part of their surgical care. To better understand patient factors which could prevent app use in a surgical setting, we performed a feasibility study on an app designed to facilitate self-care following colorectal surgery.

Methods:
This was a prospective mixed-methods feasibility study performed at an urban public safety net hospital. Following colorectal surgery, patients were approached for enrollment and offered a smartphone app which uses previously validated content to provide recommendations based on symptoms. Patients were asked to use the app daily for 14 days after discharge. Demographics and usability data were collected at enrollment. The System Usability Scale (SUS) was used to measure usability. The SUS was repeated at follow up and then we performed a structured interview covering domains such as ease of use, willingness to use, and utility of use. Chart and app log review identified phone calls and ER visits related to surgery.

Results:
We screened 75 patients, enrolled 14 (19%), and completed follow up interviews with 10 (13%). Reasons for non-enrollment included: lack of a suitable device (16 patients, 21%), willingness to participate (14, 19%), language barriers (12, 16%), inclusion criteria (11, 15%), and other reasons (8, 11%). The unplanned ER visit rate was 43% (6/14), with a 14% (2/14) readmission rate. The app addressed 67% (4/6) of the presenting complaints in the ER, but no patients reported those complaints in the app. The app was used once or not at all by 4/14 patients (29%); the remaining 10 (71%) used it a median of 7 times (6-13). SUS scores were >90th percentile at first use (raw score 94, IQR 86-96) and follow up (88, 83-95). Four patients who were interviewed (40%) reported daily app use. Feeling ill was the most common (3/6, 50%) reason for less frequent use. All 10 patients interviewed reported being able to fill out the app themselves, 9 denied difficult or confusing questions, 8 felt the app fit into their daily routine, and 1 felt there were too many questions. Six felt the app’s recommendations could be trusted, 3 weren’t sure, and 1 did not trust the app.

Conclusion:
Smartphone based interventions have the potential to improve care coordination and patient perceptions of communication. Patient barriers to app use include device availability, interest, compliance when feeling ill, and incomplete or inaccurate symptom reporting. Use patterns in this study fell short of goals outlined at enrollment, suggesting the need for highly engaging apps. Further study is needed to find ways to overcome these barriers as well as methods of integrating apps into surgical care pathways.

17.17 Assessment of the Quality of Google Glass Images For Burn Wound Assessment

Posted on December 22, 2014

P. H. Chang1,2,3, P. H. Chang1,2,3  1Shriners Hospitals For Children-Boston,Boston, MA, USA 2Massachusetts General Hospital,Boston, MA, USA 3Harvard School Of Medicine,Brookline, MA, USA

Introduction:   Wearable technology has emerged as a new source of medical devices.  Google Glass has been trialed by several surgeons of various specialties for its unique combination of video and photographic recording and transmission, portability, and hands-free use which offers obvious advantages for the surgeon who is sterilely scrubbed in.  Our department sought to assess the use of Google Glass as an image capturing device for assessment of burn wounds especially with regards to quality of images and ease of use. 

Methods:
A retrospective analysis of pediatric burn patients admitted between May 2014 and June 2014 was performed.  Inclusion criteria included all burn patients with at least 1% TBSA and a minimum depth of injury of partial thickness.  Photographic consent was obtained from all parents of patients as per hospital regulations.   The Glass device was used to take pictures of the patient's burn wounds.  The hospital photographer then took pictures using a Nikon D7100 24 MP camera.  The two images were compared side by side by an experienced burn surgeon and assessed as to the quality of image and ability to make an accurate diagnosis based on the images.  Notes were taken by the Glass user as to issues that arose while taking the pictures.

Results:
5 patients had burn wounds assessed in this time period using both the Google Glass and the traditional hospital photographer.  Despite the lower resolution of the Glass device (5 MP), 5 of 5 patients' Glass images were assessed by the experienced burn surgeon to be of adequate quality to provide the same information as the higher resolution pictures obtained by the hospital photographer.  The hospital photographer was able to obtain pictures faster on average (approximately 1 second per image) compared to the Glass device (approximately 5 seconds per image).   Issues that arose from use of the Glass included:

1)  overheating and subsequent shutdown of the Glass device while in the heated environment of the burn operating room.
2)  the lack of ability to focus the Glass lens and thus require the Glass user to be in close proximity to the burn wounds to obtain optimal pictures

3)  the need to take extra care to disable automatic upload of images to Google Plus account to prevent transfer of sensitive patient information to Google servers which are by definition non HIPPA compliant.

Conclusion:
The Google Glass device is able to capture clinically accurate images of burn wounds.  However, there are limiting factors to the technology as it currently stands that would need to be addressed before it could be fully utilized in a burn surgery practice.
 

17.18 What Happened Last Night?! – Variability in Night Shift ICU Care

Posted on December 22, 2014

J. Driver1, P. Y. Wai1, M. A. Zapf1, A. Kothari1, K. Y. Wolin1, P. C. Kuo1  1Loyola University Chicago Stritch School Of Medicine,Maywood, IL, USA

Introduction: ICU patients comprise the sickest patient population in the hospital. They are presumed to receive unwavering "around the clock" care. However, this assumption has not been previously investigated and anecdotal observations suggest that night time care is variable.  To determine potential differences in night time ICU care, patient care parameters were analyzed comparing hourly data from day, evening and night shifts. We hypothesized that variability in ICU care occurs during the night shift and impacts important patient outcomes. 

Methods: EPIC electronic medical record data from 15,493 patients in 5 ICUs from 2008-2013 at a major urban academic medical center were retrospectively analyzed for hourly urine output (U/O), mean arterial pressure (MAP), frequency of MD and RN EMR access and total fluid output during day, evening and night shifts. Variation in hourly U/O was selected as a surrogate marker for overall attention to care.  ICUs included: CCU, NeuroICU, CardiothoracicICU, MICU, and SICU. Mean night shift values were compared to the mean combined day and evening shift values. Statistical analysis was performed using paired t-tests or linear mixed effect modeling; p values < 0.05 were considered significant.

 

Results: There was reduced MAP (-0.75 mmHg/hr*) and reduced U/O (-18.9 mL/hr*) during the night shift. Paradoxically, frequency of care giver EMR access was significantly decreased at night (-33.4 times/hr*) and correlated with decreased U/O*, increased length of stay* and increased overall in-hospital mortality*. The model of resident and attending MD coverage and ICU specialty did not correlate with these parameters. (*p<0.0001) 

Conclusion: Our results demonstrate that attentiveness (measured by frequency of EMR access) correlated with surrogate care parameters (U/O) and outcome measures (length of stay and mortality). We conclude that variations in night shift ICU care may be due to caregiver inattention. Corrective strategies to increase patient monitoring, such as scheduled night shift ICU team rounding, should be identified.  

15.15 The publication gender gap in academic surgery

Posted on December 22, 2014

R. C. Wright1, C. Mueller1  1Stanford University,Pediatric Surgery,Palo Alto, CA, USA

Introduction:  Terms such as “glass ceiling” and “sticky floor” are still commonly used to describe the role of women in academic surgery, and multiple studies have documented disparities between men and women in the field.  In spite of the awareness and continued efforts to alleviate this gap within surgery, gender inequalities remain. 

Methods:  In this investigation the researchers examined the differences in published literature by male and female academic surgeons according to amount and impact. Websites for departments of surgery of three large academic centers were reviewed. Only full-time faculty were included in the analysis. Surgeons’ gender and academic rank were determined by their online biographies. Over a two week span all H-indexes, number of articles published, and other bibliometrics were determined using the Web of Science database. 

Results: A one-way ANOVA showed a significantly higher H-index for men than women (p<.05). In addition, one-way ANOVA showed significantly more articles published by men than women (p<.05). These differences are most dramatic at the rank of associate professor where the H-index for men is three times that of the women. The rank of full professor showed men had double the number of articles published. 

Conclusion: These findings align with previous research which shows a disparity between males and females as they climb the academic ladder. Since publishing research articles is a vital part of advancement in academic medicine, gender disparities in this realm may have major effects on the promotion process.  Future investigations may focus on the reasons behind this publication disparity.

 

15.16 Readmissions Following Major Cancer Surgery in Older Adults Within a Large Multihospital System

Posted on December 22, 2014

R. C. Langan1,2, C. Huang3, K. Harris1,2,3, S. Colton1, A. L. Potosky2,3,4, L. B. Johnson1,2,3,4, N. M. Shara2,3,5, W. B. Al-Refaie1,2,3,4  1Georgetown University Hospital,Department Of Surgery,Washington, DC, USA 2MedStar-Georgetown Surgical Outcomes Research Center,Washington, DC, USA 3MedStar Health Research Institute,Washington, DC, USA 4Lombardi Comprehensive Cancer Center,Washington, DC, USA 5Georgetown-Howard Universities Center For Clinical And Translational Sciences,Washington, DC, USA

Introduction:  Readmissions are a focus of emerging efforts to improve the quality and affordability of healthcare. Yet, little is known about reasons for readmissions after major cancer surgery in the expanding elderly population (≥ 65 years) who are also at increased risk of adverse operative events. We sought to identify 1) the extent to which older age impacts readmissions and 2) factors predictive of 30- and 90-day readmissions after major cancer surgery among older adults. 

Methods:  We identified 2,797 older adults who underwent seven types of major thoracic or abdomino-pelvic cancer surgery within a large multihospital system from 2003-2012.  Multivariate logistic regression analyses were conducted to identify predictors of 30- and 90-day hospital readmission. 

Results: Overall 30-day and 90-day readmission rates were 16% and 24% with the majority of readmissions occuring within 15-days of discharge. Principal diagnoses of 30-day readmissions included gastrointestinal, pulmonary and infections complications. 30-day readmissions were associated with > 2 comorbid conditions and ≥ 2 postoperative complications. Readmissions significanctly varied according to cancer surgery type and across treating hospitals. Readmissions did not vary by increasing age. Factors associated with 90-day readmission were comparable to those observed at 30-days (Table 1). 

Conclusion: In this large multi-hospital study of older adults, multi-morbidities, procedure type, increased complications and the treating hospital predicted 30- and 90-day readmissions. These findings point toward the potential impact of hospital-level factors behind these readmissions. Our results also point towards the importance of assessing the influence of readmission on other important cancer care metrics; patient reported outcomes and the completion of adjuvant systemic therapies.

 

15.17 The Readability, Complexity, and Suitability of Online Patient Material for Breast Reconstruction

Posted on December 22, 2014

C. R. Vargas1, P. Koolen1, D. J. Chuang1, B. T. Lee1  1Beth Israel Deaconess Medical Center,Surgery / Plastic And Reconstructive Surgery,Boston, MA, USA

Introduction:
Limited health literacy affects nearly half of American adults and has been shown to adversely affect patient participation, satisfaction, healthcare costs, and overall outcomes.  As unprecedented numbers of patients search the internet for health information, the accessibility of online material is more important than ever before.  The aim of this study was to evaluate available breast reconstruction resources on the internet with regard to reading grade level, degree of complexity, and suitability for the intended patient audience using three validated tools.

Methods:
The ten most popular patient websites for "breast reconstruction" were identified using the largest internet search engine.  The content of each site was assessed for readability using the Simple Measure of Gobbledygook (SMOG) analysis, complexity using the PMOSE/iKIRSCH formula, and suitability using the Suitability Assessment of Materials (SAM) instrument.  Resulting scores were analyzed both overall and by website.

Results:
Readability analysis revealed an overall average grade level of 13.4, with a range from 10.7 (MedlinePlus) to 15.8 (Wikipedia).  All sites exceeded the recommended 6th grade reading level.  Complexity evaluation revealed a mean PMOSE/iKIRSCH score of 6.2, consistent with "Low" complexity and equivalent to a high school level.  Websites ranged from "Very Low" complexity (BreastReconstruction.org, WebMD, National Cancer Institute, MedicineNet.com) to "High" complexity (Wikipedia).  Suitability assessment overall produced a mean 39.7% score, interpreted as "Not Suitable" for the intended patient audience.  Four sites (American Society of Plastic Surgeons, American Cancer Society, MedlinePlus, and National Cancer Institute) were found to have "Adequate" suitability scores when examined individually; the remaining six were "Not suitable".

Conclusion:
Available online patient material for breast reconstruction is too difficult for many patients to read.  Although overall table and list complexity of the websites is acceptable for average Americans, the content, literacy demand, and format is largely unsuitable for the intended patient audiences.  Attention to specific measures shown to improve readability and suitability is needed in designing appropriate material and minimizing disparities related to limited patient health literacy.

15.18 Patient Preferences in Access to Post-Mastectomy Breast Reconstruction

Posted on December 22, 2014

C. R. Vargas1, M. Paul1, O. Ganor1, M. Semack1, B. T. Lee1  1Beth Israel Deaconess Medical Center,Surgery/Plastic And Reconstructive Surgery,Boston, MA, USA

Introduction:
There is currently considerable variability at our institution in the timing of consultation with a plastic and reconstructive surgeon following initial diagnosis of breast cancer and meeting with a breast surgeon.  Providers in both specialties have expressed differing opinions regarding the ideal method of scheduling, and no published data exists regarding what patients prefer.  We aim to elucidate patients' preferences for the timing of plastic surgery consultation as part of the preoperative evaluation and planning process that follows a new diagnosis of breast cancer.

Methods:
A 16-question electronic survey instrument was developed based on formative patient comments and discussion between the breast and plastic surgery teams.  The survey was administered to all patients referred to the plastic and reconstructive surgery clinic during their initial consultation visit to discuss immediate post-mastectomy breast reconstruction between December 2013 and July 2014.  Surveys were administered in private consultation rooms by the clinic nurse and all data was collected anonymously.  Descriptive analysis was performed for each survey question.

Results:
A total of 31 unique responses were collected during the 7 month study period.  The largest number of patients (48%) indicated that they would prefer to see a plastic surgeon one week after their first consultation with a breast surgeon.  Only one patient reported a desire to see both surgeons on the same day.  Most patients indicated that having a family member or friend accompany them to the appointment (45%) and having time to process their cancer diagnosis before seeing the plastic surgeon (32%) were the key factors in deciding when they would like to discuss reconstruction.  All patients reported having had a discussion with their breast surgeon about reconstruction during the first appointment, and 55% said they had researched reconstructive options independently prior to consultation with the plastic surgeon.  All patients reported being "satisfied" or "very satisfied" with the process of meeting with both surgeons as well as with the information they received.

Conclusion:
The majority of patients in our study indicated a preference for delay between initial consultation with a breast surgeon and initial consultation with a plastic surgeon.  Overall, patients were satisfied with the process of meeting with each surgeon separately and with the information they received.  Incorporating patient preferences into the preoperative evaluation and planning process allows patients to optimize available support from loved ones and to begin coping with their diagnosis.
 

15.19 Spatial Interactions of Market and Socioeconomic Factors in Kidney Transplantation

Posted on December 22, 2014

J. T. Adler1,2, H. Yeh2,4, J. F. Markmann2,4, L. L. Nguyen1,3,4  1Brigham And Women’s Hospital,Center For Surgery And Public Health,Boston, MA, USA 2Massachusetts General Hospital,Transplant Surgery,Boston, MA, USA 3Brigham And Women’s Hospital,Vascular Surgery,Boston, MA, USA 4Harvard School Of Medicine,Brookline, MA, USA

Introduction: Prior work has demonstrated that the number of kidney transplants per population (KTP) is dependent on market factors such as competition and the number of transplant centers.  Socioeconomic status (SES) also plays a role in KTP.  We hypothesize that both of these important factors are spatially correlated in KTP, demonstrable by neighboring areas influencing each other more than remote areas.

Methods: The Herfindahl Hirschman Index (HHI), a standard measure of market competition, was calculated for each Health Service Area (HSA from Dartmouth Atlas) from 2000-2013 using zip code information of kidney transplant recipients from the Scientific Registry of Transplant Recipients.  Global Moran’s I, a measure of spatial dependency, was used to test market competition for spatial autocorrelation.  Areal interpolation was used to identify areas of concentrated market competition. Three standard spatial regression models were constructed to analyze the relationship between market competition and KTP adjusted for SES.

Results: Market competition exhibits moderate spatial autocorrelation (Global Moran’s I 0.27, P < 0.0001).  It is unevenly distributed in the United States and mirrors the general population (Figure).  The spatial lag model was the best fit by AIC criterion, suggesting a diffusion model among neighboring HSAs.  Under the spatial lag model, market competition was strongly associated with an increase in KTP by 27.5 ± 1.7 (P < 0.0001).  Markers of SES associated with an increase KTP included percent crowding (1.2 ± 0.2, P < 0.0001), percent with a college education or greater (0.39 ± 0.12, P = 0.0001), and percent unemployed (0.89 ± 0.21, P < 0.0001).  Lower median property value (per ten thousand dollars) was associated with slightly decreased KTP (0.03 ± 0.0007, P < 0.0001).

Conclusions: Competition and SES effects diffuse among neighboring HSAs in KTP.  This emphasizes a role understanding spatial autocorrelation in factors influencing KTP beyond market and SES factors.  Efforts to improve access to kidney transplantation should consider such issues in planning transplant center location, organ allocation, and organ sharing.

15.20 Does Transplant Center Flagging have Unintended Consequences? Spillover Effects of Medicare Policy

Posted on December 22, 2014

L. H. Nicholas1,2, D. Segev2  1Johns Hopkins University School Of Public Health,Health Policy & Management,Baltimore, MD, USA 2Johns Hopkins University School Of Medicine,Surgery,Baltimore, MD, USA

Introduction: The Center for Medicare and Medicaid Services (CMS) increasingly uses its role as one of the largest healthcare payers to influence quality of care.  Despite good intentions, policies designed to purchase high-quality care may have unintended consequences.  In 2007, CMS instituted aggressive Conditions of Participation (COP) for Transplant Centers, which flag transplant programs with worse than expected patient or graft survival for reviews that can be time-consuming and have major financial consequences; CMS can require low-performing centers to stop treating Medicare and Medicaid patients or enter into a Systems Improvement Agreement limiting the scope of transplant programs.  It is unknown whether flagging within a particular organ impacts the volume or quality of transplants of non-flagged organs.

Methods: We use data from the Scientific Registry of Transplant Recipients from 2004 – 2011 to study the intended and unintended consequences of CMS flagging for kidney and liver transplants, the most commonly transplanted organs, on other transplant programs within a center.  We used difference-in-differences regression models to compare outcomes at 130 transplant centers that were versus were not flagged (or informed of poor performance) before and after COP implementation.  Prior to the COP, hospitals received regular performance report cards but there were no sanctions against poor performers.  This approach allows us to separate transplant center’s response to the threat of CMS sanctions from any changes in behavior that are driven by information about quality of care they are currently providing. 

Results: 69 centers were flagged or notified of underperformance for kidney transplant and 43 for liver transplant at least once during the study period.  Simply receiving information about performance was not related to statistically significant changes in volume for either the low-performing organ or other programs within a transplant center.  However, programs flagged for poor performance in the COP period reduced volume for the flagged organ (54 fewer transplants per year after the second kidney flag, p < 0.01; 18 fewer transplants after the second liver flag, p < 0.10).  Transplant centers also reduced lung transplant volume following flags for both kidney (-12, p < 0.10 for first flag, -14, p < 0.01 for second flag) and liver transplant (-20, p < 0.05 for first flag, -5.3 for second flag).

Conclusion:  Transplant centers respond to the threat of CMS sanctions by reducing transplant volume for organ programs with and without potential quality programs.  Reduced access to transplant across organ types may be an unintended consequence of CMS efforts to improve the quality of transplant care. 

 

16.01 The Role of Breast MRI in Ductal Carcinoma in situ: Has it Improved Clinical Outcomes?

Posted on December 22, 2014

L. S. Sparber1, R. S. Chamberlain1,2,3  1Saint Barnabas Medical Center,Deptarment Of Surgery,Livingston, NJ, USA 2Saint George’s University,School Of Medicine,St. George’s, , Grenada 3New Jersey Medical School,Rutgers University – Department Of Surgery,Newark, NJ, USA

Introduction: For over three decades, screening mammography has played a central role in the early detection of in situ breast tumors in the United States.  More recently, breast magnetic resonance imaging (MRI) has emerged as a potentially more sensitive imaging modality than traditional mammography, but whether its use should be limited to adjunct screening for those at high risk or be universally utilized remains controversial.  While breast MRI undoubtedly detects subtler breast abnormalities, it is unclear whether this has resulted in an improvement in treatment decisions for patients with Ductal Carcinoma in situ (DCIS).

Methods: A comprehensive search for all published clinical studies on the use of MRI and its impact on DCIS management (2010-2014) was conducted using PubMed and Google Scholar.  The search focused on the value of MRI to guide treatment strategies, including mastectomy rates, re-excision rates and the overall benefit of this added imaging modality.  Keywords searched included: “breast MRI”, “mastectomy”, “DCIS”, “Ductal Carcinoma in situ”, and “surgical planning” in all possible combinations.

Results:  Six studies involving 3,296 patients have been published (Table 1). Pilewskie et al (2014) reported the largest study involving 2,321 DCIS patients (596 DCIS patients in MRI group; 1,725 patients in non-MRI group).  Within this group, 904 DCIS patients underwent radiation therapy [RT]; and 1,391 patients did not.  In the non-RT subgroup there was no association with the performance of an MRI and lower loco regional recurrence rates (p = 0.28). Three additional studies analyzed the impact of MRI on DCIS mastectomy rates, with Allen et al reporting no significant difference in mastectomy rates if an MRI was performed (p = .62). In contrast, Itakura et al reported increased mastectomy rates in patients undergoing preoperative MRI (p < .001). Re-excision rates were investigated in three studies, and preoperative performance of an MRI did not statistically impact these rates favorably or negatively. Pilewskie et al, demonstrated that breast conserving surgery was more successful in the non-MRI group (p = .06), whereas Allen et al and Kropcho et al found the results to not be statistically significant (p = .41 and p = .414, respectively). Across all studies, preoperative MRI was judged not routinely beneficial in DCIS patients.    

Conclusion: Breast MRI is associated with an increased sensitivity compared to other breast imaging technologies; however, it does not appear to improve clinical outcomes in patients with DCIS when added to conventional breast assessment.  Moreover, routine breast MRI in DCIS may contribute to an increase in unnecessary mastectomies.

 

16.02 Effects of drinking hydrogen-rich water on muscle fatigue caused by acute exercise in athletes

Posted on December 22, 2014

A. Nakao1, J. Kotani1, K. Kohama1, T. Nishimura1, T. Yamada1, S. Miyakawa2  1Hyogo College Of Medicine,Emergency, Disaster And Critical Care Medicine,Nishinomiya, HYOGO, Japan 2University Of Tsukuba,Doctoral Program In Sports Medicine,Tsukuba, IBARAKI, Japan

Introduction: Muscle contraction during short intervals of intense exercise causes oxidative stress, which can play a role in the development of overtraining symptoms, including increased fatigue, resulting in muscle microinjury or inflammation. Recently it has been said that hydrogen can function as antioxidant, so we investigated the effect of hydrogen-rich water (HW) on oxidative stress and muscle fatigue in response to acute exercise.

Methods: Ten male soccer players aged 20.9 ± 1.3 years old were subjected to exercise tests and blood sampling. Each subject was examined twice in a crossover double-blind manner; they were given either HW or placebo water (PW) for one week intervals. Subjects were requested to use a cycle ergometer at a 75 % maximal oxygen uptake (VO2) for 30 min, followed by measurement of peak torque and muscle activity throughout 100 repetitions of maximal isokinetic knee extension. Oxidative stress markers and creatine kinase in the peripheral blood were sequentially measured.

Results:Although acute exercise resulted in an increase in blood lactate levels in the subjects given PW, oral intake of HW prevented an elevation of blood lactate during heavy exercise. Peak torque of PW significantly decreased during maximal isokinetic knee extension, suggesting muscle fatigue, but peak torque of HW didn’t decrease at early phase. There was no significant change in blood oxidative injury markers (d-ROMs and BAP) or creatine kinease after exercise.

Conclusion:Adequate hydration with hydrogen-rich water pre-exercise reduced blood lactate levels and improved exercise-induced decline of muscle function. Although further studies to elucidate the exact mechanisms and the benefits are needed to be confirmed in larger series of studies, these preliminary results may suggest that HW may be suitable hydration for athletes.

 

16.03 Factors Associated with Readmission and Length of Stay Following Gastric Bypass

Posted on December 22, 2014

S. X. Sun2, C. Hollenbeak3, A. Rogers1,2  1Penn State Hershey Medical Center,Minimally Invasive Surgery,Hershey, PA, USA 2Penn State Hershey Medical Center,General Surgery,Hershey, PA, USA 3Penn State Hershey Medical Center,Outcomes Research And Quality,Hershey, PA, USA

Introduction:
Interest is growing in preventing readmissions as payers start to link reimbursement to readmission rates.  The purpose of this study was to assess factors that contribute to 30-day readmission rates for patients undergoing gastric bypass for obesity and determine whether these readmissions may be preventable. 

Methods:
Data from the Pennsylvania Health Care Cost Containment Council (PHC4) were queried for all patients undergoing elective gastric bypass for obesity in 2011 (n=4,505). The outcomes measured were length of stay (LOS) and 30-day readmission. Univariate comparisons between characteristics of readmitted (n=298) and non-readmitted patients were performed using t tests and chi-square tests. Readmission was modeled using logistic regression; LOS was modeled using linear regression and controlled for potential confounders.  

Results:
Of the 298 (6.6%) patients who were readmitted, the most common cause for readmission was bleeding (11.84%) followed by infection (8.88%), and abdominal pain (7.89%). On multivariate analysis, African American race, open gastric bypass, and history of myocardial infarction or rheumatoid arthritis were associated with increased odds of readmission within 30 days of the index hospitalization. Longer LOS was also predictive of readmission (OR 1.10, p=<0.0001). Determinants of LOS were assessed using linear regression. Patients who were above age 51, and those with history of congestive heart failure, peripheral vascular disease, and kidney disease were more likely to have longer lengths of stay. Black race, open surgery, and discharge to an extended care facility were also predictive of prolonged hospital stays. 

Conclusion:
This study showed that the most common causes of 30-day readmission following elective gastric bypass was bleeding, infection and abdominal pain. Black race, open surgery and comorbid conditions, such as heart disease, were associated with higher odds of readmission and longer lengths of stay. Even though it is difficult to alter patient comorbidities, our results show that it may be beneficial to optimize these comorbid conditions before gastric bypass surgery as this may lead to lower readmission rates and shorter lengths of stay. 
 

16.04 Analysis of Internet Information on Lateral Lumbar Interbody Fusion

Posted on December 22, 2014

R. Belayneh1, A. Mesfin2  1Howard University College Of Medicine,Washington, DC, USA 2University Of Rochester School Of Medicine And Dentristry,Orthopaedic Surgery,Rochester, NY, USA

Introduction:
The Internet is a common resource for health and medical information. Previous studies have shown the Internet’s shortcomings in presenting comprehensive information regarding surgical procedures. Lateral lumbar interbody fusion (LLIF) is a surgical technique that is being increasingly used. To our knowledge, there are no studies evaluating the quality of information available on the Internet regarding LLIF. The purpose of this study is to examine information on the Internet about LLIF and determine the completeness and accuracy of the information provided.

Methods:
The top 35 websites providing information on the “lateral lumbar interbody fusion” from four search engines (Google, Yahoo, Bing, DuckDuckGo) were identified. 140 websites were evaluated. Each website was categorized based on authorship (academic, private, medical industry, insurance company, other) and we analyzed: appropriate patient inclusion and exclusion criteria, surgical treatment alternatives, non-surgical treatment alternatives, claimed benefits, complications and risks, industry-sponsored literature, peer-reviewed literature, description and diagram of procedure, direct contact information of the author, and date of last update. 

Results:
78 unique websites were identified after excluding duplicate and inaccessible websites. 46.2% of websites were authored by a private medical group, 26.9% by an academic medical group, 5.1% by biomedical industry, 2.6% by an insurance company, and 19.2% by other sources. 68% of websites reported patient inclusion criteria and only 24.4% reported exclusion criteria. Benefits of LLIF were reported in 69.2% of websites and 36% of websites reported potential complications of LLIF. Alternative surgical options were discussed in 50% of websites and non-surgical options were discussed in 7.7% of websites. 21.8% of websites contained references to peer-reviewed literature while 32.1% contained industry-sponsored literature.

Conclusion:
Overall, the quality and completeness of information regarding LLIF on the Internet is poor.  The majority (46.2%) of Internet information on LLIF is provided by private medical groups. Only 36% of websites discuss potential complications of LLIF and most of the cited literature (32%) is from the biomedical industry.  Spine surgeons and spine societies can assist in improving the quality of the information on the Internet regarding LLIF.
 

16.05 Risk Factors for Pediatric Surgical Readmissions: An Analysis of the Pediatric NSQIP Database

Posted on December 22, 2014

G. M. Taylor1, M. C. Shroyer1, A. B. Douglas1, R. T. Russell1  1University Of Alabama At Birmingham, Children’s Of Alabama,Pediatric Surgery,Birmingham, AL, USA

Introduction: Hospital readmissions account for a large proportion of health care expenditures. The patient characteristics associated with surgical readmissions in children help define preoperative patient risk factors, which may be modifiable but also may be used for preoperative family counseling.  The objective of this study was to characterize readmission rates and factors associated with readmission among children following surgery.

Methods: The Pediatric National Surgical Quality Improvement Project (NSQIP) is a multicenter clinical registry from 50 participating sites collecting data to measure the quality of children's surgical care. The 2012 Pediatric NSQIP public use file (PUF) was queried for pediatric general surgical patients who had an unplanned readmission in 2012.  Detailed patient and case characteristics were analyzed.  Univariate and multivariate logistic regression were utilized to identify patient characteristics, clinical variables, and comorbidities predictive of unplanned readmissions.

Results: 18,643 cases were analyzed from the 2012 Pediatric NSQIP PUF file. Of these, 1111 patients (6%) experienced unplanned readmissions within 30 days of surgery.  Significant preoperative variables/comorbidities associated with readmission are included in Table 1. In addition, children in the older age groups (30 days-6months, 6 months-2 years, 2-5 years, 5-12 years, and > 12 years) were more likely to be readmitted than those < 30 days old.  Variables in the model that did not predict readmission included race, a preoperative diagnosis of diabetes, cerebral palsy, chronic lung disease, cystic fibrosis, major/severe cardiac risk factors, enteral or parenteral nutritional support at the time of operation, history of prematurity, and emergent/urgent operation. Infectious complications were the reason for 40% of unplanned readmissions.

Conclusions: Certain patient risk factors and comorbidities were associated with an increased risk of unplanned readmission. Though we may not be able to directly affect these risk factors, we can utilize them to counsel high risk patients and their families preoperatively about the likelihood of readmission due to these risk factors.  Infectious complications were the most common reason for readmission.

16.06 Variability in Surgical Skin Preparation Adherence in Common Pediatric Operations

Posted on December 22, 2014

J. M. Podolnick2,3,4, L. R. Putnam2,3,4, S. Sakhuja2,3,4, C. M. Chang2,3,4, M. T. Austin2,3,4, K. P. Lally2,3,4, K. Tsao2,3,4  4Children’s Memorial Hermann Hospital,Houston, TX, USA 2University Of Texas Health Science Center At Houston,Department Of Pediatric Surgery,Houston, TX, USA 3Center For Surgical Trials And Evidence-based Practice,Houston, TX, USA

Introduction:

Skin antisepsis agents are commonly used in an effort to decrease surgical site infections (SSI). However, surgeon preference, anatomical site considerations, and patient age may influence proper agent utilization. Despite institutional adoption of evidence-based guidelines, we hypothesized that adherence to skin preparation guidelines is variable in pediatric operations.

Methods:

A retrospective cohort study of eight common pediatric operations (laparoscopic appendectomy, fundoplication, gastrostomy tube placement, pyloromyotomy, laparoscopic cholecystectomy, abscess incision and drainage, inguinal hernia repair, and stoma takedown) was performed to evaluate the skin prep agents utilized over a one year period.  The skin prep used for each operation was recorded as well as patient age, gender, operative time, prep nurse, surgeon, and anatomical site of prep (torso, extremity, pelvis/perineum). Correct prep agent was determined based on adherence to our institutional guidelines based on best-evidence and best-practice in pediatric hospitals. Logistic regression and the chi squared test were performed; p<0.05 was considered significant.

Results:

183 cases were reviewed with an overall adherence of 58% to skin prep guidelines. Adherence was highest for laparoscopic appendectomies and laparoscopic cholecystectomies (92% and 96%, respectively) and lowest for inguinal hernia repairs and stoma takedowns (32% and 8%, respectively). A total of five different skin prep agents or combinations were used with at least two different agents/combinations used per case type; all five were used for appendectomies during the study period (Table). Factors associated with non-adherence included type of operation, surgeon, and patient age.

Conclusion:

Significant variability in adherence to correct skin prep guidelines exists for common pediatric operations. Contributing factors include type of operation, surgeon, and patient age. Consistent practice and adherence to evidence-based guidelines for skin preparation requires targeted interventions in order to optimize skin antisepsis and minimize risk of SSI.

16.07 ~~Hospital Departmental Variation in Children’s Surgical Outcomes

Posted on December 22, 2014

A. M. Stey1, B. L. Hall2,6, M. Cohen2, C. Y. Ko2,5, S. Rangel4, K. Kraemer2, R. Moss3  1Mount Sinai,New York, NY, USA 2American College Of Surgeons,Chicago, IL, USA 3Nationwide Childrens Hospital,Columbus, OHIO, USA 4Boston Children’s Hospital,Boston, MA, USA 5University Of California Los Angeles,Los Angeles, CA, USA 6Washington University In Saint Louis,Saint Louis, MO, USA

~~Introduction: Institution wide efforts in quality improvement have improved patient safety across disciplines. Achieving improvements in surgical outcomes may be dependent on department specific factors. The aim of this study was to determine if outcomes in children’s surgical subspecialties within the same institution were similar or different to outcomes in general pediatric surgery.

Methods: 2011-2012 ACS-NSQIP-P data were sorted into six specialties; general surgery, plastics, urology, otolaryngology, orthopedics and neurosurgery among 50 hospitals. 30-day composite morbidity (occurrence of 17 postoperative complications) was the primary outcome. Multivariate hierarchical models were used to estimate risk-adjusted hospital odds ratio of morbidity for each specialty. Spearman correlation, ranking of these odds ratios was performed. General surgery was treated as the reference since it had the largest case volume.

Results: Correlations in an institutions general surgery outcomes, and specialty outcomes were moderate (R=0.3-0.5, p<0.01) with the exception of orthopedics (R=0.1, p=0.5). Median difference in hospital performance rank between general surgery and specialties ranged from 8-14 ranks. Median difference in hospital decile performance ranking ranged from 1-3 deciles. 1-2 hospitals ranked in the best decile in general surgery were also in the best decile in specialty care for a 20-40% concordance in best decile designation.  0-2 hospitals ranked in the worst decile in general surgery were also in the worst decile in specialty care for a 0-40% concordance in worst decile designation. There was some overlap in risk adjusted specialty performance within hospitals but outlier departments within hospitals were observed (Figure).

Conclusion:  There is variability in surgical outcomes between specialties within the same institution. This suggests that unique department specific factors may drive surgical outcomes.  Every hospital likely has an area which could be the focus of quality improvement.

 

Figure Legend:

Odds ratio is given in order for general surgery, urology, otolaryngology, plastics, neurosurgery and orthopedics for each hospital ranked by dummy identifier.
 

16.08 Risk Factors and Preventability in Reducing Early Hospital Readmission after Liver Transplantation

Posted on December 22, 2014

C. E. Rogers1, P. Baliga1, K. Chavin1, D. Taber1  1Medical University Of South Carolina,Transplant Surgery,Charleston, Sc, USA

Introduction:  There is an increasing pressure for hospitals to reduce early hospital readmission (EHR) rates for high-cost, high-risk surgical procedures.  Studies have also shown EHRs to be a measure of inpatient quality of care. 

Methods:  The aims of this study were to determine the predominant risk factors associated with EHR, to develop a risk model and to determine the etiologies, timing and preventability of readmissions in liver transplant (LTX) patients. All patients who received a LTX between Jan 2011 – May 2014 were included. Patients who experienced graft loss within one month after LTX were excluded. 

Results: A total of 207 LTX recipients were included, 48% (n=67) were readmitted within 30 days (EHR). Risk factors for EHR included African American race (13% vs. 38%, p=0.006), primary diagnosis of biliary atresia (1% vs. 8%, p=0.025) and donor history of stroke (33% vs. 54%, p=0.007). Although not statistically significant, diagnosis of hepatitis C (34% vs. 45%, p=0.119), an increase in pre-transplant hemoglobin (10.8 vs. 11.3 gm/dL, p=0.119) and a decrease in serum albumin (3.1 ± 3.1 vs. 2.6 ± 0.8 gm/dL, p=0.188) also correlated with an increased risk for EHR.  A history of a previous liver transplant (10% vs. 0%, p=0.007) and dialysis within a week prior to transplant (8% vs. 2%, p=0.067) appear to be protective against EHR. These 8 factors were then used as variables in a logistic regression analysis to develop a risk model that demonstrated a negative predictive value of 71.4%, a positive predictive value 67.8%, and an overall predictive value of 70.3%. The secondary analysis revealed that of the patients readmitted within 7 days (n=39), 25% were due to known or ongoing medical problems, which were identified, on average, 2 days (range1-4) prior to the EHR and potentially preventable.  Graft loss was significantly higher in LTX with EHR (Figure 1). 

Conclusion: This analysis identified specific factors for EHR that can potentially predict which patients are at high-risk for readmission.  Future analyses should attempt to prospectively validate this model and target the high-risk patients through interventions designed to minimize EHR and improve overall quality of patient care.

 

16.09 The “Halo Effect” in Trauma Centers: Does it Extend to Emergency General Surgery?

Posted on December 22, 2014

N. Nagarajan1, S. Selvarajah1, H. Alshaikh1, F. Gani1, H. Alturki1, A. Najafian1, C. K. Zogg1, D. T. Efron1, E. B. Schneider1, A. H. Haider1  1Johns Hopkins University School Of Medicine,Center For Surgical Trials And Outcomes Research, Department Of Surgery,Baltimore, MD, USA

Introduction:  Trauma Centers (TC) have been shown to have a “halo effect”, resulting in improved outcomes for non-trauma conditions. It remains unclear if these improvements extend to outcomes for emergency general surgery (EGS). Using emergent colectomy in patients with diverticulitis as an index condition, the objective of this study was to compare outcomes between TC and NTC.

Methods:  The Nationwide Emergency Department Sample (2006-2011) was queried for patients (≥16 years) who underwent an emergent colectomy (ICD9: 173*, 457*, 458*) with a primary diagnosis of diverticulitis (ICD9: 562.11, 562.13). Outcomes studied included mortality, total charges (in 2011 dollars) and length of stay (LOS). Mortality in TC and NTC was compared using logistic regression, controlling for age, sex, Charlson Comorbidity Index (CCI), type of insurance, income quartile, partial/total colectomy, presence of peritonitis, perforation, and hospital region, clustering by hospital. Unadjusted total charges and LOS were analyzed with non-parametric tests, then were adjusted for all of the above and mortality. Adjusted total charges and LOS were analyzed using generalized linear models with gamma and Poisson distributions, respectively.  

Results: A total of 25,396 patients were included; of whom 5,189 (20.4%) were treated at TC and 20,207 (79.6%) at NTC. Median age [60 years (IQR: 49-73), p = 0.959] and proportion of females (51.6% vs. 51.3%, p = 0.395) were similar between TC and NTC, but there were significant differences in insurance status (p = 0.027) and median household income (p <0.001) (Table I). Unadjusted mortality at TC did not significantly differ from NTC, median charges and were significantly different (Table I). After controlling for patient, procedure and hospital-level characteristics, the odds of mortality was significantly higher in TC (OR=1.24, 95% CI, 1.02-1.51). Estimated mean charges ($127,801 vs. $116,464, p = 0.004) and LOS (IRR=1.06, 95% CI, 1.05-1.11) were also significantly higher in TC after adjustment.

Conclusion: The improved outcomes reported for other non-trauma conditions in TC were not observed for patients undergoing an emergent colectomy for diverticulitis after accounting for demographic and hospital-level characteristics.  Future research is needed to discern if differences in the clinical course of patients in TC compared to those in NTC are affecting our findings.

« Previous Page
Next Page »