33.03 Challenges in the Management of Adrenal Incidentalomas: One Year at an Urban Level I Trauma Center

Posted on January 31, 2019

T. Feeney1,2, S. Talutis1,2, M. Janeway1,2, P. Sridhar1,2, A. Gupta3, P. Knapp4, J. Moses5, D. McAneny1,2, F. T. Drake1,2  1Boston Medical Center,Surgical Endocrinology,Boston, MA, USA 2Boston University School of Medicine,Surgery,Boston, MA, USA 3Boston University School of Medicine,Radiology,Boston, MA, USA 4Boston University School of Medicine,Medicine,Boston, MA, USA 5Boston University School of Medicine,Pediatrics,Boston, MA, USA

Introduction: Incidental findings, including adrenal masses, are increasingly common, and their management is challenged by poor communication, determination of responsibility, the infrequency of clinical significant, and competing health considerations. Clear guidelines exist for the evaluation of incidental adrenal masses, but barriers to implementation are not well understood. The goal of the current study was to evaluate usual care practice patterns for follow-up of incidental adrenal masses at an urban Level 1 trauma and tertiary referral center.  

Methods:  A retrospective review of medical records for all patients with an incidentally-discovered adrenal mass at a tertiary medical center in Boston, MA and associated ambulatory care locations. All individuals ≥18 years old with a mass identified during 2016 were included. Patterns of evaluation, follow-up, and associated adrenal diagnoses were determined.

Results: 244 patients were identified from approximately 19,171 eligible scans. Median age = 63 years; median size = 1.5cm. Most patients (85%) had a single lesion, and 51% of imaging was performed during an ED evaluation. Of the 244 patients, 31 (12.7%) appropriately did not undergo adrenal evaluation due to severity of comorbid conditions. Among the remaining 213, 38 were lost to follow-up and only 20 had a PCP evaluation. Of those with an identifiable PCP, 118 (80%) had a PCP within our health system. Dedicated adrenal evaluation (imaging and/or labs) was performed by a PCP in only 9% of patients (16/175). Patients were more likely to undergo follow-up if the mass was characterized in the radiology report as “indeterminate” versus “benign” (65% vs 30% p<0.001). Of 130 patients with hypertension, only 2 (3.2%) underwent aldosterone/renin screening. Of 53 patients with co-morbid conditions potentially related to Cushing’s Syndrome, 0 patients had appropriate hormonal evaluation. Four patients (1.9%) had hormonal evaluation for pheochromocytoma. Among all 244 patients, our review yielded 71 patients with some diagnosis listed in the chart (including, simply, “adrenal nodule.”) including 1 cortisol secreting mass and 8 metastases.  

Conclusion: Under usual care, most patients with an incidental adrenal mass do not undergo workup described in published guidelines. A potential contributing factor is detection during ED evaluations, which challenges follow-up. Nine of 213 adrenal masses (4.2%) were clinically significant; however, according to available literature, we would have expected up to 25%, or 53 masses, to be. This discrepancy suggests that published estimates are not necessarily generalizable to all healthcare settings or that usual care is inadequate to detect clinically significant lesions. This discrepancy between guidelines and usual care offers a clear quality improvement opportunity, and we have implemented a prospective initiative to facilitate communication and evaluation of incidental adrenal masses.

33.02 Post-thyroidectomy Neck Appearance and Impact on Quality of Life in Thyroid Cancer Survivors

Posted on January 31, 2019

S. Kurumety1, I. Helenowski1, S. Goswami1, B. Peipert1, S. Yount2, C. Sturgeon1  2Feinberg School Of Medicine – Northwestern University,Department Of Medical Social Sciences,Chicago, IL, USA 1Feinberg School Of Medicine – Northwestern University,Department Of Surgery,Chicago, IL, USA

Introduction:  There is a paucity of patient-reported data on thyroidectomy scar perception. The magnitude and duration of the impact of thyroidectomy scar on quality of life (QOL) is not known. We hypothesized that age, sex and race would predict scar perception, and that worse scar perception would correlate with lower?QOL. Furthermore, we hypothesized that over time, scar perception would improve.

Methods: Adults >18 years who had undergone thyroidectomy for cancer?(n=1743)?were recruited from a support group and surveyed online. Demographics, clinical characteristics, and treatment history were assessed. Scar perception was scored on a 5-point Likert scale.?QOL was evaluated via PROMIS-29. Respondents were grouped and compared based on their responses. The relationship between scar perception, patient characteristics, and QOL were?analyzed?with univariable and multivariable?models. Kruskal-Wallis, Fisher’s exact test, and cumulative logistic regression were used to compare?categorical variables. The relationship?between PROMIS domains and scar perception?were?analyzed using Spearman partial correlation coefficients?(r)?adjusted for age and years after surgery.?Holms-Bonferroni was used to correct for multiple comparisons.

Results: Increasing age?was associated with?better?scar perception (OR 0.975/year; 95% CI 0.967-0.983; p<0.001).?71% of respondents age >45?years?reported no concern over scar, compared to only 53% of respondents?<45;?p < 0.0001. Increased time since surgery?was?also associated with?improved?scar perception (OR 0.962/year; 95% CI 0.947-0.977; p<0.0001), but there was no statistically significant difference between current and baseline neck appearance >2 years after surgery. On multivariable analysis, age >45 years (OR 0.65; [0.52-0.81] p=0.0001), >2 years since surgery (OR 0.57; 95% CI 0.46-0.70; p<0.0001), and higher self-reported QOL (OR 0.77; 95% CI 0.67-0.89; p=0.0003) were independent predictors of better self-reported scar appearance. In patients <2 years after surgery (n=568), the PROMIS domains of anxiety (rs=0.19; p<0.0001), depression (rs=0.21; p<0.0001), social function (rs=-0.18; p<0.0001), and fatigue (rs=0.21; p<0.0001) had weak but statistically significant associations with worse scar appearance. Sex and race/ethnicity were not associated with scar perception.

Conclusions: This is the largest study conducted in the U.S to evaluate scar perception after thyroidectomy, and the first to use PROMIS measures.??Age >45, >2 years since surgery, and higher self-reported QOL were independent predictors of better scar perception.??There was no significant difference between perception of current and baseline neck appearance in the group of respondents >2 years after thyroidectomy. There was a weak correlation between scar perception?and?PROMIS domains in patients who had surgery within 2 years. The impact of thyroidectomy scar on QOL appears to be mild and transient and plateaus after 2 years.

33.01 Intraoperative recurrent laryngeal nerve oedema affects post-operative voice quality

Posted on January 31, 2019

J. C. Lee1,2 1The Alfred, Melbourne, Victoria, Australia 2Monash Health, Dandenong, Victoria, Australia

Introduction: Over the last 150 years thyroid surgery has been transformed from a procedure with high mortality to a very safe one. However, recurrent laryngeal nerve (RLN) injuries continue to occur. This study examined the differential palsy rates between the left and right RLNs, and the role of intraoperative nerve swelling as a risk factor of postoperative palsy.

Methods: Thyroidectomy data of patients of the Monash University Endocrine Surgery Unit were collected from 13 institutions, including demographics, pathology, and change in RLN diameter (subgroup). Voice quality was scored subjectively using the Voice Disorder Index and objectively using the Dysphonia Severity Index (DSI), before and after operation, in a subgroup.

Results: A total of 5,334 RLNs were at risk in 3,408 thyroidectomies in this study. The overall RLN palsy rate was 1.5%, greater on the right side than the left for bilateral cases (P = .025), and greater on the left side than the right for unilateral cases (P = .007). The diameter of the right RLN was larger than the left RLN, both at the beginning and end of the dissection (P = .001). The RLN diameter increased by approximately 1.5-fold (P < .001). In hemithyroidectomy patients, the greater the increase in recurrent laryngeal nerve diameter, the worse the post-operative DSI score (P = .03). Patients who underwent either hemi- or total thyroidectomy both reported significant deterioration of voice. However, on objective assessment, only total thyroidectomy patients showed significant deterioration (Mean DSI 4.0 ± 0.3 – 2.5 ± 0.3, P < .01). Conclusion: As we embark on new and innovative thyroidectomy techniques, it is important to review the anatomical and functional details of the RLN. It is also paramount that we do not compromise patient outcomes while developing minimally invasive techniques.

32.10 Cost Awareness of Common Surgical Supplies is Severely Limited Regardless of Role and Experience

Posted on January 31, 2019

R. Sorber1, D. Stobierski2, G. Dougherty2, C. Kang3, Y. Lum1,2  1The Johns Hopkins University School Of Medicine,Department Of Surgery,Baltimore, MD, USA 2Johns Hopkins University School Of Medicine,Bloomberg School Of Public Health,Baltimore, MD, USA 3The Johns Hopkins University School Of Medicine,School Of Nursing,Baltimore, MD, USA

Introduction:

Increasing concerns regarding healthcare costs have triggered interest in reducing operating room waste, but the overall level of awareness of costs among the team members making intraoperative decisions regarding supply selection remains unclear. This work characterizes the knowledge of supply costs among surgeons as well as operating room staff in a large academic hospital and seeks to examine the contribution of operating room role and years of experience with regards to cost awareness. 

Methods:

This work is a cross-sectional study of surgeons, surgical trainees, operating room nurses and surgical technicians (n=372) across all surgical specialties at the Johns Hopkins Hospital. Participants completed a survey reporting frequency of use and estimated cost for the eleven commonly used surgical supplies. They were also asked to render opinions on the role of cost in surgical decisionmaking and their desire for increased access to cost information. Responses were stratified by respondent role as part of the surgical team and years of experience. All data was analyzed using Stata to perform ANOVA and Χ2 testing as well as linear regression modeling.

Results:

Cost estimates ranged widely, with most respondents overestimating supply costs of most inexpensive items by 1.3-3.2 times the actual cost and underestimating the two most expensive items (0.87-0.89 times actual cost). There was no significant difference in accuracy of cost estimation when stratified by role, years of experience, item, or frequency of item use. The vast majority of respondents (88.7%) expressed agreement that cost should factor into surgical decisionmaking as well as a desire to learn more regarding cost of common supplies.

Conclusion:

Accurate knowledge of the cost of common surgical supplies is severely limited among surgeons, surgical trainees and operating room staff with no correlation to years of experience or frequency of use. While concerning, this lack of knowledge coexists with a strong desire to augment cost awareness. Improved access to cost information across disciplines has high potential to inform surgical decisionmaking and potentially decrease operating room waste.

 

32.09 Debunking the July Effect: Systematic Review and Novel Difference-in-Difference Analysis

Posted on January 31, 2019

C. K. Zogg1,2,3, D. Metcalfe3, S. A. Hirji2, K. A. Davis1, A. H. Haider2  1Yale University School Of Medicine,New Haven, CT, USA 2Brigham And Women’s Hospital,Center For Surgery And Public Health,Boston, MA, USA 3University of Oxford,Oxford, United Kingdom

Introduction: The arrival of new residents at the beginning of the academic year has long been associated with perceived adverse patient outcomes. Numerous studies in recent years have sought to prove/disprove the ‘July Effect.’ The objective of this study was to provide a definitive answer, combining data on mortality, morbidity, and unplanned readmission through a systematic review/meta-analysis and expanded difference-in-difference (DID) analysis of seasonal variation in outcomes for teaching vs non-teaching hospitals across seven common medical and surgical conditions.

Methods: 1) Systematic review and meta-analysis of studies published prior to July 31, 2018. 2) DID analysis of adult patients, ≥18y, with primary diagnosis/procedure codes for AMI, CVA, pneumonia, elective CABG, elective colectomy, craniotomy, or hip fracture contained within the 2012-2015 Nationwide Readmissions Database. Weighted models compared disease-specific differences in 30- and 90-day mortality, readmission, and median index hospital length of stay (LOS) between patients admitted to teaching vs non-teaching hospitals in July-August vs September-June and April-May.

Results: A total of 85 studies met inclusion criteria. Of these, 12 (14.1%) reported evidence in support of a July Effect for any outcome (1/13 high-quality studies). An additional 14 (16.5%) suggested that evidence was mixed (3/13 high-quality studies). 57/85 assessed mortality, of which 25 were eligible to be included in the random effects meta-analysis (Figure), OR(95%CI): 1.00(0.97-1.03). 48/85 assessed major morbidity, of which 26 were included in the random effects meta-analysis, 1.02(0.99-1.05). One met inclusion for readmission, 0.90(0.80-1.23). Data assessment similarly revealed no significant differences in 30- or 90-day mortality when comparing teaching vs non-teaching hospitals in July-August vs April-May (e.g. absolute 30-day DID[95%CI] hip fracture: +0.1[-0.7 to +0.9] percentage-points). When compared relative to September-June, AMI showed a slight 30-day difference, +0.4(0.1-0.8) percentage-points, that was not significant for p<0.001. Similar results were observed for 30- and 90-day readmission (e.g. 30-day hip fracture: -0.2[-2.0 to +1.6] percentage-points) and median index hospital LOS (0.0[0.0-0.0] days).

Conclusion: An influx of recent studies has challenged pre-existing notions of the July Effect for major adverse outcomes: mortality and morbidity. While evidence refuting the July Effect in readmission is scarcer, DID assessment of common medical and surgical conditions demonstrated that the July Effect does not exist. Taken together, the results suggest that fears surrounding the July Effect are unfounded and that further studies might be unwarranted.

32.08 Quantifying Documentation Burden Using the Electronic Medical Record.

Posted on January 31, 2019

G. J. Eckenrode1, H. Yeo1  1Weill Cornell Medical College,Surgery,New York, NY, USA

Introduction:
Time spent on clinical documentation is frequently cited as a contributing factor in physician burnout. Many physicians believe that the amount of patient documentation, both in terms of number of documents required as well as total length of documents, is increasing and has created a significant burden. There is a paucity evidence regarding the actual volume of increase and the amount of documentation taking place.

Methods:
We used a database of text-based clinical documents extracted from the inpatient electronic medical record (EMR) of a single institution. This database was established in 2014 and aggregates all electronic medical information for all patients treated by a large gastrointestinal surgery center within the institution. It contains notes dating from the widespread adoption of electronic medical documentation in approximately 2006 to the present day. We extracted all physician-visible text-based notes for each patient’s entire hospital stay for analysis. The number of notes and the word count for each note was calculated for each patient for each day.

Results:
The database contains 141,480 unique patient identification numbers and 10,925,542 physician-visible inpatient notes. Notes prior to 2007 were excluded due to low daily volume reflective of low document capture by the database. We found 1,591 note types as labeled by the EMR. The range of patients receiving notes on any given day ranged from 250 – 350 patients in 2007 with a steady increase to 400 – 1100 patients at the end of 2017. The average number of notes per patient was 77 with a range of 1 – 10,000. The average number of notes per patient per day remained constant over time, with a range of 4-6 but the number of words per patient per day rose constantly with time from 4,000 – 6,000 in 2008 to 12,000 – 14,000 per day at the end of 2017, a 3-fold increase.  

Conclusion:
While the number of notes per patient per day has been constant over time, the number of words per note have increased markedly. Over the past decade, at this single institution patient documentation has increased in both quantity and complexity, requiring more work from physicians to create and manage, increasing the burden of clinical care.
 

32.07 How Much is Enough? A Crowd-Sourced Public Opinion Survey on Minimum Surgeon Volumes

Posted on January 31, 2019

J. Danford1, D. Underbakke2, B. Sirovich1, S. Wong1,2, M. Sorensen1,2  1Geisel School of Medicine at Dartmouth,Hanover, NH, USA 2Dartmouth-Hitchcock Medical Center,General Surgery,Lebanon, NH, USA

Introduction:  In the past 15 years, numerous investigators have demonstrated an association between surgeon volumes and outcomes.  From this body of literature, there have been attempts to quantify minimum volume standards for specific surgeries. Despite the growing body of literature on the subject, little has been reported on the impact of this research on public perception of surgeon competency.

Methods:  A survey on public perception of the importance of surgeon volumes was designed using a modified Delphi technique and completed by participants using Amazon Mechanical Turk, an online crowdsourcing marketplace. Respondents completed a 38-question survey on their opinion of minimum volume standards and other factors that may influence their choice of surgeon. They were also asked to estimate minimum volume standards for four different surgeries, and to consider the implications of published minimum volume numbers in two diagnostic scenarios. 

Results: The survey was completed by 2,056 people. The respondents were 51% male, 49% female. Median age range was 30-39 years old.

Overall, 81% (n=1,666) of people agreed that surgeons should be subject to minimum volume standards. Only 19% (n=384) reported having prior knowledge about a link between surgeon volumes and outcomes. 

Respondents accurately estimated the published suggested minimum volume standard for inguinal hernias of 25. For knee replacement surgery, respondents estimated a minimum of 30 per year: 1.5 times the published minimum volume standard of 20. For mitral valve repair, respondents estimated 44 per year: 4.4 times the published minimum volume standard of 10. For pancreaticoduodenectomy, respondents estimated 44 per year: 8.8 times the published minimum volume standard of 5.

When posed with the scenario of needing an inguinal hernia repair, 77% (n=1,584) said they would require their surgeon to have met a minimum volume standard to proceed with surgery. If told their surgeon performed 25 per year (the published suggested minimum volume standard), 55% (n=1,127) of respondents would feel comfortable proceeding with surgery. However, when posed with needing a pancreaticoduodenectomy, 92% (n=1,877) said they would require their surgeon to have met a minimum volume standard. And when told their surgeon did 5 per year (the established minimum volume standard), only 13% (n=265) would feel comfortable proceeding with surgery.  

Conclusion: This survey suggests that surgical volumes are important to the lay public. However, it also demonstrates the general public’s unrealistic expectations of minimum volume standards and inability to interpret surgical volume numbers when attempting to use them to judge a surgeon’s competence. This study has implications for patients, surgeons, hospitals, and policy makers when considering the implementation of minimum volume standards and how best to educate the public about this aspect of choosing a surgeon.
 

32.06 Enhanced Recovery After Surgery for Hysterectomy Shortens Hospital Stay and Reduces Care Disparity

Posted on January 31, 2019

S. M. Stapleton1,2, R. M. Sisodia1,2, D. C. Chang1,2, N. P. Perez1,2, B. V. Udelsman1,2, M. G. Del Carmen1,2, K. D. Lillemoe1,2  1Massachusetts General Hospital,Surgery,Boston, MA, USA 2Harvard School Of Medicine,Brookline, MA, USA

Introduction:  Multi-modal approaches for enhanced recovery after surgery (ERAS) represent evidence-based protocols designed which standardize peri-operative care, to improve patient outcomes and reduce cost to the health system. ERAS protocols were first developed for use in the field of colorectal surgery. More recent efforts seek to expand implementation throughout surgical and nonsurgical specialties. We aim to evaluate implementation of an ERAS protocol within the department of obstetrics and gynecology. We hypothesize that care standardization will reduce length of stay in a field that does not routinely perform bowel surgery.

Methods:  An observational study at a tertiary academic medical center was performed for the 12-months pre- and 6-months post-ERAS implementation. Female patients with ICD10 codes for elective hysterectomy were included. In-hospital deaths were excluded. Endpoints assessed included length of stay (LOS), and likelihood of same-day discharge or readmission. Multivariable analysis adjusted for ERAS, minimally invasive vs. open hysterectomy, procedure performed for malignancy, and case start time. Difference-in-difference analyses were performed by race.

Results: We analyzed 1004 hysterectomies, 88.9% (n=812) lap and 98.3% (n=987) for benign disease. Hospital duration was significantly reduced post-ERAS (24.2 [IQR11.4-28.5] hrs. vs. 13.1 [IQR10.6-27.7] hrs., p=<0.01). Additionally, hospital duration was significantly reduced post-ERAS for patients staying <24 hrs. (13.0 hrs. pre vs. 11.9 hrs. post, p=0.01), but interestingly the opposite trend was observed for patients staying >24 hrs. (47.8 hrs. pre vs. 65.2 hrs. post, p=0.02). Furthermore, rates of same-day discharge increased significantly post-ERAS (49.1% for pre vs. 63.2% for post, p=<0.01). There was no significant difference in readmission rates (15.7% pre vs. 18.1% post, p=0.34). When stratifying into pre-ERAS vs. post-ERAS, same-day discharge rates were 50.4% vs. 45.5% pre-ERAS, and 63.3% vs. 62.9 post-ERAS for whites vs. non-whites respectively (figure). Lastly, cases started before 1pm were significantly more likely to be discharged on the same day as the surgery (OR 1.26, p=0.07).

Conclusion: As an example of value-based care, ERAS is effective in reducing hospital duration by increasing rates of same day discharge through identifying borderline patients who would otherwise stay an additional day. Additionally, ERAS is effective in reducing racial disparity in care, suggesting that standardization of care pathways may reduce bias in decision making. Lastly, future ERAS protocols may consider supplementing current practices with systems level interventions, such as starting complex cases that are eligible for same-day discharge before 1 pm.

 

32.05 What Data Do Patients Want to Use in Choosing a Provider? A National Survey of Patient Preferences

Posted on January 31, 2019

R. J. Ellis1,2, D. B. Hewitt3, J. K. Johnson4, K. Y. Bilimoria1  1Northwestern University,Department Of Surgery, Surgical Outcomes And Quality Improvement Center,Chicago, IL, USA 2American College of Surgeons,Chicago, IL, USA 3Thomas Jefferson University,Department Of Surgery,Philadelphia, PA, USA 4Northwestern Feinberg School of Medicine,Center For Healthcare Studies In The Institute Of Public Health And Medicine,Chicago, IL, USA

Introduction:  Patient utilization of healthcare quality reporting has been suboptimal despite attempts to encourage use of publicly available data. Lack of utilization may be due to discordance between reported quality metrics and what patients want to know when making healthcare choices. The objectives of this study were (1) to identify hospital- and physician-level measures of quality that patients would prefer presented in public reporting and (2) to explore the relative importance of these factors in how patients assess healthcare quality during decision making.

Methods:  Interviews and focus groups were used to develop and refine a survey exploring the relative importance of healthcare quality measures. Measures were studied across all domains of healthcare quality, including hospital-level measures (e.g., location, accreditations, hospital-level outcomes) and physician-level measures (e.g., years of experience, training program, adherence to national guidelines). The survey was administered to online survey panels through SurveyGizmo to obtain a census balanced national sample. Likert scale responses were compared using non-parametric tests of central tendency. Rank order responses were compared using analysis of variance testing adjusted for multiple comparisons. Associations with binary outcomes were analyzed using multivariable logistic regression models.

Results: The survey was sent to 11,125 individuals with 4,672 responses (42.0% response rate). Census balancing yielded 2,004 surveys for analysis. Of those, 1,213 (60.5%) reported previously researching healthcare online. Measures identified as most important were hospital reputation (considered important by 61.9%), physician years of experience (51.5%), and primary care physician recommendations (43.2%). Relatively unimportant factors included adherence to national guidelines (17.6%), risk of requiring temporary nursing home care (17.5%), and hospital academic affiliation (13.3%, p<0.001 for all compared to most important factors). Outcome measures were not among the most important factors, with the risk of death considered important by only 35.1% of respondents. Patients were unlikely to rank outcome measures as the most important factors in choosing healthcare providers, irrespective of age, gender, educational status, or income.

Conclusion: In selecting providers, patients valued hospital reputation, physician experience, and primary care physician recommendations. Publicly reported metrics like guideline adherence and patient outcomes were considered less important, despite a national push to focus on outcomes in public reporting. Public quality reports contain information that patients perceive to be of relatively low value, which may contribute to low utilization of public healthcare quality information. Development of reporting systems focused on relevant, patient-centered information may improve patient utilization of publicly reported quality data.

32.04 Behind the Mask: Gender Bias Experiences of Female Surgeons

Posted on January 31, 2019

K. L. Barnes1, L. McGuire3, G. Dunivan1, A. Sussman4, R. McKee2  1University Of New Mexico HSC,FPMRS/Obstetrics And Gynecology,Albuquerque, NM, USA 2University Of New Mexico HSC,Colorectal Surgery/General Surgery,Albuquerque, NM, USA 3University Of New Mexico HSC,School Of Medicine,Albuquerque, NM, USA 4University Of New Mexico HSC,Family And Community Medicine,Albuquerque, NM, USA

Introduction: The number of female surgeons continues to rise, however reports of sexism and stigmatization in the clinical setting continue. Although, overt sexism is becoming increasingly rare, underlying prejudices held against women are frequently expressed as microaggressions- subtle discriminatory or insulting actions that communicate demeaning or hostile messages at the interpersonal level.  We sought to assess the frequency and severity of gender-based microaggressions experienced by female surgeons.

Methods: This mixed methods approach utilized both focus groups and questionnaires to explore female surgeons’ experiences of gender bias in the form of microaggressions. The Sexist Microaggression Experiences and Stress Scale (Sexist MESS), a validated, 44-item questionnaire, was used to quantify the frequency and psychologic impact of gender-based microaggressions. This questionnaire consists of six domains, with higher scores indicating more frequent or severe microaggression impact. We conducted focus groups with female surgeons to explore their unique experiences of workplace gender bias and developed 15 additional questions. These questions were added to the survey and sent to all female resident, fellow and attending surgeons at a single academic institution.

Results: Four focus groups including 23 female trainee and attending surgeons were conducted revealing four emerging themes: Exclusion, Adaptation, Increased Effort, and Resilience. The survey response rate was 64.3% (65/101 surgeons). Survey data showed that the frequency and severity of microaggressions was higher in 5 of 6 domains for trainees compared to attending surgeons (Table 1), with the exception of “Inferiority”. When Obstetrician Gynecologists (OB/GYN) were compared to all other surgeons, rates of reported microaggressions were similar in all domains except “Leaving Gender at the Door”. Non-OB/GYN surgeons reported more pressure to downplay, hide or avoid characteristics and behaviors associated with femininity in order to succeed. The variables of non-white race, currently providing childcare, and number of years in practice after training did not demonstrate statistical significance.

Conclusion: The extent and psychological impact of microaggressions experienced by female surgeons varies based on level of training, with higher rates reported by trainees compared to attending surgeons. The type of surgical specialty practiced made little difference, with OB/GYN and non-OB/GYN surgeons describing similar experiences. Higher frequency and severity scores for the domain “Leaving Gender at the Door” reported by non-OB/GYN surgeons may be explained by the lower proportion of women in these fields compared to OB/GYN. 

32.03 Prospective Identification of Costly Surgical Episodes

Posted on January 31, 2019

K. R. Chhabra1,3,4, U. Nuliyalu4, J. B. Dimick2,3,4, H. Nathan2,3,4  1Brigham And Women’s Hospital,Department Of Surgery,Boston, MA, USA 2University Of Michigan,Department Of Surgery,Ann Arbor, MI, USA 3University Of Michigan,IHPI Clinician Scholars Program,Ann Arbor, MI, USA 4University Of Michigan,Center For Healthcare Outcomes And Policy,Ann Arbor, MI, USA

Introduction:
Surgery accounts for almost half of inpatient spending, much of which is concentrated in a subset of high cost patients. A method of prospectively identifying high cost patients, i.e. “hot spotting,” may help manage population health spending, but we lack an optimal way to predict which patients will have high-cost surgical episodes.

Methods:
Using 100% Medicare claims data, we identified patients aged 65-99 undergoing elective inpatient surgery (CABG, colectomy, total hip/knee replacement) in 2014. We calculated price-standardized Medicare payments for the surgical episode from admission through 30 days after discharge (episode payments). Based on predictor variables from 2013, e.g. Elixhauser comorbidities, hierarchical condition categories, Medicare’s Chronic Conditions Warehouse (CCW), and total spending, we constructed models to predict the costs of surgical episodes in 2014. We used general linear mixed models incorporating hospital random effects and adjusting for age, sex, and race, testing fit with R2 and kappa statistics (κ) using quintiles of spending.

Results:
A cost prediction model based on CCW score performed well in predicting payment variation for all procedures (R2 0.16-0.22, κ 0.13-0.15; all P<0.001). Other models also had statistically significant R2 and κ but had inferior predictive performance to CCW. The costliest quintile of patients as predicted by the model captured 40-50% of the patients in each procedure’s actual costliest quintile. For example, in CABG, 48% of the costliest quintile was predicted by the model’s costliest quintile. A greater share of the costliest quintile was identified when the prediction threshold was lowered; e.g. in CABG 73% of the actual costliest quintile was identified by combining the model’s 2 top quintiles of predicted cost.  

Conclusion:
Expensive surgical patients can be prospectively identified using readily available data on patients’ chronic conditions. The sensitivity of the cost prediction model can be tailored as desired. For instance, if attempting to identify as many potentially expensive patients as possible, one may lower the threshold for detection by combining quintiles of predicted cost. 

32.02 Impact of Medicare Readmissions Penalties for Surgical Conditions

Posted on January 31, 2019

K. R. Chhabra1,2,4, A. M. Ibrahim2,3,4, J. R. Thumma2, J. B. Dimick2,3,4  1Brigham And Women’s Hospital,Department Of Surgery,Boston, MA, USA 2University Of Michigan,Center For Healthcare Outcomes And Policy,Ann Arbor, MI, USA 3University Of Michigan,Department Of Surgery,Ann Arbor, MI, USA 4University Of Michigan,IHPI Clinician Scholars Program,Ann Arbor, MI, USA

Introduction:
The Hospital Readmissions Reduction Program, announced in 2010 to penalize excess readmissions for patients with select medical diagnoses, was expanded in 2013 to include targeted surgical diagnoses beginning with hip and knee replacements. The impact of these procedure-specific penalties for targeted procedures is not well understood.

 

Methods:  
A retrospective review of 2,627,974 patients who underwent total hip replacement or total knee replacement from 100% fee-for-service Medicare claims.  We used an interrupted time series model to assess hospital rates of readmission before the Hospital Readmissions Reduction Program was announced (2008-2010), during implementation for medical conditions (2010-2013) and after specific hip/knee replacement penalties were announced (2013-2016). We also assessed trends in length of stay and the use of observation status.

 

Results:
From 2008 to 2016, readmission rates declined for total hip replacement (7.8%–5.6%) and total knee replacement (7.0%–5.2%). Readmission rates were decreasing in 2008-2010, but the decline accelerated after the announcement of Hospital Readmissions Reduction Program in 2010 (slope change ?0.06 to -0.09). Readmissions continued to decrease after targeted surgical procedures were announced in 2013, but at a slower rate (slope change -0.09 to -0.05). During the same time period, mean length of stay decreased (3.6 d–2.4 d for total hip replacement, 3.6–2.5 d for total knee replacement). There was no change in the trend of observation status use before and after the Hospital Readmission Reduction Program.

 

Conclusions:
The major reduction in surgical readmissions rates occurred after the initial announcement of the Hospital Readmission Reduction Program, rather than after penalties for specific surgical procedures were announced. These findings suggest that the initial policy (for medical conditions) had a broad spillover effect, leading to improvements in surgical readmission rates as well.

32.01 Human Performance Analysis in Surgery – Cognitive and Technical Performance Improvement Opportunity

Posted on January 31, 2019

J. Suliburk1, C. Ryan1, Q. Buck1, C. Pirko1, N. Massarweh1, N. Barshes1, S. Awad1, S. R. Todd1, H. Singh2, T. Rosengart1  1Baylor College Of Medicine,Department Of Surgery,Houston, TX, USA 2Baylor College Of Medicine,DeBakey VA Medical Center For Innovations In Quality, Effectiveness And Safety / Department Of Internal Medicine,Houston, TX, USA

Introduction: Surgical quality improvement efforts have largely focused on data registries and process improvement strategies to overcome the role of human performance deficiencies (HPD) in catalyzing adverse outcome.  Limited data are available in quantifying the prevalence and types of HPD in surgical complications. The purpose of this prospective study was to develop and deploy a novel taxonomy tool for analyzing cognitive, technical and team HPD during the provision of surgical care to understand how to improve safety and quality.

Methods:   A prospective multicenter study involving 3 adult affiliate hospitals (level 1 trauma center, quaternary care university hospital and a VA hospital) at a large academic medical center was conducted over a 6-month interval in 2018.  An HPD tool was developed through systematic literature review according to PRISMA guidelines followed by Delphi consensus among medical error experts.  This tool classified HPD into 5 major categories related to cognitive, technical and team dynamic functions (Table).  Training of all surgeons in error taxonomy and categorization occurred thru an initial 2-week run-in period.  We then used the tool in weekly concurrent reporting of complications to categorize HPDs for all major adverse surgical outcomes across our 3 study site quality improvement conferences.  Surgeons self-assigned preliminary HPD classification to case complications, which were then adjudicated by a 3-person investigator panel following a service-wide case presentation and discussion.   

Results:  5365 cardiothoracic, surgical oncology, transplant, elective general surgery, acute care surgery, and vascular surgery cases were analyzed.  The overall major complication rate was 5% (188 complications).  Of these, 56% (n = 106) were HPD-related:  50% execution error, 31% cognitive dissonance, 13% communication error, 4% teamwork error, and 5% rules violation. The average number of HPDs per case was 1.8 ± 0.9. The frequency and distribution of HPDs was similar across sites, with cognitive bias in decision of care being most common (HPD Class IA.3) subtype, followed by recognition error (HPD Class IIA). HPD most commonly occurred postoperatively (58%), followed by intraoperatively (32%) and preoperatively (10%).

Conclusion:  HPD was identified in over half the instances of major surgical complications at a major academic medical center, most typically related to cognitive dissonance and execution of care. The prevalence of these HPDs suggests opportunity for enhanced education and training to reduce the incidence of HPD contributing to adverse outcomes.    The newly developed taxonomy provides a framework to facilitate quality improvement in understanding human error in surgery.

 

31.10 Variability Between Lateral and Anterior-Posterior (AP) Sacral Ratios in Anorectal Malformations

Posted on January 31, 2019

H. Ahmad1, D. R. Halleran1, A. Akers1, V. Alexander1, M. Levitt1, R. J. Wood1  1Nationwide Childrens Hospital,Center For Colorectal And Pelvic Reconstruction,Columbus, OH, USA

Introduction: The sacral ratio (SR) has been used as a tool to evaluate sacral development in patients with anorectal malformations (ARM) and to help (along with the type of ARM and spinal status) to predict future bowel control. Although the ratio can be calculated using images from either the AP or lateral planes, lateral images are believed to produce more reliable ratios, given that the calculation is not influenced by the tilt of the pelvis. The congruency of the sacral ratio in the AP and lateral planes has not been previously investigated. We therefore aimed to assess the variability in the AP and lateral sacral images.

Methods: We reviewed all patients with ARM treated at our institution  between 2014 and 2018 who had both an AP and lateral image of their sacrum. The SR was calculated using the ratio of the distance from the sacroiliac joint to the tip of the coccyx to the distance from the top of the iliac crest to the sacroiliac joint. All ratios were calculated by a pediatric radiologist. Variation between the SRs as determined by the AP and lateral images were compared across all patients and by ARM type using sacral ratio categories (0-0.39, 0.40-0.69, >0.70)  that were developed for the purpose of counseling families.

Results: 561 patients were included in the study. SRs in the AP plane varied by an average of 17% (IQR 4,25, range 0-154). The AP SR overestimated the lateral SR in 23% (N=128) and underestimated the lateral SR in 63% (N=354) of patients. The variability in measurements decreased with increasing sacral development, as patients with a severe hypodevelopment (SR <0.4, N=39) demonstrated a variation of 27%, patients with moderate hypodevelopment (SR 0.4-0.69, N=193) demonstrated a variation of 18%, and patients with normal sacral development (SR >0.7, N=329) demonstrated a variation of 15%. The difference in these groups was statistically significant (p=0.03).

Conclusion: The SR determined by images in the AP plane varied significantly from that measured using lateral images. These results demonstrate that the AP sacral ratio can lead to a significant misinterpretation of the degree of sacral development which would impair the ability to accurately counsel families on their child’s future continence potential. Based on these data, we recommend the lateral SR to be used as the preferred measure. The AP view remains valuable to assess for hemisacrum. 

 

31.09 Predicting Crohn's Disease Surgery Complications: Harvey Bradshaw Index vs ACS NSQIP Risk Calculator

Posted on January 31, 2019

K. R. McMahon1, C. Cordero-Caballero1, A. Afzali3, S. Husain2  1The Ohio State University Wexner Medical Center,College Of Medicine,Columbus, OHIO, USA 2The Ohio State University Wexner Medical Center,Division Of Colon & Rectal Surgery,Columbus, OHIO, USA 3The Ohio State University Wexner Medical Center,Division Of Gastroenterology, Hepatology, And Nutrition,COLUMBUS, OHIO, USA

Introduction: Gastroenterologists commonly use the Harvey-Bradshaw Index (HBI) to assess the severity of Crohn’s disease (CD) and to guide medical therapy. Surgeons, on the other hand, often use the ACS NSQIP Surgical Risk Calculator to determine surgical risk when treating patients with CD. However, the ACS NSQIP calculator does not account for CD as a risk factor even though it has been shown to be an independent predictor of poor postoperative outcomes. The utility of the HBI to predict surgical complications has not been studied. The aim of our study was to compare the ability of HBI and ACS NSQIP Surgical Risk Calculator to predict surgical complications in patients with Crohn’s disease. We hypothesized that HBI is a superior method of predicting surgical complications in this patient population.

Methods: A retrospective chart review was done to identify patients who underwent surgery for CD and the post-operative complications. Patients who had an HBI calculated prior to, but within one month of surgery, were identified. The ACS NSQIP Surgical Risk Calculator was used to calculate each patient's predicted risk. The group was divided into high and low risk based on the calculator’s listed probability for any complication. Patients were also divided into a low disease activity and high disease activity based on their HBI. Fisher’s exact test, unpaired t-test, and chi-square distribution were used for statistical analysis.

Results: 

A total of 61 patients were included. The average age was 37 years old. 40% were male and 60% female. The overall complication rate was 33%.

There was no significant difference between the high disease activity and low disease activity HBI groups in age, gender, ASA class, steroid use, or NSQIP calculated risk of any complication. There was no significant difference between the NSQIP calculated high and low-risk groups in age, gender, steroid use, or HBI. The higher risk group did correspond to a higher ASA class; this relationship achieved statistical significance (p=0.0113).

The high disease activity HBI group had significantly more surgical complications than the low disease activity group. Additionally, the high disease activity group also had a significantly longer length of hospital stay (table 1). There was no significant difference between the NSQIP calculated high and low-risk groups for surgical complications or length of hospital stay (table 1).     

Conclusion: HBI score appears to be a better predictor of postoperative outcomes than the commonly used ACS NSQIP Surgical Risk Calculator. Further study is needed to examine the relationship between HBI and surgical risk prospectively and in a larger population of patients.   
 

31.08 Do Prolonged Operative Times Obviate the Potential Benefits Associated with Laparoscopic Colectomy?

Posted on January 31, 2019

P. J. Sweigert1, E. Eguia1, A. N. Kothari1, K. A. Ban1, M. H. Nelson1, M. S. Baker1, M. A. Singer1  1Loyola University Medical Center,Department Of Surgery,Maywood, IL, USA

Introduction:
Prior studies have demonstrated that minimally invasive (MIS) approaches to colectomy are oncologically equivalent and associated with shorter lengths of stay, and reduced morbidity in comparison to open approaches to colectomy. There is also increasing evidence that prolonged operative time, especially that greater than 3 hours, is associated with increased rates of postoperative morbidity. Few studies examine the impact of operative time (OT) on the potential benefits afforded by MIS colectomy. We sought to determine if benefits associated with MIS colectomy are maintained in cases where OT is significantly prolonged. 

Methods:
The American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) procedure targeted colectomy database was queried to identify adult patients who underwent elective left (LC) and right (RC) colectomy with anastomosis between 2011 and 2016. Emergent or converted cases, patients with preoperative infections, and observations with missing OT data were excluded. Forward stepwise multivariable logistic regression adjusting for demographic and clinical risk factors was used to compare outcomes for prolonged (4th quartile) MIS cases to average (2nd-3rd quartile) open cases with 30-day mortality or serious morbidity as the primary outcome of interest. Secondary 30-day outcomes included any morbidity, mortality, anastomotic leak, surgical site infection (SSI), and prolonged length of stay (LOS). 

Results:
18,274 patients underwent RC and 54,550 LC during the study period. RC was most commonly performed for colon cancer (48.6%). Median OT for open RC was 132 min (IQR 92-189).  That for MIS RC was 135 min (IQR 103-175), p=0.010. LC was also most commonly performed for colon cancer (44.9%). Median OT for open LC was 171 min (IQR 119-242). That for MIS LC was 173 min (IQR 129-231), p=0.001. No difference was seen in the adjusted primary outcome when prolonged MIS cases were compared to average open RC (OR 0.818, 95% CI: [0.660, 1.014]).  Prolonged MIS cases did, however, show a significant benefit in comparison to average length open procedures for adjusted mortality or serious morbidity in LC (OR 0.824, 95% CI: [0.729-0.931]). Prolonged MIS approach was associated with significantly lower morbidity, SSI and LOS for both RC and LC relative to average open colectomy (Table 1). 

Conclusion:
MIS approaches to colectomy are associated with improved rates of postoperative complications, serious morbidity and 30-day mortality relative to open approaches to colectomy.  These benefits are maintained even with OT that extends beyond 3 hours.  Surgeons performing elective MIS colectomy are justified in persisting with prolonged attempts at MIS colectomy. 
 

31.07 Postoperative Length of Stay Following Colorectal Surgery Impact Readmissions

Posted on January 31, 2019

X. L. Baldwin1, P. D. Strassle1, S. Lumpkin1, K. Stitzenberg1  1University Of North Carolina At Chapel Hill,Chapel Hill, NC, USA

Introduction:
Enhanced recovery pathways have led to shorter lengths of stay (LOS) after colorectal surgery. There continues to be controversy about the relationship between LOS and readmission following colorectal surgery. The purpose of this study was to evaluate the association between LOS and readmissions in a nationally representative sample. We hypothesized that shorter LOS would increase readmission rate. 

Methods:
Hospitalizations of adult patients aged 18 – 85 years old, who underwent colon and/or rectal resection between January 2010 and August 2015 in the National Readmission Database were eligible for inclusion. Patients who were treated in December, who were not residents of the state in which they underwent surgery, who died, or who had a LOS <1 day were excluded. Multivariable logistic regression was used to assess the effect of LOS on 30-day readmission, adjusting for patient demographics, comorbidities, hospital characteristics, and inpatient complications.

Results:
We assessed 376,376 hospitalizations. Median LOS was 5 days (IQR 4-8) and 14% of patients (n=51,087) were readmitted within 30 days. As LOS increased, the incidence of readmissions also increased (7% patients with 1 day LOS to 29% in patients with LOS ≥20 days, p<0.0001), Figure 1. After adjustment for patient demographics, comorbidities, inpatient complications, and hospital characteristics, a 5-8 day LOS was associated with a 50% increase in odds of 30-day readmission (OR 1.53, 95% CI 1.05, 1.14, p<0.0001), a 9-12 day LOS was associated with a 100% increase in odds (OR 2.06 95% CI 1.99, 2.13, p<0.0001), and a LOS ≥13 days was associated with an almost 150% increase in odds (OR 2.45 95% CI 2.36, 2.55, p<0.0001), compared to a 1-4 day LOS.

Conclusion:
Contrary to our hypothesis, we found that an increased length of stay resulted in increased readmission rates following colorectal surgery. Shorter LOS decreased the odds of 30-day readmission, regardless of a patient’s pre-existing comorbidities and inpatient complications. Future studies should examine factors associated with prolonged hospitalization and identify possible interventions to decrease readmission in these patients.
 

31.06 Effect of Elective Sigmoidectomy for Diverticulitis on Bowel Function Patient-Reported Outcomes

Posted on January 31, 2019

J. L. Goldwag1,2, R. V. Lyn3, L. R. Wilson1,2, M. Z. Wilson1,2, S. J. Ivatury1,2  1Dartmouth Hitchcock Medical Center,The Department Of Surgery,Lebanon, NH, USA 2Dartmouth Medical School,Lebanon, NH, USA 3Dartmouth College,Hanover, NH, USA

Introduction:
Diverticular disease is common worldwide. A subset of these patients will choose to undergo elective surgical resection due to symptoms or complicated disease. The effect of elective sigmoid colon resection for diverticular disease on bowel function is unclear. The aim of this study is to evaluate changes in bowel function following elective sigmoid resection for diverticular disease.

Methods:
This is a prospective, observational study. We included all patients seen at our institution from May 2015 to July 2018 who underwent elective sigmoid resection for diverticular disease. We used the Colorectal Functional Outcome Questionnaire (COREFO), a validated bowel function questionnaire that assesses bowel function in five domains and a global function score. Scores range from zero to 100 with a higher score indicating worse function. We obtained questionnaire data at baseline as well as at postoperative follow-up. Patients were included if they completed both questionnaires and underwent elective sigmoid resection.  Patients were excluded if they remained in bowel discontinuity or did not complete both questionnaires. A paired t-test was used to compare baseline and post-intervention scores.

Results:
49 patients met criteria for inclusion in this study. The median time between questionnaire completion was 70 days (IQR: 56 to 85). The mean age was 60 ± 12 years with 57% female patients. 36 (73%) patients underwent sigmoidectomy alone and 13 (27%) underwent sigmoidectomy with fistula repair. Six patients (12%) had a diverting loop ileostomy in addition to sigmoidectomy and underwent a subsequent reversal. Overall, there was no difference in Total COREFO score from baseline to post-intervention.  There were also no differences in any of the five COREFO domains (Figure 1).

Conclusion:
Bowel function does not change in the postoperative period following elective sigmoid resection for diverticular disease. Surgeons should counsel patients, especially symptomatic ones, that bowel function will be no different at time of postoperative follow-up.
 

31.04 Transverse Abdominis Plane Block Vs Intrathecal Analgesia in Colorectal Surgery: A Randomized Trial

Posted on January 31, 2019

D. Colibaseanu1, O. Osagiede2, A. Merchea1, C. Thomas6, E. Bojaxhi3, J. Panchamia5, A. Jacob5, S. Kelley4, K. Mathis4, A. Lightner4, J. Naessens6, D. W. Larson4  1Mayo Clinic – Florida,Section Of Colon And Rectal Surgery,Jacksonville, FL, USA 2Mayo Clinic – Florida,Health Sciences Research,Jacksonville, FL, USA 3Mayo Clinic – Florida,Department Of Anesthesiology,Jacksonville, FL, USA 4Mayo Clinic,Division Of Colon And Rectal Surgery,Rochester, MN, USA 5Mayo Clinic,Department Of Anesthesiology,Rochester, MN, USA 6Mayo Clinic,Health Sciences Research,Rochester, MN, USA

Introduction: Transversus abdominis plane (TAP) block is an effective alternative to neuraxial analgesia in abdominal surgery; however, limited evidence supports its use over traditional analgesic modalities in colorectal surgery. We compared the analgesic efficacy of liposomal bupivacaine TAP block and intrathecal (IT) opioids in a prospective randomized trial. The primary outcomes were the mean pain score and morphine milligram equivalents (MME) used within the first 48 hours post-surgery. Secondary outcomes included length of stay, standardized costs, postoperative ileus, and intravenous patient-controlled analgesia use. 

Methods: Patients were recruited from two campuses of a single institution. Two hundred and nine patients undergoing elective small bowel or colorectal resections were enrolled. They were randomized to receive either bilateral TAP block or single-injection IT analgesia with hydromorphone. Patients were assessed at 4, 8, 16, 24, and 48 hours post-surgery.

Results: Two hundred patients completed the trial (TAP =102, IT N=98). The TAP group had a mean pain score 1.7 points higher than the IT group 4 hours post-surgery, persisting up to 16 hours post-surgery. There was evidence of higher MME use < 24 hours post-surgery in the TAP group compared to IT (median difference: 10.0 MME, 95% CI 3.0 – 20.5 MME). No difference in MME was observed between the two groups at 24 and 48 hours, or in secondary outcomes.

Conclusions: Intrathecal opioids provided better immediate postoperative pain control compared to liposomal bupivacaine TAP block, lasting up to 16 hours post-surgery. Both modalities provided adequate pain control in patients enrolled in this study, and should be considered as part of a multimodal postoperative analgesic plan for patients undergoing elective colorectal surgery.

 

 

31.05 Opioid Tolerance Impacts Major Abdominal Surgery Outcomes in Patients on Enhanced Recovery Pathway.

Posted on January 31, 2019

M. H. Zaman1, O. P. Owodunni2, M. Ighani2, M. Grant3, D. Bettick4, S. Sateri3, T. Magnuson2, S. Gearhart2  1The Johns Hopkins University School Of Medicine,Urology,Baltimore, MD, USA 2The Johns Hopkins University School Of Medicine,Surgery,Baltimore, MD, USA 3The Johns Hopkins University School Of Medicine,Anesthesia,Baltimore, MD, USA 4The Johns Hopkins Bayview Medical Center,Quality,Baltimore, MD, USA

Introduction: Chronic opioid exposure can lead to a state of tolerance in patients where increasing doses of opioids are necessary to reduce pain; this can make postoperative management difficult. An Enhanced Recovery Pathway (ERP) is an evidence-based intervention that focuses on optimizing recovery and postoperative outcomes. The effectiveness of an ERP depends on the degree of compliance with the pathway. We wish to determine the effects of opioid tolerance in patients undergoing abdominal surgery on an ERP and its impact on compliance and postoperative outcomes. 

Methods:  From January 2013 to June 2017, patients undergoing major abdominal surgery prior to and following ERP-implementation were included. Patients <18 years and having emergency surgery were excluded. Compliance was measured to 14 perioperative pathway variables and high-compliance was defined as achieving ≥75%. Opioid tolerance was defined as any patient taking a prescribed opioid medication equivalent to 60 mg of Morphine per day for one week prior to surgery.  CR-POSSUM scores were used for risk-adjusted analyses. Outcomes of interest include length of stay (LOS), major-complications (Clavien-Dindo“CD”≥2), and 30-day readmission rates. 

Results: 1251 patients (605 pre-ERP and 646 ERP patients) were included. A total of 221 patients were opioid tolerant. Opioid tolerant patients were more likely to be younger (56 vs. 59 years, P=0.002), have disseminated cancer (11% vs. 5 %, P=0.003) and have an open procedure (69% vs. 60%, P=0.01) than non-tolerant patients. When comparing opioid-tolerant patients prior to (107 patients) and following EPR implementation (114 patients), there was no difference in demographic and clinical characteristics; however, more opioid-tolerant patients following ERP implementation had a laparoscopic procedure (42% vs. 19%, P<0.001).  In a multivariable analysis, opioid tolerance was associated with an increase in major complications (OR 1.24, p=0.032) and in readmissions (OR 1.42, p=0.005).  Among the ERP cohort, opioid-tolerant patients were less likely to be highly compliant with ERP variables than non-tolerant patients (35% vs. 54%; p<.001). Opioid tolerance was associated with a higher median LOS (5 days vs. 4 days; p<0.02) and a higher readmission rate (24% vs. 13%; p<0.01) than non-opioid tolerant ERP patients. In opioid-tolerant patients, high compliance with ERP was associated with a decreased odds of major complications (OR: 0.10, p<.001) and a reduction in the readmission rate (OR: 0.7, p=0.003). Opioid tolerance was an independent predictor of non-compliance with ERP (OR: 0.44, p<.001).

Conclusion: We provide evidence that opioid tolerance is associated with less favorable outcomes in patients undergoing major abdominal surgery and on an ERP; this is likely due to lack of pathway compliance.  Minimizing opioid use prior to elective major abdominal surgery may improve compliance and postoperative outcomes.

« Previous Page
Next Page »