53.08 Preoperative Opioid Use is Associated with Higher Costs and Morbidity after Abdominal Surgery

Posted on January 18, 2016

D. C. Cron1, M. J. Englesbe1, K. L. Carrier3, C. Bolton2, M. Joseph2, S. Moser2, J. F. Waljee1, P. E. Hilliard2, S. Kheterpal2, C. M. Brummett2 1University Of Michigan,Surgery,Ann Arbor, MI, USA 2University Of Michigan,Anesthesiology,Ann Arbor, MI, USA 3University Of Michigan,College Of Pharmacy,Ann Arbor, MI, USA

Introduction:
Opioids are increasingly used to manage chronic pain, and chronic opioid users are among the most challenging patients to care for perioperatively. Given the epidemic of opioid-related morbidity and mortality, it is critical to understand how opioid use impacts surgical outcomes. In this context, we examined the effect of preoperative opioid use on clinical and financial outcomes in adults undergoing major elective abdominal surgery.

Methods:
We retrospectively reviewed 2460 patients drawn from a statewide surgical quality collaborative database who underwent elective abdominopelvic surgery at a single center between 2008-2014. We identified preoperative opioid use from the medical record at the anesthesia preoperative evaluation (day of operation). Opioid use was categorized first as any opioid use (binary) and then as none vs. short-acting only vs. long-acting +/- short-acting. Outcomes included 90 day total hospital costs, hospital length of stay (LOS), discharge destination, 30 day hospital readmissions, and 30 day major complication rates. Standard multivariate regression techniques were used to adjust for case complexity and patient-specific risk factors (age, ASA class, functional status, smoking, medical comorbidities).

Results:
497 patients (20%) used opioids preoperatively. After covariate adjustment, opioid users (compared to opioid-naïve) had higher mean costs ($27091 vs. $24307, P<0.01), longer mean LOS (6.0 vs 5.2 days, P<0.01), increased readmissions (10% vs 6%, P<0.01), higher complication rates (20% vs. 16%, P=0.04), but there was no significant difference in non-home discharge rates (14% vs. 11%, P=0.11). Surgical site infection, sepsis, and postop transfusion were significantly more common in opioid users. When categorized by short vs. long-acting opioids, both levels of opioid use were significantly associated with costs, LOS, and readmissions in multivariate analyses. The effect of long-acting opioids was stronger than that of short-acting opioids across all outcomes (except discharge destination). The figure shows a stepwise increase in adjusted mean costs and LOS across categories of opioid use, with long-acting opioid users having the highest costs and LOS.

Conclusion:
Opioid use is common prior to abdominal surgery and is associated with increased healthcare utilization and morbidity, even after risk-adjustment. Preoperative opioids represent a potentially modifiable risk factor and a novel target for surgical quality improvement. Going forward, strategies to reduce opioid use and implement alternative pain management options may improve the quality and value of surgical care in this vulnerable patient population.

53.09 Defining Expectations: Rates of Admission After Surgical Resection for Gastrointestinal Malignancies

Posted on January 18, 2016

K. T. Collier1, J. Tong1, G. Karakousis1, S. Dasani1, R. R. Kelz1 1University Of Pennsvlvania,Department Of Surgery,Philadelphia, PA, USA

Background: Despite efforts to select medically appropriate patients for surgery, some patients remain high risk. We sought to examine admissions following surgical resection of 4 intra-abdominal malignancies to better prepare patients for the surgical experience and to identify potential patients at risk for multiple admissions.

Methods: Patients with an established diagnosis of colon, rectal, stomach or pancreas cancer who were admitted to the hospital for the first time since diagnosis and underwent surgical resection In New York or California (2008-2011) were identified for study inclusion. Patients were excluded if they had a post-operative complication during the index admission. Rates of multiple admissions (MuAdm) at or exceeding the 90th percentile of admissions within each cancer type were tabulated. MuAdm was defined as >=2 admissions for colon and rectal cancer and >=3 admissions for pancreas and stomach cancer patients. Univariate analysis was performed to determine patient characteristics known at the time of admission that were associated with MuAdm. Independent multivariate logistic regression models were developed to identify patient cohorts at risk for MuAdm within each malignancy type. The Bonferroni correction was used to adjust for multiple comparisons.

Results: Of 32,601 patients, 22,761 underwent surgery for colon cancer, 5,875 rectal cancer, 1,594 pancreas cancer and 2,371 stomach cancer. The median age of patients for was 68 (IQR: 57-77), median number of comorbidities was 3 (IQR:3-5), 51.5% of patients were female, and 19% of study population was admitted through the emergency department. The overall 1-yr rate of any post-operative admission was 37.91%; with 33.9% for colon cancer, 47.4% rectal cancer, 49.6% pancreas cancer and 45.34% stomach cancer (p<0.001). The rate of MuAdm was 14% for colon cancer, 19.3% rectal cancer, 12.74% pancreas cancer and 11.05% stomach cancer (p<0.001). The number of MuAdm ranged from (2-26) for colon cancer patients, (2-16) for rectal cancer patients, (3-11) for pancreas cancer patients and (3-12) for stomach cancer patients. Factors associated with MuAdm on multivariate analysis, after Bonferroni adjustment for multiple comparisons, are shown in Table 1.

Conclusions: At least 40% of patients undergoing surgical resection of an intra-abdominal malignancy without complications will be admitted to the hospital at least one time in the subsequent year. For some patients, there will be multiple admissions. Risk factors for multiple admissions differ by tumor type. In patients with multiple risks, alternatives to the standard of care should be considered.

53.10 Impact of Queues and Congestion in Cost of Surgical Opportunity, A Modeling Based Analysis

Posted on January 18, 2016

J. D. Rueda1, A. Ruiz Patiño2 1University Of Maryland,School Of Pharmacy,Baltimore, MD, USA 2Hospital Universitario San Ignacio,Bogota, DC, Colombia

Introduction:
Congestion in the postanesthesia care unit (PACU) leads to formation of a waiting queue for patients transferred after surgery. As patient recovery is performed in the operating room, incoming surgeries are delayed. The purpose of this study was to establish the impact of this phenomenon on surgical lost time and finantial resources.

Methods:
An operational mathematical study based on the queuing theory was performed. Average queue length and average hourly queue waiting time were evaluated. Calculations were based on mean patient daily flow, PACU length of stay, occupation and current number of beds. Data was prospectively collected during a period of two months and the variables entry and exit time were recorded for each patient taken to the PACU. Data were imputed in a computational model made with MS Excel®. To account for data uncertainty, a series of deterministic sensitivity analyses for all independent variables were performed.

Results:
With a mean patient daily flow of 40.3, an average PACU length of stay of 4 hours 25 minutes and 15 beds, average total lost surgical opportunity time, based on occupation analysis, was estimated at 2.36 h (CI95%0.36 – 4.74h). Lost of potential surgical hour equals $1,592. Sensitivity analysis (Figure1) shows that an increase of 2 beds (investment cost of $18,234) is required to solve queue formation.

Conclusion:
Guided by queuing theory, solving congestion in the PACU could lead to an increase in number of daily surgeries and a better cost-opportunity relationship compared to the current capacity.

53.05 Variable Surgical Outcomes After Hospital Consolidations: Implications for Local Healthcare Delivery

Posted on January 18, 2016

V. Chang2,3, A. N. Kothari1,2, R. M. Yau2, J. M. Albright1, S. Besser5, R. H. Blackwell2,4, G. N. Gupta2,4, P. C. Kuo1,2 1Loyola University Medical Center,Department Of Surgery,Maywood, IL, USA 21:Map Surgical Analytics Research Group,Maywood, IL, USA 3Loyola University Chicago Stritch School Of Medicine,Maywood, IL, USA 4Loyola University Medical Center,Department Of Urology,Maywood, IL, USA 5Depaul University,College Of Computing And Digital Media,Chicago, IL, USA

Introduction:
While many healthcare policy analysts claim that hospital consolidations benefit both health system financial health and patient care by lowering cost and improving quality, very little can be found in literature that examines the latter claim. With more hospital consolidations as an inevitable part of our future healthcare ecosystem, we investigated the relationship between hospital consolidations and surgical outcomes. We hypothesize that hospital consolidations improve surgical outcomes.

Methods:
We conducted a retrospective study spanning 2007-2013 in the states of CA and FL using the Healthcare Cost and Utilization Project State Inpatient Database, AHA Annual Survey Database, and Medicare’s Case Mix Index data. We identified 19 hospitals that consolidated over the study period and propensity matched the 19 hospitals with 19 independent hospitals using patient and hospital characteristics. Differences-in-differences analysis was used to compare the change in the risk-adjusted complication rate of 7 elective surgeries 1 year prior to consolidation and the year following consolidation between the consolidation hospitals and matched control group. Our complication index included post-op pulmonary failure, MI, renal failure, PNA, DVT, PE, SSI, hemorrhage, and GI bleed. Based on ProPublica’s surgeon scorecard, the elective surgeries included laparoscopic cholecystectomy, transurethral prostatectomy, cervical fusion of the anterior column (anterior technique), lumbar and lumbosacral fusion of the anterior/posterior column (posterior technique), total hip replacement, and total knee replacement.

Results:
Of the 7 procedures studied, 2 procedures saw a decrease in complication rate (lumbar and lumbosacral fusion of the posterior column posterior technique, DID=-0.6, p<0.01; total hip replacement, DID=-0.6, p<0.01), 3 procedures saw an increase in complication rate (TURP, DID=4.1, p<0.01; cervical fusion of the anterior column anterior technique, DID=1.5, p<0.01; total knee replacement, DID=0.3, p<0.01), and 2 procedures saw no change in complication rate (laparoscopic cholecystectomy, lumbar and lumbosacral fusion of the anterior column posterior technique, both p>0.05) following a hospital consolidation.

Conclusion:
Arguments have been made that consolidated health care systems can share high-performing clinical services and infrastructure resources such as electronic medical records to improve quality. Our results indicate that hospital consolidations result in improved outcomes, but not uniformly. Undoubtedly, many factors affect surgical outcomes following consolidation, such as volume, hospital perioperative resources, reason for consolidating, leadership, and physician factors.

53.06 Machine Learning and Cloud Computing for Enhanced Surgical Risk Prediction

Posted on January 18, 2016

Z. C. Dietch1, K. C. Lichtendahl2, Y. Grushka-Cockayne2, J. M. Will2, R. S. Jones1, R. G. Sawyer1 1University Of Virginia,Department Of Surgery,Charlottesville, VA, USA 2University Of Virginia,Darden Graduate School Of Business,Charlottesville, VA, USA

Introduction: Accurate risk prediction enhances surgical care by impacting perioperative planning, informed consent, patient selection, and outcome measures. Advances in machine learning, computing, and data availability have led to superior predictions in many disciplines by automating the process of variable selection in the context of vast amounts of data. Based on wisdom-of-crowds literature, we hypothesized that an ensemble, or average, of machine-learning prediction models would outperform mortality predictions issued by the American College of Surgeons National Surgical Quality Improvement Program (ACSNSQIP).

Methods: Preoperative characteristics and mortality outcomes from the ACSNSQIP Participant Use Files (PUF) 2005-2013 were used to form mortality predictions. The data were split into training and testing sets. The free open-source R statistical software was utilized on Microsoft’s Azure cloud environment to parallel-process variable selection and mortality predictions. Two machine-learning algorithms—regularized logistic regression and random forest—were fit to the training set to generate probabilistic mortality predictions for each observation in the testing set. The two model predictions were averaged to form an ensemble prediction. The primary outcome was the improved accuracy of mortality predictions of the ensemble compared to those of the ACSNSQIP model. The Brier score, a measure of accuracy represented by the mean squared difference between the predicted probability and actual outcome for a set of individual observations, was used to evaluate performance. Significance was determined using the Amisano-Giacomini test.

Results: Models were trained on 1,847,818 records, and predictions were tested on 615,939 records. The regularized logistic regression selected 93 predictors, including 11 new variables engineered using ACSNQIP PUF data. The random forest utilized 86 selected predictors, including 17 newly engineered variables. Predictions of the ensemble outperformed ACSNSQIP predictions as measured by Brier scores where lower is better (0.009854 vs. 0.009904, p=0.0065), representing a mean improvement of 0.5% over ACSNSQIP predictions. A graphical representation of model calibration is presented (Figure).

Conclusion: Advances in machine learning and cloud computing have enabled rapid, robust, and affordable predictive analysis. Utilizing such capabilities, we generated a predictive model that outperformed the ACSNSQIP model. Despite lacking institution-specific data used by the ACSNSQIP in their model, we outperformed ACSNSQIP predictions using otherwise identical data. Because outcomes vary by institution, we expect further improvement by including institution-specific data in our model.

53.07 Discretionary Surgical Procedures: International Variations in Utilization Among Developed Countries

Posted on January 18, 2016

J. T. Adler1, A. P. Loehrer1, E. A. Mort1, K. D. Lillemoe1, D. C. Chang1 1Massachusetts General Hospital,Boston, MA, USA

Introduction: Healthcare delivery and utilization varies internationally. It is unknown how much inpatient procedure utilization varies among developed countries, especially among discretionary procedures. We hypothesized the rates of discretionary admissions are higher in the US than other developed countries.

Methods: Previously published seven discretionary surgical procedures (appendectomy, abdominal aortic aneurysm [AAA] repair, carotid endarterectomy [CEA], open reduction and internal fixation [ORIF] of the femur, hysterectomy, and prostatectomy), were analyzed. The Global Comparators dataset from Dr Foster Intelligence, a global hospital benchmarking collaborative that includes all inpatient admissions to 18 academic medical centers in Australia, Holland, the United Kingdom (UK), and the United States (US) from 2008-2013 was used. With the UK as a reference, multivariable logistic regressions were performed for hospital admissions for each procedure versus all other hospital admissions, adjusting for age and gender differences.

Results: A total of 6,956,354 admissions were analyzed. Utilization patterns varied widely (Table); a higher OR indicates more common utilization of admissions for a particular procedure. The US had the highest rates of inpatient admissions for four procedures: AAA repair (OR 2.60, p<0.001), CEA (OR 2.81, P<0.001), hysterectomy (OR 1.98, p<0.001), and prostatectomy (OR 5.13, p<0.001). US also had significantly higher, although not the highest, rate for ORIF (OR 1.49, p<0.001) than UK. In contrast, US had lower rates for appendectomy (OR 0.79, p<0.001) relative to UK, both of which had the highest rates in Australia.

Conclusions: International utilization patterns vary considerably, with significantly higher rates of inpatient admissions for discretionary surgical procedures in the US. This association was especially true for resource-intensive procedures, but not for less-intensive operations. Analyses like this may lead to better understanding of healthcare system performance.

53.02 The value of colon surgery care bundles in Michigan

Posted on January 18, 2016

D. Z. Semaan1, A. P. Meka1, T. Jaffe1, J. Papin IV1, U. Okoro1, C. Hwang1, A. J. Mullard1, D. Campbell1, M. Englesbe1 1University Of Michigan,Department Of Surgery,Ann Arbor, MI, USA

Introduction:

Surgical site infections (SSIs) following colectomy are morbid and expensive. Previous studies within the Michigan Surgical Quality Collaborative (MSQC) have led to the implementation of a six-element perioperative care bundle for colectomy. The aim of this study was to determine the value (quality and cost) implications of implementation of these bundled perioperative care measures in Michigan.

Methods:

We identified 3,387 patients in the MSQC database who underwent colon surgery from 2012 to 2015. The patient population was stratified into two groups based on compliance: low compliance for adherence to 0-2 perioperative care measures and high compliance for adherence to 3-6 perioperative care measures. Factors potentially associated with SSI were tested using univariate statistical tests. Payment data (patient level) was aggregated in collaboration with Blue Cross Blue Shield of Michigan. Model adjustment covariates included patient demographic factors, clinical factors, case mix, and clustering of patients within hospitals. Adjustments were made to payments to manage hospital-level contractual differences in BCBSM episode payments. A hierarchical generalized linear model was created to test for independent associations between perioperative care measures, SSI and cost.

Results:

The high compliance and low compliance group had risk-adjusted SSI rates of 8.2% (95% CI, 7.21-9.2%) and 16.0% (95% CI, 12.9-19.1%), respectively. This correlates to an absolute risk reduction of 7.8% (p<0.01).

When compared with low compliance, the high compliance group had an absolute risk reduction of 3.6% (p<0.01), 2.9% (p<0.01) and 1.3% (p<0.01) for SSI rates in superficial space, deep space, and organ space, respectively.

Low compliance cases had an average episodic cost of $20,046 (95% CI, $17,281-$22,812) while high compliance cases had a total episodic cost of $15,272 (95% CI, $14,354-$16,192). This showed a $4,474 (95% CI, $1,859-$7,688) reduction in cost with high compliance cases (p<0.01). Included in this decrease is a $1,986 (14.8%) reduction in facility base payment and a $2,274 (43.9%) reduction in total professional payment. (Figure)

Conclusion:

High compliance with bundle elements shows a statistical reduction in surgical site infections and episode costs. The largest domain of cost savings is professional fee payments. Further implementation of perioperative care bundles will improve the value of colon surgery in Michigan.

53.03 90-day post-procedural outcomes of emergent stenting versus surgery in malignant large bowel obstruction

Posted on January 18, 2016

J. S. Abelson1, H. Yeo2, J. W. Milsom2, A. Sedrakyan3 1Weill Cornell Medical College,Department Of Surgery,New York, NY, USA 2Weill Cornell Medical College,Division Of Colon And Rectal Surgery,New York, NY, USA 3Weill Cornell Medical College,Division Of Comparative Effectiveness And Outcomes Research,New York, NY, USA

Introduction: Stenting for malignant large bowel obstruction (MLBO) has been used for palliation and bridge to surgery. There is little data comparing stenting to other interventions in a ‘real world’ setting. We report 90-day outcomes using a large database.

Methods: The NY State Department of Health SPARCS database was used to identify patients with emergency treatment for obstructing colon cancer. There were 4 relevant management groups: Group 1(stenting alone); Group 2(stenting followed by resection within 14 days;’bridge to surgery’); Group 3(primary resection with anastomosis); and Group 4(stoma creation with or without resection). Outcomes were major complications, readmission, reoperation, mortality, LOS, and discharge disposition.

Results: 2,435 patients were treated for MLBO from Oct. 2009 to Dec. 2013. 235 patients underwent stenting alone (Group 1). 57 patients had stenting as their initial procedure followed by resection within 2 weeks (Group 2). 1,388 patients had a primary resection with anastomosis (Group 3) and 755 had a stoma creation (Group 4). Group 2 had a trend towards better outcomes; given the small number (n = 57) and inability to reach statistical significance, these patients were removed from analysis. Among the remaining 3 groups, there were no statistically significant differences between age, gender, or comorbidity but whites were more likely to undergo primary resection (58.8%, 33.1%, 8.1%, p=0.01). High volume centers were more likely to perform stenting compared to low volume centers (19.4% vs 3.5%;p<0.01). LOS was shortest in the stent group (8 days vs 12 days (group 3) and 13 days (group 4); p<0.01) There were no differences in major complication rates between Group 1, 3, and 4 during the index hospitalization and 90 days post-procedure. Patients in Group 1 had the highest rate of 90-day readmissions compared to Group 3 and 4(42.6% vs 26.9% and 30.3%;p<0.01). The percent of patients who underwent reoperation within 90 days in Group 1, 3, and 4 was 17.0%, 2.7%, and 4.0% respectively(p<0.01). The 90-day mortality rate was highest in Group 4 followed by Group 1 and 3(15.6%, 10.6%, 9.4%; p<0.01). Patients in Group 1 were more likely to be discharged to hospice (7.6%) as compared to Group 4 (3.6%) and Group 3 (1.5%) (p<0.01).

Conclusion: Stenting remains an uncommon intervention for patients admitted and treated emergently for MLBO. Not surprisingly, it is most commonly performed at high volume centers, perhaps due to access to advanced endoscopy. The lack of widespread adoption of this intervention could be due to higher 90-day readmission and reoperations rates compared to surgical treatment and use predominantly in high volume centers.

53.04 Readmissions for Surgical Site Infections: A Novel Measure to Guide Hospital Readmission Reduction Efforts

Posted on January 18, 2016

R. F. Shah1, E. Pavey1, A. Dahlke1, M. Ju1, R. Merkow1, A. Yang1, R. Rajaram1, K. Bilimoria1 1Surgical Outcomes And Quality Improvement Center (SOQIC),Department Of Surgery, Feinberg School Of Medicine, Northwestern University,Chicago, IL, USA

Introduction: Readmissions have become a major focus of pay-for-performance programs in the past 5 years. Investigating the reasons for readmissions is necessary in order to help hospitals identify targets for improvement. Surgical site infections (SSI) have been found to be the main contributing factor to readmissions. Hospitals may differ in how they manage post-discharge SSIs (e.g. if they manage patients in an outpatient clinic), and thus readmissions for SSI may be a useful hospital-level quality indicator. Our objectives were to (1) investigate the variation that exists between hospitals with their rates of readmissions for SSI and (2) assess patient- and hospital level factors associated with readmissions.

Methods: Patients undergoing a colectomy at hospitals enrolled in the American College of Surgeons National Surgical Quality Improvement Program throughout January 1, 2012, and December 31, 2013 were included in the study population. Readmission rates and indications for those readmissions were assessed. Proportional hazards models were developed to examine risk-adjusted hospital variation and the association of patient and hospital factors with readmissions for SSI. We then ranked the hospitals in our study by their odds ratios, and calculated the correlation of this rank with overall hospital risk-adjusted SSI rates as well as overall hospital risk-adjusted readmission rates

Results: Based on 30,876 patients from 145 hospitals, the rate of readmissions for SSI was 1.07%. Hospital risk-adjusted odds ratios for readmissions for SSIs varied from 0.40 to 2.80. 4 hospitals performed significantly better than expected (2.8%), and 5 hospitals performed significantly worse than expected (3.5%). Hospital performance regarding readmissions for SSI was not correlated with overall hospital risk-adjusted SSI rates or readmission rates (r=.30 and r=.026, respectively). Risk factors for readmissions for SSIs include class III obesity (HR, 1.95; CI, 1.3 – 2.93; P < .01), ASA class III (HR, 1.79, CI, 1.32-2.3, P<.001), smoking status (HR 1.32; CI, 1.01 – 1.71, P = .04), current steroid use (HR 1.58; CI, 1.09 – 2.28, P = .02), operative times over 100 minutes (P < .001), and open surgery (P<.01). There was a non-significant trend for larger hospitals to have higher readmission for SSI rates (P=.18). A resident-to-bed ratio under 0.3 was also a risk factor for readmission for SSI (P<.01) compared to higher ratios.

Conclusions: Readmission for SSI rates represents a unique aspect of quality beyond that offered by just SSI and overall readmission rates alone. The rate of unplanned readmissions for SSI is a novel quality indicator that may provide actionable quality improvement information for hospitals.

52.09 Patient Preparation for Transitions of Surgical Care: Failing to Prepare is Preparing to Fail

Posted on January 18, 2016

L. Martin1, S. Finlayson1, B. Brooke1 1University Of Utah,Surgery,Salt Lake City, UT, USA

Introduction:

Transitions of care before and after surgery are a critical time for patients to gain essential information to ensure they understand surgical care plans, avoid adverse events and receive the best outcomes. This study was designed to evaluate how patients prepare for transitions of surgical care, and to explore the association between utilization of health information resources with readmission.

Methods:

A cross-sectional survey was conducted in March 2015 among nationwide members of Health Alliance web-based social communities who reported having direct experience with surgery as either a patient or caregiver. Respondents were asked to report on how prepared they were for the transition of care before and after their most recent surgery, what health information resources they used to get prepared, and whether they required readmission within 30-days following surgery. Survey results were analyzed using bivariate methods.

Results:

Of 1872 surveyed individuals, 93% were patients and 79% had undergone surgery within the past year. Respondents’ exposure to surgery represented a broad spectrum of ten major disciplines with Orthopedic (28%) and General Surgery (14%) being most common. The majority of respondents felt very prepared prior to undergoing surgery (64%) and before leaving the hospital after surgery (65%). As compared to patients who reported being unprepared, those who felt prepared for transitions of surgical care were most likely to be given multiple health information resources before surgery (97% vs. 79%, p=<0.001) and before leaving the hospital (91% vs. 85%, p=0.02), which included face-to-face meetings, written instructions, internet sites, videos, and smartphone applications. Feeling prepared was significantly associated with the number of resources that patients were provided by their surgical team and used before surgery (Figure 1) and before leaving the hospital. 30-Day readmission was significantly lower among patients who felt prepared either before (8% vs. 23%; p=<0.001) or after surgery (9% vs. 23%, p=<0.001). Furthermore, use of any post-operative resource was associated with a significantly lower 30-day readmission (10% vs 31%, p=<0.001) as compared to patients who reported no health information utilization.

Conclusion:

Patients who report using more health information resources before and after surgery felt more prepared for transitions of surgical care and have a lower rate of 30-day readmission. Ensuring that patients have access to multiple sources of health information and feel prepared for transitions of surgical care is critical to patient satisfaction and obtaining the best clinical outcomes.

52.10 Systematic Review on the Effect of Health Information Technology in Surgical Patient Care

Posted on January 18, 2016

J. Robinson1,2, H. Huth4, G. P. Jackson2,3 1Vanderbilt University Medical Center,Department Of General Surgery,Nashville, TN, USA 2Vanderbilt University Medical Center,Department Of Biomedical Informatics,Nashville, TN, USA 3Vanderbilt University Medical Center,Department Of Pediatric Surgery,Nashville, TN, USA 4Vanderbilt University,Nashville, TENNESSEE, USA

Introduction: Government regulations, such as ‘meaningful use’, and consumer demand have driven widespread implementation and adoption of health information technologies. The emergence of electronic health records (EHRs), computerized order entry (CPOE), and patient portals has transformed the way health information is stored, used, and communicated to patients. The effect of such technologies on surgical practice has not been well studied. Our objective is to systematically review the available evidence regarding the impact of health information technology (HIT) on the care of surgical patients.

Methods: A systematic search of Medline, EMBASE, CINAHL, and Cochrane Library to identify data-driven, non-survey, studies on the effect of HIT, EHR, CPOE or patient portals on surgeons and their patients was performed. Experts in clinical informatics were also queried for known relevant articles to compile for review. Two authors independently reviewed abstracts to select relevant publications that met the inclusion criteria, followed by a full-text review of relevant articles, according to the PRISMA guidelines.

Results: A total of 2909 citations from 4 databases were identified, with 2879 publications remaining after duplicates were removed and the addition of expert-recommended articles. 190 articles were retrieved for full-text analysis, and 23 articles, including 21 observational studies and 2 RCTs, were found applicable. EHR and CPOE systems improved appropriate and timely antibiotic administration for surgical procedures in 6 observational studies. 5 observational studies indicated electronically generated operative notes have increased accuracy, completeness, and timely availability in the medical record compared to dictated notes. The Internet as an informative resource was inadequate for surgical procedures in 2 observational studies; however, a tailored web-based program decreased perioperative anxiety of patients in a single small-scale RCT. In observational studies, patient web portals showed high adoption and utilization, encompassing an increasing proportion of outpatient surgical encounters.

Conclusion: Overall, the quality of evidence on HIT in surgical practice found was moderate to low with only 2 RCTs identified. Currently available data suggest EHR, CPOE, and electronic notes are effective at improving the accuracy of drug administration and operative reports, but few other effects are known. While patient portals are in use, data is lacking on the quality or effect of portal encounters, and the Internet alone has not proven to be a reliable information source. With the increasing use of HIT in surgical patient care, further evidence-based research is required to optimize its utilization and efficacy.

53.01 Relationship of Postoperative Surgery Clinic Visits and Readmission after Gastrointestinal Surgery

Posted on January 18, 2016

R. H. Hollis1,2, L. A. Graham1,2, J. S. Richman1,2, M. S. Morris1,2, M. T. Hawn3 1University Of Alabama At Birmingham,Dept. Of Surgery,Birmingham, AL, USA 2Birmingham VA Medical Center,Center For Surgical, Medical Acute Care Research And Transitions (C-SMART),Birmingham, AL, USA 3Stanford School Of Medicine,Dept. Of Surgery,Stanford, CA, USA

Introduction: Though postoperative surgery clinic visits are routinely performed, little is known about the relationship between surgery clinic visits and hospital readmissions. We hypothesized that timing of post-discharge surgery clinic visits is associated with readmission risk following gastrointestinal surgery.

Methods: Using VASQIP data we identified patients discharged following gastrointestinal surgery with at least a two-day postoperative length of stay in VA facilities between years 2008-2014. Our independent variable of interest was the occurrence of a postoperative general surgery clinic visit identified in the VA Corporate Data Warehouse (CDW). Our outcomes were 10-day and 30-day hospital readmission rates. To evaluate the association of clinic visit and hospital readmission, we used cox-proportional hazard models with general surgery clinic visit as a time-varying covariate and controlled for patient, procedure, and inpatient stay factors associated with readmission. To assess whether timing of clinic visits was associated with readmission, we performed a facility level analysis and tested the association between facility rate of readmission and mean time from discharge to initial clinic visit using Pearson correlation. Facilities performing less than twenty surgeries were excluded.

Results: Of 85,219 patients discharged after undergoing inpatient gastrointestinal surgery, hospital readmissions occurred in 6218 (7.3%) patients within 10-days and in 10,451 (12.3%) within-30-days. A general surgery visit was recorded in 28,933 (33.9%) patients within 10 days following discharge and in 59,279 (69.5%) within 30-days. Overall, the median time to readmission was 8 days (IQR: 4-16), and the median time to clinic visit was 11 days (IQR 7-15). Only 2,810 (26.9%) of patients readmitted within 30-days had a preceding general surgery clinic visit. Having a postoperative general surgery clinic visit was associated with a significantly decreased hazard of readmission within 10 days (HR 0.64, 95%CI 0.59-0.70) and 30-days (HR 0.63, 95%CI 0.60-0.66) of discharge. In examination of 114 facilities, longer mean time to initial postoperative clinic visit was associated with a higher facility rate of 10-day (r=0.24, p=0.01) and 30-day readmission (r=0.30, p<0.01).

Conclusion: Following discharge from inpatient gastrointestinal surgery, postoperative general surgery clinic visits were associated with decreased hazard of 10-day and 30-day hospital readmissions. Facilities with longer time to postoperative clinic visits were associated with higher readmission rates. These findings highlight early postoperative surgical clinic visits as a potential important focus for reducing readmission rates.

52.05 The Gastrostomy Tube Consultation: An Opportunity for Palliative Care Assessment and Intervention

Posted on January 18, 2016

C. M. McGreevy1, S. Pentakota1, A. Kunac1, O. Mohamed1, K. Sigler1, A. C. Mosenthal1, A. Berlin1 1Rutgers-New Jersey Medical School,Surgery Department,Newark, NJ, USA

Introduction: General surgeons are frequently consulted for placement of a gastrostomy tube, and patients requiring feeding access are often seriously ill. Current guidelines for quality palliative care recommend that all patients with a potentially life-limiting illness receive a palliative care assessment when feeding tube placement is considered. We aimed to characterize the extent of unmet palliative care need in patients receiving gastrostomy tubes by examining palliative care processes and outcomes in this population.

Methods: This is a retrospective study of all adult non-trauma inpatients who underwent gastrostomy tube placement in the year 2013. Patients were identified based on procedure codes. We abstracted data regarding demographics, diagnosis, indications, palliative care processes, and outcomes via chart review. The primary outcome was receipt of palliative care assessment prior to tube placement. Secondary outcomes included functional status at discharge as measured by Glasgow Outcome Scale (GOS) or Modified Rankin Scale (MRS), as well as in-hospital and 6-month mortality. We used counts and proportions to describe study variables and multivariable logistic regression to identify factors associated with receipt of palliative care.

Results: One hundred twenty-eight patients met inclusion criteria. All but 3 had a serious or life-limiting illness. Of the remaining 125 patients, head and neck malignancy (37%) was the leading indication, followed by acute cerebrovascular accident (27%), prolonged respiratory failure (15%), and other neurologic disorders (14%). Only 14% of patients in whom a tube was placed had a palliative care assessment prior to the procedure. Only indication and race/ethnicity were statistically significantly associated with this care pattern. No head and neck malignancy patients received a palliative care assessment, and non-black patients were much less likely to receive palliative care assessment prior to gastrostomy placement (OR: 0.28 (0.11-0.69)). 14% of patients required tube change due to a complication within 1 year of placement. In-hospital and 6-month mortality were 6% and 16%, respectively. 62% of survivors to discharge suffered from significant functional disability defined as a GOS of ≤3 or a MRS of ≥4: unable to walk or attend to bodily needs without assistance.

Conclusion: Despite expert consensus guidelines, the majority of patients with serious or life-limiting illness did not receive palliative care assessment prior to placement of a gastrostomy tube. While considered routine procedures, patients requiring feeding access have high mortality rates and poor functional outcomes. This suggests that consultation for feeding tube placement is an appropriate trigger for palliative care assessment and intervention to ensure treatment is aligned with patient preferences. Surgeons can promote high-quality, patient-centered care by taking an active role in this process.

52.06 Enhanced Recovery After Surgery Programs Improve Patient Outcomes and Recovery: A Meta-Analysis

Posted on January 18, 2016

C. S. Lau1,3, R. S. Chamberlain1,2,3 1Saint Barnabas Medical Center,Surgery,Livingston, NJ, USA 2New Jersey Medical School,Surgery,Newark, NJ, USA 3St. George’s University School Of Medicine,St. George’s, St. George’s, Grenada

Introduction: Enhanced recovery after surgery (ERAS) programs have been developed with the aim to improve patient outcomes and accelerate recovery after surgery. ERAS programs are a multimodal approach, with interventions during all stages of care: preoperative, intraoperative, and postoperative. ERAS programs have been proposed to improve patient outcomes and reduce health care costs. This meta-analysis examines the impact of ERAS programs on patient outcomes and recovery.

Methods: A comprehensive literature search of all published randomized control trials (RCTs) assessing the use of ERAS programs in surgical patients was conducted using PubMed, Cochrane Central Registry of Controlled Trials, and Google Scholar (1966-2015). Keywords searched included ‘enhanced recovery’ and ‘fast track’. Studies using at least 4 components of the ERAS program were included. Primary outcomes analyzed were length of stay (LOS), overall mortality, readmission within 30 days, and total costs. Total complications, time to first flatus, and time to first bowel movement were also analyzed.

Results: 42 RCTs involving 5,241 patients (2,595 receiving ERAS and 2,646 receiving standard of care) were analyzed. ERAS programs significantly reduced LOS by 2.35 days (MD = -2.345; 95%CI, -2.733 to -1.958; p<0.001), total complications by 38.0% (RR=0.620; 95%CI 0.545 – 0.704; p<0.001), and total costs (SMD= -0.789; 95%CI, -1.093 to -0.485; p<0.001). LOS reductions varied by type of surgery, with a 3 day reduction after orthopedic surgery (p=0.017) and no significant reduction after cardiovascular surgery (p=0.073). Return of gastrointestinal (GI) function was also significantly improved, as measured by earlier time to first flatus (SMD= -0.987; 95%CI, -1.389 to -0.585, p<0.001) and time to first bowel movement (SMD= -1.074; 95%CI, -1.396 to -0.752; p<0.001). Overall mortality was reduced by 29.2% (RR=0.708; 95%CI 0.377 – 1.330; p=0.283). Overall, there was no difference in readmission rates within 30 days (RR=1.151; 95%CI 0.822–1.612, p=0.412); however, readmission rates within 30 days after upper GI surgeries nearly doubled with the use of ERAS programs (RR=1.922; 95%CI 1.111 – 3.324; p=0.019).

Conclusion: ERAS programs are associated with a significant reduction in LOS, total complications, total costs, as well as earlier return of GI function. Overall mortality rates remained similar, but readmission rates varied significantly depending on the type of surgery. ERAS programs are effective and a valuable part in improving patient outcomes and accelerating recovery after surgery. Additional studies are required to determine the specific components of the ERAS program that are most beneficial.

52.07 Discrete Choice Survey App for Individualized Peripheral Arterial Disease Treatment Selection

Posted on January 18, 2016

M. A. Corriere1, R. Barnard1, S. Saldana1, R. J. Guzman3, D. Easterling1, D. Boone2, A. Hyde2, G. L. Burke1, E. Ip1 1Wake Forest University School Of Medicine,Winston-Salem, NC, USA 2Wake Forest University Schools Of Business,Winston-Salem, NC, USA 3Beth Israel Deaconess Medical Center,Vascular Surgery,Boston, MA, USA

Introduction: Multiple treatments are available for peripheral arterial disease (PAD), but an evidence-based ‘best’ choice is often unclear. Providers lack efficient methods for identifying patients’ goals and values, and treatment selection is often based on ambiguous criteria with limited patient input. We developed and pilot tested an app to help providers understand patients’ treatment preferences, facilitating shared decision making by identifying compatibile treatment choices.

Methods: Focus groups were used to explore themes, terminology, and treatment attributes patients with severe PAD (rest pain and/or tissue loss) consider important. Transcripts were analyzed using thematic content analysis. An iPad app was developed to characterize how individuals prioritize key treatment attributes, including: treatment type (i.e., medication, percutaneous procedure, or surgery), anticipated level of symptomatic improvement, odds of success versus failure, risk, and durability. These attributes were presented using 14 randomly ordered pair-wise choice tasks with varying levels for each attribute based on an orthogonal array design. Responses were used to determine part-worth utilities and generate patient-specific scores indicating the relative importance of each attribute.

Results: 26 patients with severe PAD participated in focus groups, and a separate group of 34 participants completed the discrete choice survey in clinic. Mean choice task completion time was 12.8±6.0 minutes. Aggregate importance scores indicated that treatment type was most influential (41%), followed by chance of successful treatment outcome (24%), risk (18%), durability (10%), and level of symptomatic response (7%). Analysis on a per-patient level, however, demonstrated that attributes are prioritized differently between individuals (Figure), often supporting individualized choices based on the patient’s personal goals and values.

Conclusion: Patient treatment preferences are based on complex values systems that differ between individuals with similar PAD severity. Choice-based survey tools can be used to evaluate patient preferences in an objective and quantitative fashion, allowing providers to access information that is otherwise difficult and time-intensive to obtain during a standard clinic visit. Understanding patient’s goals and values has potential to facilitate individualized treatment choices based on the patient’s unique priorities, goals, and values.

52.02 Post-Operative Outcomes Following Elective Colorectal Surgery in the Obese Population

Posted on January 18, 2016

F. C. Patel1, A. Gullick1, A. DeRussy1, D. I. Chu1, J. Grams1, M. Morris1 1University Of Alabama Birmingham,Department Of Surgery,Birmingham, Alabama, USA

Introduction: Obesity remains a growing epidemic in the United States. Studies have suggested that obesity may worsen post-operative outcomes such as surgical site infection (SSI), but many of these studies categorized patients only as obese or non-obese. By further stratifying the obese population, we aim to investigate the role of obesity classes in determining post-operative outcomes for patients undergoing elective colorectal surgery.

Methods: Patients who underwent elective colorectal surgery were queried from the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) 2011-2013 cohort and stratified by body mass index category into underweight (<18.5 kg/m2), normal weight (18.5-24.9 kg/m2), overweight (25-29.9 kg/m2), and class I (30-34.9 kg/m2), II (35-39.9 kg/m2), and III (>40 kg/m2) obesity. Univariate and bivariate comparisons were made with Chi-square and Wilcoxon Rank Sums tests to determine differences among categorical and continuous variables, respectively. Logistic regression analyses identified risk factors for 30-day mortality, 30-day readmission, and SSI (defined as superficial/deep SSI or wound disruption but excluding organ space infection).

Results: Of 74,891 patients who underwent elective colorectal surgery, 3,265 (4.4%) were underweight, 21,685 (29%) normal weight, 24,705 (33%) overweight, 14,797 (19.8%) class I obesity, 6,324 (8.4%) class II obesity, and 4,115 (5.5%) class III obesity. Comorbidities included non-insulin-dependent diabetes (10%), smoking (17.5%), and hypertension (49.2%). SSI rates in the overall cohort were 8.7% and ranged from the highest in class III obesity to the lowest in normal weight patients (15% vs. 6.5%, p<0.001). Fully adjusted modelling showed an increased risk of post-operative SSI with increased obesity class: Overweight (OR 1.34, CI 1.24-1.44), Class I (OR 1.68, CI 1.55-1.82), Class II (OR 2.32 CI 2.10-2.55), and Class III (OR 2.56 CI 2.20-2.74). Underweight patients were at increased risk of 30-day mortality (OR 1.34 CI 1.01-1.79), but obesity did not predict mortality. No weight categories were associated with an increased risk of readmission.

Conclusion: Obesity has a dose dependent association with SSI following elective colorectal surgery, but is not associated with readmission or 30-day mortality. BMI may account for some of the variation in post-operative outcomes such as SSI. In order to improve post-operative outcomes, pre-habilitation including supervised weight loss may play an important role prior to elective surgery.

52.03 Big Data In Surgery: Modeling How Post-Surgical Complications Increase Risk For Further Complications

Posted on January 18, 2016

S. I. Feld1, S. E. Tevis1, A. G. Cobian2, M. W. Craven2, G. D. Kennedy1 1University Of Wisconsin,Department Of Surgery,Madison, WI, USA 2University Of Wisconsin,Department Of Biostatistics And Medical Informatics,Madison, WI, USA

Introduction:

Patients who suffer from post-operative complications have longer hospital stays, higher rates of readmission and mortality, and higher cost of care. Many studies have evaluated predictors of complication development. The goal of this study was to assess the temporal relationships among post-operative complications. Knowledge of these relationships will improve our ability to select targeted interventions to prevent cascades of these complications.

Methods:

The American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) database includes preoperative risk factors, intraoperative variables and 30-day postoperative outcomes for patients who underwent major inpatient and outpatient surgical procedures. This study includes cases from this database from 2005 – 2013. Data included all ACS NSQIP-defined complications within 30 days post operation. Machine learning methods were used to model the temporal dependencies between complications. A Markov chain model was developed to model the development of subsequent complications given knowledge of the complications a patient has experienced.

Results:

The model was best at predicting death, coma longer than a day, cardiac arrest, septic shock, renal failure, pneumonia, unplanned re-intubation, longer than 2 days on a ventilator and bleeding transfusion (greater than 75% sensitivity at the 75% specificity threshold). The risk for later complications depends on the complications a patient has experienced. We found some complications to be more likely to occur in insolation while others are likely to be associated with a second. Figure 1 shows the risk for complications given a prior complication (normalized by the overall likelihood of diagnosing that complication), with red indicating greater risk and blue less risk than baseline. Complications such as cardiac arrest or MI, renal insufficiency or failure, stroke, intubation, septic shock and coma contributed to complication cascades to a much greater extent than other complications. For example, a patient who has a coma has on odds ratio of >2 of dying within 30 days of the operation while the odds ratio for death following a diagnosis of pneumonia is <0.5.

Conclusions:

A Markov Chain Model combining information about prior complications and time to occurrence after surgery can inform the likelihood of future complications. The present study utilized a novel method to determine several associations among post-operative complications which will contribute to our ability to better target interventions for high-risk post-operative patients.

52.04 Barriers to Translational Research: Specimen Attrition in a Prospective Cancer Tissue and Databank

Posted on January 18, 2016

A. Gangi1, W. Sun1, S. Yoder3, M. Fournier2, M. C. Lee1 1Moffitt Cancer Center,Comprehensive Breast Program,Tampa, FL, USA 2Moffitt Cancer Center,Tissue Core Facility,Tampa, FL, USA 3Moffitt Cancer Center,Molecular Genomics Core Facility,Tampa, FL, USA

Introduction: Translational cancer research is increasingly reliant on existing tissue and data banks for retrospective tissue and data studies with potential clinical impact. Despite significant infrastructure and funding, these investigations are hindered by a multitude of unanticipated hurdles.

Methods: This is a funded, prospective/retrospective, IRB-approved, tissue and data study incorporating SNP analysis of breast cancer patients treated at an NCI designated, comprehensive cancer center. All cases and controls had archival tissue and were at least 5 years from their incident breast cancer diagnosis. Large-scale prospective institutional and programmatic databanks, tissue banks, and medical record chart review were leveraged. Tissue specimens were reviewed by a research pathologist for verification and adequacy. Sequencing of 10 selected genes at 30x coverage was targeted. Causes of attrition in the 1:1 matched case-control population were evaluated. Matched pairs were not matched by tissue type.

Results: 2927 stage I-III breast cancer patients were reviewed to identify 130 matched case-control pairs (260 total subjects) with specimens and data. Because of paired matching, elimination of one specimen resulted in the elimination of a pair if a replacement could not be identified. Of 260 specimens (130 pairs), 192 tissue samples (96 matched pairs) were identified (73.8%) over an 18-month time frame. 156 were archived FFPE (slides or blocks) and 36 were frozen (solid tissue, blood, extracted DNA). Thirty-four of the 130 data-matched pairs were unusable (26.1%): in 28 pairs, tissue could not be identified for one of the paired subjects. This was due to: missing blocks/slides, compromised tissue cases, or specimens that did not represent normal tissue due to inflammation or necrosis. In 2 other pairs, adequate tissue for macrodissection could not be identified. Four additional pairs were excluded due to insufficient DNA yields at extraction for one of the paired specimens. Of 192 specimens, 150 (75 pairs) passed the quantifiable DNA quality control studies (Qubit). Of 75 pairs sent for DNA sequencing, 47 pairs (36.2%) had evaluable data. The remaining 28 had insufficient hybridization for 30x coverage.

Conclusion: Despite significant infrastructure and resources, retrospective tissue studies are fraught with evaluable specimen loss at each step of the process. When assessing feasibility of retrospective tissue studies, investigators should consider a significant dropout rate within populations of archival tissue specimens.

51.09 Do Risk Calculators Accurately Predict Surgical Site Infection in Ventral Hernia Repair?

Posted on January 18, 2016

T. O. Mitchell1, J. L. Holihan1, E. P. Askenasy2, J. A. Greenberg3, J. N. Keith4, R. G. Martindale5, J. Roth6, B. E. Henchcliffe1, C. W. Hannon1, J. Mo1, M. K. Liang1 1University Of Texas Health Science Center At Houston,Department Of Surgery,Houston, TX, USA 2Baylor College Of Medicine,Department Of Surgery,Houston, TX, USA 3University Of Wisconsin,Department Of Surgery,Madison, WI, USA 4University Of Iowa,Department Of Surgery,Iowa City, IA, USA 5Oregon Health And Science University,Department Of Surgery,Portland, OR, USA 6University Of Kentucky,Department Of Surgery,Lexington, KY, USA

Introduction: Current risk assessment models for surgical site occurrence (SSO) and surgical site infection (SSI) following ventral hernia repair (VHR) have limited external validation. Our aim was to determine 1) if existing models stratify patients into groups by risk and 2) which model best predicts the rate of SSO and SSI.

Methods: Patients who underwent VHR and were followed for at least one month were included. Using two datasets–a retrospective multicenter database (Ventral Hernia Outcomes Collaborative; VHOC) and a single-center prospective database (Prospective)–each patient was assigned a predicted risk with each of the following models: Ventral Hernia Risk Score (VHRS), Ventral Hernia Working Group (VHWG), Centers for Disease Control (CDC) Wound Class, and Hernia Wound Risk Assessment Tool (HW-RAT). Patients in the Prospective database were also assigned a predicted risk from the American College of Surgeons National Surgical Quality Improvement Project (ACS-NSQIP). Areas under the receiver operating characteristic (ROC) curve (AUC) were compared to assess the predictive accuracy of the models for SSO and SSI. Pearson’s Chi-Square was used to determine which models were able to risk-stratify patients into groups with significantly differing rates of actual SSO and SSI.

Results: The VHOC database (n=795) had an overall SSO and SSI rate of 23% and 17%. The AUCs were low for SSO (0.56, 0.54, 0.52, 0.60) and SSI (0.55, 0.53, 0.50, 0.58). The VHRS (p=0.01) and HW-RAT (p<0.01) significantly stratified patients into tiers for SSO while the VHWG (p<0.05) and HW-RAT (p<0.05) stratified for SSI (Table 1). In the Prospective database (n=88), 14% and 8% developed a SSO and SSI. The AUC were low for SSO (0.63, 0.54, 0.50, 0.57, 0.69) and modest for SSI (0.81, 0.64, 0.55, 0.62, 0.73). The ACS-NSQIP (p<0.01) stratified for SSO while the VHRS (p<0.01) and ACS-NSQIP (p<0.05) stratified for SSI. In both databases VHRS, VHWG, and CDC overestimated risk of SSO and SSI while HW-RAT and ACS-NSQIP underestimated risk for all groups.

Conclusion: All five existing predictive models have limited ability to risk-stratify patients and accurately assess risk of SSO. However, both the VHRS and ACS-NSQIP demonstrate modest success in identifying patients at risk for SSI. Continued model refinement is needed to improve upon the two highest performing models (VHRS and ACS-NSQIP) along with investigation to determine if modifications to perioperative management based on risk stratification can improve outcomes.

51.10 Good Catches and Near Misses: The Hidden Benefit of Surgical Safety Checklists

Posted on January 18, 2016

M. B. Diffley1, L. R. Putnam1,2, A. Hildebrandt1, K. Caldwell1, A. Minzenmayer1, S. Covey1, K. T. Anderson1,2, A. Kawaguchi1,2, D. Pham1,2, L. S. Kao3, K. P. Lally1,2, K. Tsao1,2 1University Of Texas Health Science Center At Houston,Pediatric Surgery,Houston, TX, USA 2Children’s Memorial Hermann Hospital,Pediatric Surgery,Houston, TX, USA 3University Of Texas Health Science Center At Houston,General Surgery,Houston, TX, USA

Introduction: The effectiveness of surgical safety checklists (SSCs) in reducing post-operative morbidity and mortality is difficult to measure when these outcomes are rare, as in pediatric surgery. However, SSCs may have additional benefits. Good catches and near misses are defined as events which can lead to patient harm but are prevented from occurring. We hypothesized that SSCs increase good catches and near misses.

Methods: A direct observational study from May-July 2015 was conducted. During the performance of the 19-point, pre-incision phase of the SSC, five trained observers documented checklist adherence and good catch or near miss events. The events were organized into five categories by a patient safety expert: communication failures, medication issues, equipment issues, process issues, and safety issues. Regression analysis was used to evaluate the association between events and the performance of the checklist (adherence to all checkpoints), surgical specialty and case duration. Inter-rater reliability (kappa) was determined for checklist adherence.

Results:Among 212 cases from 9 pediatric surgical subspecialties, SSCs resulted in detection of at least one event in 37 (17%) cases. The most common events were related to process issues (32.4%); the least common events were due to equipment issues (2.7%, Figure). Median (range) pre-incision checkpoints completion was 18 (7-19). Pediatric cardiovascular surgery cases had the highest event rate of 33%: one good catch or near miss for every 3 cases. Kappa for the pre-incision checklist was 0.70 (95%CI 0.63-0.88). On regression analysis, there were no significant associations between the number of events per checklist and checklist adherence, surgical specialty, or case duration. The median number of events did not vary with case length (<30 min = 1, 31-60 min = 1, 61-120 min = 1, >121 min = 1).

Conclusion: Surgical safety checklists, when performed with high fidelity, can detect good catches and near misses. Evaluation of SSCs often focuses only on morbidity and mortality, while good catches and near misses are not reported. Identification and categorization of these events should be routinely measured since they provide targets for focused quality improvement which may lead to reduced errors and adverse events.

« Previous Page
Next Page »