15.09 Money Well Spent? Cost/Utilization Analysis of Prophylactic IVC Filter in High Risk Trauma Patients

Posted on January 25, 2017

M. N. Carlin1,2, A. Daneshpajouh1,2, J. Catino1, M. Bukur3  1Delray Medical Center,Trauma,Delray, FL, USA 2Larkin Community Hospital,General Surgery,South Miami, FL, USA 3Bellevue Hospital Center,Trauma,New York, NY, USA

Introduction:  Placement of Inferior Vena Cava Filters (IVCF) for venous thromboembolic (VTE) prophylaxis is a common and questionable practice in high risk trauma patients. We sought to examine our utilization of prophylactic IVCF at a Level I trauma center. Our primary endpoints were daily cost of IVCF prophylaxis, time to IVCF, duration between IVCF and chemoprophylaxis, and number of patients needed to treat (NNT) to prevent VTE.

Methods:  This was a retrospective review of high risk trauma patients undergoing prophylactic IVCF over a 5-year period (2010-2014). Demographic, Physiologic, Injury, Procedural and Outcome data were abstracted from the registry. Daily cost of IVCF prophylaxis was obtained by dividing the Medicare cost of IVCF by number of days without chemoprophylaxis. NNT was calculated using VTE risk in trauma patients without IVCF as controls.

Results:  Over the 5-year period 146 patients had a prophylactic IVCF. The mean age was 56.3 years (SD ± 24.2), 67.8% were male, and 76% were Caucasian. The predominant mechanism of injury was falls (45.9%) followed by motor vehicle accidents (20.5%). The Median ISS was 25 (IQR 16-29), with a Head AIS 3 (IQR 3-5). Operative intervention was required in 24.7%, with Orthopedic (25.3%) and Craniotomy (21.9%) procedures being the most common.

Median time to IVCF was 78 hours (IQR 48-144). The most common indication for IVCF was closed head injury (48.6%) followed by spinal injuries (30.8%). 57.5% of the patients received chemoprophylaxis in addition to IVCF, with a median time to administration of 96 hours after IVCF (IQR 24-192). The median cost of IVCF prophylaxis was $759/day (IQR $361-1897) compared to $4.32 for LMWH prophylaxis. During the study period 0.26% of patients had a pulmonary embolism (0 in IVCF group). The estimated NNT to prevent 1 additional PE with IVCF was 379 (95%CI 265-661). 

Conclusion: Prophylactic IVCF placement is a costly practice (170-fold increase over chemoprophylaxis) with relatively low benefit in our investigation. Duration of anticipated time without chemoprophylaxis and appropriate patient selection should be considered prior to routine IVCF placement.
 

15.08 Readmission After COPD Exacerbation (RACE) Scale: Determining 30-Day Readmission Risk

Posted on January 25, 2017

R. S. Chamberlain1,2,3, C. S. Lau1,2, B. L. Siracuse1, R. S. Chamberlain1,2,3  1Saint Barnabas Medical Center,Surgery,Livingston, NJ, USA 2Saint George’s University,Grenada, Grenada, Grenada 3Rutgers University,Surgery,Newark, NEW JERSEY, USA

Background: Chronic obstructive pulmonary disease (COPD) or emphysema affects over 13 million Americans, accounts for over 500,000 hospitalizations annually, and is a frequent co-morbidity in surgical patients, altering or curtailing surgical therapy.  The Hospital Readmission Reduction Program, established by the ACA requires the Centers for Medicare and Medicaid Services to reduce payments to hospitals with excess readmissions for COPD as of 2015. This study sought to develop a predictive readmission nomogram that could identify COPD patients at higher readmission risk, as well as surgical risk, and permit the implementation of readmission risk reduction strategies a priori.

Methods: Demographic and clinical data on 342,907 patients from New York and California (derivation cohort) and 260,553 patients from Washington and Florida (validation cohort) were abstracted from the State Inpatient Database (2006-2011), and the Readmission After Chronic Obstructive Pulmonary Disease Exacerbation (RACE) Scale was developed to predict 30-day readmission risk.

Results: 30-day COPD readmission rates were 7.50% for the derivation cohort and 6.67% for the validation cohort. Factors including age <65 (OR 1.58; 95% CI, 1.50-1.67), male gender (OR 1.15; 95% CI, 1.12-1.18), African American (OR 1.09; 95% CI, 1.05-1.14), 1st income quartile (OR 1.09; 95% CI, 1.05-1.14), Medicare (OR 1.49; 95% CI, 1.41-1.56), Medicaid (OR 1.84; 95% CI, 1.74-1.95), anemia (OR 1.07; 95% CI, 1.03-1.11), congestive heart failure (OR 1.08; 95% CI, 1.05-1.12), depression (OR 1.17; 95% CI, 1.13-1.22), drug abuse (OR 1.16; 95% CI, 1.08-1.23), and psychoses (OR 1.17; 95% CI, 1.12-1.23) were independently associated with increased readmission rates, p<0.01. The RACE Scale was created. When it was applied to the validation cohort, it explained 94% of readmission variability within the cohort.

Conclusions: The RACE Scale reliably predicts an individual patient’s 30-day COPD readmission risk based on specific factors present at initial admission. The RACE Scale is a risk stratification model that can identify high-risk patients and permit implementation of patient-specific readmission-reduction strategies to improve patient care, surgical outcomes, as well as reducing readmissions and healthcare expenditures. 

15.06 Clinical Outcomes of Airlifts versus Ground Transports in Pediatric Trauma Interfacility Transfers

Posted on January 25, 2017

R. T. Russell1, R. L. Griffin2, G. F. Smith1, R. H. Hollis3, I. I. Maizlin1  1University Of Alabama at Birmingham,Pediatric Surgery,Birmingham, Alabama, USA 2University Of Alabama at Birmingham,Epidemiology,Birmingham, Alabama, USA 3University Of Alabama at Birmingham,Surgery,Birmingham, Alabama, USA

Introduction: Airlift transport provides an important service by allowing patients to reach definitive treatment rapidly. However, whether this costly resource in interfacility pediatric trauma transfers is appropriately utilized is unclear. Our objective was to evaluate differences in clinical outcomes of pediatric trauma patients following interfacility transfer by ground versus airlift transportation.

Methods: A matched cohort study was conducted among a population of severely injured pediatric trauma patients transferred from referring facilities, based on the trauma registry at a single pediatric trauma facility. Demographics, injury characteristics, clinical characteristics at arrival, and clinical outcomes were prospectively collected. Each patient transported by air was matched to a patient transported by ground based on age ± 3 years, NISS ± 5, injury mechanism, and distance from the place of injury to the hospital ± 30 miles. Demographic, injury, and clinical characteristics were compared between transport types using McNemar's Q and a paired t-test for categorical and continuous variables, respectively.

Results: A total of 273 air-transported patients were matched to ground-transported patients. Patients transported by air were more likely to be African-American (p=0.0135); in addition, despite matching, patients transported by air were slightly younger (p=0.0057) and had an average distance of 5 miles further from definitive treatment (<0.0001). After matching, air-transported patients had a higher injury severity score (p=0.0003), were more likely to have a Level-1 trauma team activation (p<0.0001), be admitted to the ICU from the ED (p=0.0030), and be intubated at the referring facility (p=0.0004). The only clinical outcome that differed between transport groups was hospital length of stay, which was two days longer on average for the air-transported patients (p=0.0240). However, there was no difference in regards to time to first procedure, ICU days, ventilator support days, and hospital disposition (including death).

Conclusion: Despite differences in demographics and injury characteristics, clinical outcomes for patients transported by air were no different than outcomes for patients transported by ground. Consequently, further investigation is required into more accurately identifying those patients who would benefit most from interfacility air transport.

15.05 A Surgery Scheduling Model to Improve Operating Room Utilization at a Tertiary Care Medical Center

Posted on January 25, 2017

M. Jain1, M. Jafarnia2, N. Nayyar2,3, Y. Wang2, B. L. Gewertz1, R. Jain2,3  1Cedars-Sinai Medical Center,Department Of Surgery,Los Angeles, CA, USA 2University Of Southern California,Department Of Electrical Engineering,Los Angeles, CA, USA 3Vivace Systems,Los Angeles, CA, USA

Introduction:

Operating room time is one of the most expensive resources in the hospital. Effective surgery scheduling is crucial to reducing hospital costs. The greatest challenge in optimal surgery scheduling is the uncertainty in operative duration (ORD). Total ORD varies by procedure, nature of case (emergent vs. elective, primary vs. redo, etc.), surgeon, and a variety of other factors. The use of algorithms and machine learning methods can greatly improve the prediction of ORD and operating room utilization.

Methods:

To predict ORD by procedure and surgeon, we introduced 3 different models that estimate ORD, and then combined them to obtain a Hybrid Model of ORD. In Model 1, the estimate is based upon the historical data for ORD for a particular procedure performed by a particular surgeon. In Model 2, we combine the estimate obtained in Model 1 with each surgeon’s estimate of their own required ORD. Intuitively, Model 2 gives more weight to surgeons with more accurate personal ORD estimates. In Model 3, the estimate is based upon the historical data for ORD for a particular procedure performed by any surgeon. A Hybrid Model then chooses the best ORD predictor for each surgeon-procedure pair. We use machine learning techniques to train the model and derive ORDs for each surgeon and each procedure. We also evaluated the effects of patient age, gender, and BMI on ORD. To create the optimal surgery schedule, we introduced a Mixed Integer Linear Program (MILP) formulation that reduces overtime and idle time cost.

A data set of ORD from a tertiary care medical center between November 2013 and March 2016 was evaluated. Predicted ORD was deemed accurate if the actual ORD was less than 20% above or below the predicted ORD. The number of cases predicted accurately was calculated.

Results:

ORD prediction accuracy increased by 33% with use of the Hybrid Model (60% vs. 45%). Patient demographics such as age, gender, and BMI did not improve ORD prediction in the Hybrid Model. Overtime costs are reduced by nearly 66% with our ORD prediction model. Finally, this model demonstrates the potential to increase the overall case load by nearly 18% with no changes in other performance metrics.

Conclusion:

Algorithms and machine learning methods and can be used to improve surgery scheduling by using historical data to more accurately predict surgeon-specific and procedure-specific ORDs.

15.04 Early Lessons from Ongoing Pilot of Telehealth Postoperative Visits after Routine Surgery

Posted on January 25, 2017

J. M. Soegaard Ballester2,5, C. Neylan3, M. F. Scott1, L. Owei1, R. Rosin5,6,7, C. W. Hanson1,2,4,5, J. B. Morris1  1University Of Pennsylvania,Department Of Surgery/Perelman School Of Medicine,Philadelphia, PA, USA 2University Of Pennsylvania,Perelman School Of Medicine,Philadelphia, PA, USA 3Rutgers University,Robert Wood Johnson Medical School,New Brunswick, NJ, USA 4University Of Pennsylvania,Department Of Anesthesia/Perelman School Of Medicine,Philadelphia, PA, USA 5University Of Pennsylvania,University Of Pennsylvania Health System,Philadelphia, PA, USA 6University Of Pennsylvania,Penn Medicine Center For Health Care Innovation,Philadelphia, PA, USA 7University Of Pennsylvania,Leonard Davis Institute,Philadelphia, PA, USA

Introduction:  Focusing on delivery of high-value care, we describe our implementation of telephone postoperative visits (TPOVs) as alternatives to in-person follow-up after routine, low-risk surgery. Our pilot sought to assess the feasibility of offering TPOVs as well as to evaluate patient satisfaction and clinical outcomes. 

Methods:  Enabling TPOVs required submitting a policy proposal to the Department of Health, evaluating implications for surgical global period reimbursement, and updating provider credentials and privileges. We offered TPOVs to all clinically eligible, in-state patients scheduled for appropriate low-risk surgeries, such as: laparoscopic cholecystectomy (LC), open inguinal hernia repair (IHR), umbilical hernia repair (UHR), and minor ventral hernia repair (mVHR). The attending surgeon determined clinical eligibility by assessing the patient’s surgical plan and global clinical picture. Out-of-state patients were excluded given current regulations governing the practice of medicine across state lines. Within 2 weeks of surgery, an advanced practitioner conducted the TPOV following a structured template addressing all postoperative milestones. Patients were discharged from routine follow-up if both they and the practitioner agreed that recovery was satisfactory. 

Results: Of 64 eligible patients, 62 opted for a TPOV between April and August 2016 (Figure 1), with most citing convenience (56%), travel (31%), and time (18%) as main motivations. The average patient opting in was 54 years old (range 22−90, 34% ≥ 65) and lived 19.4 miles from our clinic (range 0.9−121.0).

Of 24 patients completing TPOVs, all were satisfied with the TPOV as their sole postoperative visit (POV) and 21 were discharged from routine follow-up. Three required additional remote management, including diagnosis and treatment of a UTI, monitoring of a wound hematoma via electronically transmitted pictures, and review of postoperative LFTs. On average, TPOVs lasted 9.13 minutes, compared to the 86.3-minute mean end-to-end in-clinic POV time for this surgeon. Adding round-trip driving times from Google Maps, we estimate each patient saved an average of 159−228 minutes, or 95−96% of the time that they would have spent coming to clinic for their POV. 

Conclusion: We have successfully piloted TPOVs at our urban surgical practice with plans to scale our implementation. Many patients, including those with shorter travel times, find TPOVs more convenient than in-clinic visits. Moreover, estimates of time saved are compelling. Telehealth is a rapidly evolving delivery paradigm, with changing regulations and reimbursement. Widespread adoption will be contingent upon ensuring equivalent—or improved—outcomes.

15.03 MRI OVERUSE? A CRITICAL APPRAISAL OF A STATEWIDE TRAUMA SYSTEM CERVICAL SPINE MANAGEMENT

Posted on January 25, 2017

L. Kuo1, B. Frank1, C. Sharoky1, D. Holena1  1University Of Pennsylvania,SURGERY,Philadelphia, PA, USA

Introduction:  Diagnosis of cervical spine (c-spine) injuries requires a balance between resource use and diagnostic efficacy.  In 2009, EAST published c-spine management guidelines but the impact remains of these recommendations is unknown.  We hypothesized MRI use varies between institutions but that higher utilization does not correlate with higher rates of c-spine injury diagnosis. We also sought to characterize variability in MRI charges between institutions.

Methods:  We performed a retrospective review of blunt injury patients aged 18-65 years at level I & II trauma centers in Pennsylvania from 2011-2014. Transferred patients were excluded. Center-level variation in cervical spine MRI use in patients with c-spine injury was calculated, as was cost difference between high and low utilization centers. MRI and c-spine injuries were derived from ICD-9-CM codes and stratified by Glascow Coma Score (GSC) level (high, 14-15; low 3-13).  Correlation between center rates of MRI use and c-spine injury. Annual c-spine MRI charge per center was calculated using average prices from 9 different MRI and CT centers around the country.

Results: 36,316 patients at 26 centers were included (median age 45 (IQR 29-55), 81% white, 67% male). 8,708 (23%) had c-spine injuries. Mean center cervical MRI rate was 12.8% (range: 1.2% – 26.9% )(Figure). In patients with c-spine injuries, MRI rates varied from 4.1% to 82.3%. In the high GCS subset, rates ranged from 1.5% to 26.7%. with no relation between MRI rates and c-spine injury (rho = 0.10, p = 0.61). Estimating an average charge of $1419/MRI, the highest use center had MRI charges of $1920/c-spine injury compared to $69/injury at the lowest use center

Conclusion: Despite 5 year-old guidelines, MRI utilization is highly variable and uncorrelated with rates of c-spine injury.  In the current climate of high quality outcomes at lower cost, efforts to standardize care and reduce cost associated with the diagnosis of c-spine injuries should be pursued.

 

15.02 Fiscal Impact of Delayed Hospital Discharge and Transfer of Care at a Regional Level 1 Trauma Center

Posted on January 25, 2017

B. Ali1, R. Clark1, L. Webb1, S. Cali1, F. Sadiq1, D. Ozathil1, R. McKee1  1University Of New Mexico,Surgery,Albuquerque, NM, USA

Introduction:
Barrier days (BD) or days awaiting placement (DAP) are days spent in the hospital awaiting placement after medical clearance for discharge has been obtained.  Many trauma patients are uninsured or under insured. These days represent lost revenue from new admissions. They also potentially impact patient satisfaction due to long wait times for elective surgery as well as longer stays in the Emergency Room (ER). New Mexico is a resource poor state and in the last fiscal year, the University of New Mexico (UNMH) had 85% or more of its beds occupied over six months’ period in 2015. We sought to evaluate the magnitude of this problem and the cost associated with DAP in order to make a clear financial argument for dedication of resources to improve throughput. 

Methods:
This was a retrospective pilot study. 400 patients admitted to the trauma service over a period of 6 months from July 2012 to December 2012 were reviewed.  Chart review was performed to assess when patients were medically ready for discharge. Reasons for delay in discharge, number of extra days spent in the hospital, disposition and insurance type were recorded.  Cost for the entire length of stay and average cost of each barrier day was calculated. 

Results:

We identified 67 patients (16.5%) with DAP with a mean of 4.3 days and a median of 3 days. The insurance overage of these patients is shown in figure 1. Of 67 patients 35 had delays of less than or equal to 3 days while 31 patients had delays of more than 3 days. DAP represented 8.8% of total hospital charges and the aggregate cost for DAP over this 6 month period was $1.2 million (figure 2). Cost of an average barrier day was calculated to be $4010 (figure 3.) Overall length of stay had a linear relationship with DAP (figure 4). Final disposition can predict DAP as shown in fig 5.

The most common problems that we were able to identify on chart review were failure to involve social workers early in the process, delays in obtaining discharge recommendations from physical and occupational therapists, long waits for approval from receiving facility/insurance authorization, and inefficiency in communication with the primary team (figure 6). 

Conclusion:

Barrier days represent significant cost and inefficiency within our hospital system. Assessing this cost can help guide administrators decisions to dedicate more resources or change processes and weigh the cost of these changes against the cost of barrier days. 

15.01 Understanding the Financial Burden Associated with the Treatment of Colorectal Cancer

Posted on January 25, 2017

F. Gani1, M. Cerullo1, J. K. Canner1, A. E. Harzman2, S. G. Husain2, W. C. Cirocco2, M. W. Arnold2, A. Traugott2, T. M. Pawlik1,2  1Johns Hopkins University School Of Medicine,Department Of Surgery,Baltimore, MD, USA 2Ohio State University,Department Of Surgery,Columbus, OH, USA

Introduction:  Colorectal cancer (CRC) represents the second leading cause of cancer, as well as the second most expensive cancer in the United States. While the implementation of a bundled care payment model has been proposed to curtail the increasing financial burden associated with the treatment of CRC, the distribution and determinants of payments among privately insured patients remain largely unknown. The current study sought to characterize and explain differences in payments received for the treatment of CRC using a cohort of commercially insured patients.

Methods:  Patients >18 years of age, who underwent a colorectal surgery with a primary diagnosis of colon or rectal cancer were identified using the Truven Health MarketScan Database for 2010-2014. Total payments associated with surgery, chemotherapy and / or radiation therapy were calculated. Nonparametric, multivariable linear regression analysis was used to calculate and compare risk-adjusted payments between patients.

Results: A total of 32,782 patients were identified who underwent a colorectal resection for cancer. The median age of the study population was 55 years (IQR: 49-60) with 54.4% (n=17,823) being male. Comorbidities were common as 49.1% (n=16,142) patients presented with preexisting comorbidity (Charlson comorbidity index (CCI)>2). The median risk-adjusted payment for surgery was $27,726 (IQR: $20,099-$40,013), ranging from $17,528 among patients in the lowest quartile of payments to $40,968 among patients included in the highest quartile of payments (+Δ $23,440, p<0.001). Greater preoperative comorbidity (CCI=2 vs. CCI>6: $36,082 vs. $40,944) and the development of a postoperative complication (no complication vs. complication: $35,823 vs. $44,858) were associated with higher payments (both p<0.001). Following surgery, 44.0% of patients received adjuvant therapy. Marked variations in payments received for adjuvant therapy were observed ranging from $366 / cycle of chemotherapy for patients in the lowest quartile of payments to $10,426 / cycle of chemotherapy among patients included in the highest quartile of payments (+Δ $10,060, p<0.001). Among patients included within the lowest quartile of payments chemotherapy accounted for 13.7% of all payments received; in contrast, among patients included in the highest quartile, chemotherapy related payments accounted for 63.4% of all payments received (+Δ 49.7%). Payments received for radiation therapy were comparable among all patients and accounted for <1% of all payments. 

Conclusion: Payments associated with the care of CRC varied significantly despite case-mix and geographical adjustment. Variations in payments were largely due to differences in chemotherapy, and less due to differences in payments for surgery. Episode-based bundle payments for surgery and chemotherapy may differentially impact reimbursement for CRC associated care.

 

14.21 Disclosure of Pre-Referral Medical Errors: Cancer Specialists’ Attitudes and Current Practices

Posted on January 25, 2017

H. Singh1, R. M. Kauffman2, M. C. Lee3, G. P. Quinn4, L. A. Dossett1  1University Of Michigan,Department Of Surgery,Ann Arbor, MI, USA 2Vanderbilt University Medical Center,Department Of Surgery,Nashville, TN, USA 3Moffitt Cancer Center And Research Institute,Comprehensive Breast Program,Tampa, FL, USA 4Moffitt Cancer Center And Research Institute,Department Of Health Outcomes And Behavior,Tampa, FL, USA

Introduction:
Physicians are ethically obligated to disclose their own medical errors to patients. Physician-level training and risk management systems facilitate proper disclosure. No guidelines, education, training, or systems address the disclosure of medical errors discovered by specialists that have occurred prior to consultation – “pre-referral errors.” We sought to describe attitudes and practices regarding disclosure of pre-referral errors discovered across hospital systems.

Methods:
We conducted face-to-face semi-structured interviews with fellowship-trained cancer specialists at multiple NCI-designated cancer centers. Interviews (30-60 minutes in duration) were audiotaped, transcribed verbatim, and independently coded for a priori and emergent themes using the constant comparative method. Open and axial coding were applied using content analysis. 

Results:
Subjects were fellowship-trained specialists of many disciplines (n=30, 40% female, 60% surgeons). The mean age was 46 years; and both the median years of post-graduate training and independent practice were 8. Subjects exhibited a wide range of practice patterns, attitudes/beliefs, and barriers regarding the disclosure of pre-referral errors to their new patients (Table 1). Practice patterns included: no disclosure, vague or limited disclosure, event- or factor-dependent disclosure, and explicit disclosure. Attitudes ranged from no perceived added benefit to disclosing to the thought that the majority of errors are not malicious. A wide range of barriers to disclosing also exist, including: concern for referral base, incomplete information, introducing unnecessary stress for the patient, avoiding a “superior ego” image, and potentially damaging another physician’s reputation and/or livelihood.

Conclusion:
Specialists commonly contend with pre-referral errors but practice patterns, attitudes, beliefs, and barriers vary considerably among specialists. Using these data, there is potential to develop and implement disclosing mechanisms that may help mitigate and overcome identified attitudes, beliefs, and barriers.
 

14.20 Ultrasonographic Detection of Occult Inguinal Hernia

Posted on January 25, 2017

C. Shwaartz1, R. S. Lingnurkar2, B. Cohen1, M. Cohen1, H. K. Rosenberg1, C. M. Divino1  1Icahn School Of Medecine At Mount Sinai,General Surgery,New York, NY, USA 2Central Michigan University College Of Medicine,College Of Medicine,Mount Pleasant, MI, USA

Introduction:

In recent years, ultrasonography has gained popularity as an adjunct to physical examination, replacing the now abandoned contrast herniography to detect occult inguinal hernias. Despite pronounced heterogeneity in reported positive and negative predictive values for this modality, the integration of ultrasound in the diagnostic algorithm for inguinal discomfort has been advocated when physical examination alone is inconclusive. The aim of this study is to confirm this recommendation by assessing the diagnostic value of ultrasonography in detecting occult inguinal hernia, and appraise the limits of its detection rate across discrete populations. 

Methods:

We retrospectively reviewed the demography and the outcome of 137 patients presenting with inguinal discomfort between the years 2013 and 2016 in a single surgeon practice. Inclusion criteria were the following: (1) inconclusive physical examination by a single surgeon, and (2) ultrasound and interpretation by a single radiologist following physical examination. Follow up data were collected by either a clinic visit for inguinal hernia repair soon after inguinal ultrasound, or a telephone survey querying for both inguinal hernia repair during the follow up period, and eventual resolution of symptoms. Demographic factors affecting the accuracy of ultrasonography were analyzed.

Results:

137 patients were included in the study, with a median age of 49 years, of which 45% were females. 26 (19%) were tested positive and 111 (81%) were tested negative for occult inguinal hernia. A total of 18 (13%) patients underwent surgery soon after inguinal ultrasound. Of the remaining 119 patients, 101 (85%) were successfully called for follow up. 37 (31%) patients remained symptomatic on follow up, and 4 (3%) patients had undergone inguinal hernia repair during the follow up period. Positive and negative predictive values (PPV and NPV) for ultrasound in detecting occult inguinal hernia were determined to be 79.17%, and 61.05% respectively. Variance across cohort was noted: ultrasound accuracy was influenced by gender (PPV 90.91% in females vs. 69.2 in males), age (PPV 90.91% below 49 and 71.4 above 49), BMI, prior hernia surgery and comorbidities predisposing to hernia formation (lung disease, constipation, prostatism) (PPV of 85.7% with comorbidity vs. 76.4% without).

Conclusion:

Inguinal ultrasonography has a moderately high positive predictive value, but a relatively low negative predictive value for detecting occult inguinal hernia. Our study uniquely stratified these diagnostic values across discrete populations, revealing particularly high positive predictive values for females, patients aged ≤ 49, and patients with predisposing comorbidities.  These findings suggest that ultrasound may be most effective in detecting occult inguinal hernia in representative patients.

 

14.19 Patient-related and technical factors determining recovery after emergency appendicectomy

Posted on January 25, 2017

S. G. Thrumurthy1, R. Som1  1King’s College Hospital NHS Foundation Trust,Surgery,London, London, United Kingdom

Introduction:
Appendicectomy remains one of the most commonly performed emergency surgical operations, and postoperative recovery is influenced by various patient-related and technical factors. This prospective study aimed to identify how such factors affect the incidence of complications and the extent of symptom resolution after emergency appendicectomy.

Methods:
Patients who underwent emergency appendicectomy over a six month period were contacted by telephone. A standardised questionnaire was used to ascertain the duration of analgesia use, duration before return to normal physical activity, duration before return to work or school, surgical site infection rates, rates of re-presentation to community physicians or the emergency department, and rates of readmission to hospital. Patients were stratified into those who underwent laparoscopic versus open appendicectomy, smokers verses non-smokers, and body mass index (BMI) < 30 versus BMI > 30.

Results:
A total of 145 patients were included. Patients undergoing open surgery (versus laparoscopic surgery) required analgesia for significantly longer periods (22 days v. 6 days, p = 0.017), and a longer recovery period before full return to normal daily activities (48 days v. 17 days, p < 0.0001) and school/work (33 days v. 13 days, p < 0.0001). Compared to non-smokers, smokers required longer a recovery period before returning to school/work (24 days v. 17 days, p = 0.048), had a significantly higher risk of surgical site infection (relative risk [RR] 2.21, p = 0.029), and a higher risk of re-presenting to the emergency department (RR 3.21, p = 0.003) and being re-admitted to hospital within 3 months of surgery (RR 8.36, p = 0.002). Compared to patients with a BMI under 30, those with a BMI over 30 had a longer recovery period before full return to normal daily activities (49 days v. 24 days, p = 0.041) and school/work (26 days v. 17 days, p = 0.016), a higher rate of surgical site infection (RR 2.13, p = 0.044), and a higher risk of re-presenting to the emergency department (RR 3.09, p = 0.005) and being re-admitted to hospital within 3 months of surgery (RR 6.0, p = 0.008).

Conclusion:
When possible, the laparoscopic approach to appendicectomy should be adopted over open surgery to improve postoperative recovery. Patients who are smokers or obese (BMI > 30) should be warned of prolonged recovery times, and surgeons must be wary that such patients are at greater risk of surgical site infections and needing emergent or inpatient care for postoperative complications. Such patient groups may benefit from early postoperative outpatient follow-up.
 

14.18 Surgeon Attitudes Towards Prescribing Opioids

Posted on January 25, 2017

M. Alameddine1, O. Brown1, C. Hoban1, R. Kabeer1, H. Paulsen1, J. Silverberg1, B. VanWieren1, J. S. Lee1, M. J. Englebse1  1University Of Michigan,Department Of Surgery,Ann Arbor, MI, USA

Introduction:
Opioid-based pain management is contributing to an epidemic of opioid-related complications. The lack of clear guidelines in the prescription of opioid pharmaceuticals has encouraged inconsistent prescribing habits throughout the United States. Consequently, it is imperative to survey the current practice surrounding opioid-based pain management and identify detrimental trends that can be reversed via reforms in current practice guidelines. Establishing coherent protocols that unify providers in the individualized management of acute surgical pain will help to protect patients from the risk of chronic opioid dependence. The objective of this study is to characterize inconsistencies in perioperative pain management and thereby identify interventions that can most effectively encourage appropriate opioid prescribing practices. 

Methods:
A 26-question survey was constructed, piloted, and finalized through a collaboration of faculty input and student-led background research. The survey evaluated providers’ knowledge of opioid pharmacology, assessed their theoretical perioperative pain management practice in case scenarios, and surveyed their provider-specific approach to opioid prescriptive practices. All opioid-prescribing health professionals at the University of Michigan Health System were invited to participate in the study. The survey was hosted in Qualtrics and distributed via a standard email invitation and link.

Results:
Of the 66 health professionals that responded to the survey (n=201), 90% do not use a risk assessment questionnaire and 89% either rarely use, never use or are not familiar with the Michigan Automated Prescription System (MAPS). 70% of providers were confident answering questions on opioid prescribing practices in a case scenario describing an uncomplicated hernia repair, and of these 46 providers, 15 (33%) described that they would prescribe less than 30 pills for postoperative pain management, while 31 (67%) would prescribe greater than 30 pills. We then compared these subgroups using an unpaired T-test and both discussed pain management prior to surgery with, on average, only 25-50% of their patients.

Conclusion:
Despite clear evidence of an opioid epidemic in this country, our study demonstrates that health professionals are not fully utilizing available resources and prescribing practice strategies that may decrease the risk of opioid-related complications. We also saw that for the same surgical case, there is variation in the number of opioid pills that physicians would prescribe for patients post-operatively. Moving forward, the goal is to further investigate why current prescribing practices exist and try to identify potential areas for intervention and improvement. Specifically, encouraging the use of standardized prescription guidelines, risk assessment questionnaires, MAPS, and longitudinal discussions regarding postoperative pain management with patients could help to mitigate the current opioid epidemic.
 

14.17 Predicting Outcomes in Hospitalized Patients Requiring Emergency General Surgery

Posted on January 25, 2017

C. E. Sharoky1, E. A. Bailey1, M. M. Sellers1, A. J. Sinnamon1, C. Wirtalla1, D. N. Holena1, R. R. Kelz1  1Hospital Of The University Of Pennsylvania,Center For Surgery And Health Economics, Department Of Surgery,Philadelphia, PA, USA

Introduction: Acute care surgeons are charged with caring for a heterogeneous population, including patients who become acutely ill while hospitalized. Decision-making regarding these patients is often complicated, yet the majority of emergency general surgery (EGS) research has focused on the population operated on within the first two days of hospitalization. We examined outcomes of patients who had EGS at least three days after admission in order to identify preoperative and operative factors that predict mortality and postoperative length of stay >30d (LOS30) in this high risk cohort.

 

Methods: Patients >18y who had one of seven most common EGS operations (appendectomy, partial colectomy, small bowel resection, operative management of peptic ulcer disease, cholecystectomy, lysis of adhesions, exploratory laparotomy) after hospital day two on an emergent basis were identified in the ACS NSQIP registry (2011-2014). Exploratory laparotomy with no secondary procedure code was presumed non-therapeutic (NTEL). Descriptive statistics were performed. Multivariable logistic regression was used to identify predictors of 30 day mortality or LOS30 in independent models.

 

Results: Of 10,674 EGS patients who met inclusion criteria, the median age was 66 (IQR: 53-77) years. The majority of patients were white (65.5%), functionally independent (86.6%), and admitted from home (81.6%). Sepsis was the most common surgical indication (n=4,295; 40.2%). Median postoperative LOS was 8d (IQR 4-14d), and 719 (7.5%) patients experienced LOS30. Thirteen percent (n=1,424) of patients died within 30 days of operation (median time to death: 8d IQR 2-16d). Of those who died, 742 (52.1%) had a partial colectomy, 290 (20.4%) had a small bowel resection, and 600 (42.1%) failed extubation within 48 hours. Of patients who had NTEL, 171 (41.0%) died within 30 days. Factors most significantly associated with death were ASA class, age and procedure type. NTEL was the greatest operative predictor of death (OR 6.9 p<0.001). Factors most significantly associated with LOS30 were failure to extubate, ASA class and procedure type. Compared to patients whose index operation occurred during week 1, odds of death increased for each subsequent week prior to surgery (week 6 OR 2.6; CI: 1.8-3.9). Odds of LOS30 also increased weekly (week 6 OR 5.5 CI: 3.3-8.3). NTEL after week 3 had ≥50% risk of LOS30.

 

Conclusion: An important subset of patients require EGS after hospitalization. Although these patients are functionally independent on admission, mortality in this cohort is even higher than currently reported in published EGS literature. Risk of death and LOS30 increase as time from hospitalization to operation increases. Those who have NTEL are at particularly high risk of death or LOS30 following surgery. Patient and societal benefit versus risk of surgery in this cohort is complex and demands more attention from the research community.

14.16 Impact of Narcotic Analgesic Use on HIDA Scan-based Gallbladder Ejection Fractions

Posted on January 25, 2017

E. Wiesner1,2, L. Martin1,2, W. Peche1,2, J. Langell1,2  1VA Salt Lake City Health Care System,Center Of Innovation,Salt Lake City, UT, USA 2University Of Utah,Department Of Surgery,Salt Lake City, UT, USA

Introduction:  The use of gallbladder ejection fraction obtained by hepatobiliary iminodiacetic acid scan to diagnose Biliary Dyskinesia (BD) continues to be controversial. Cholescintigraphy or hepatobiliary iminodiacetic acid (HIDA) scan is a nuclear medicine imaging study that allows for the calculation of gallbladder ejection fraction (GBEF).  Patients with a clinical presentation consistent with BD and a low GBEF (<35%) are considered appropriate candidates for therapeutic cholecystectomy.  False positive HIDA-based GBEFs have been associated with the concomitant use of narcotic medications.  Narcotic analgesics are thought to reduce biliary smooth muscle motility and are typically therefore discontinued prior to conducting the HIDA scan. In this study we looked at the impact of narcotic use on HIDA scan-based GBEF in patients with suspected BD.

Methods:  We queried the Veterans Healthcare Administration National Corporate Data Warehouse from January 2005 to July 2016 for patients who underwent more than one HIDA scan. Patients undergoing HIDA for a suspected diagnosis of BD were included. Radiology reports were reviewed and the GBEF for each study was abstracted. We further categorized patients with abnormal GBEFs into those receiving concomitant narcotic analgesics during their initial HIDA scan and on subsequent HIDA scan.  A comparison was conducted to determine the impact of narcotic use on the reported GBEF in these populations.

Results: We identified 546 patients who underwent more than one HIDA scan for suspected BD during the study period.  Thirty-three percent (181) of all patients had an abnormal GBEF (average GBEF=17%) on their initial study.  Of these, 34 patients (19%) were on narcotic analgesics at the time of their initial HIDA scan (average GBEF=16%).  Of the 181 patients with a low GBEF, 45% were found to have a normal GBEF on repeat scan (average GBEF=41% and average time between studies 26.5 months), where as 100% of patients on narcotic analgesics demonstrated a normal GBEF on subsequent HIDA scan (average GBEF=74% and average time between studies 29.9 months) (p-value=0.005).  This finding was independent of continued narcotic use (26% of patients).

Conclusion: In this study, all patients who met diagnostic criteria for BD based on a low HIDA scan-based GBEF and were on narcotic analgesics at the time of the initial HIDA scan demonstrated a normal GBEF on subsequent scan.   This finding was independent of chronic long-term narcotic analgesic use.  Of the 26% of patients who remained on narcotic analgesics at the time of repeat scan, 100% were found to have a normal GBEF.  Although this study supports the discontinuation of narcotic analgesics prior to conducting a HIDA scan when possible, it also suggests that there may still be utility in conducting the repeat scan when patients who are chronic users of narcotics are unable to come off these medications. Further studies will need to be conducted to confirm these findings and determine if the effect of narcotic analgesics on biliary smooth muscle motility is lost with chronic narcotic exposure. 

 

14.15 ERAS: Eliminating the Length of Stay Differences between Open and Laparoscopic Colorectal Surgery

Posted on January 25, 2017

W. J. Farrington1, A. Gullick1, T. S. Wahl1, L. Goss1, M. Morris1, J. Cannon1, G. Kennedy1, D. I. Chu1  1University Of Alabama At Birmingham Medical Center,General Surgery,Birmingham, AL, USA

Introduction: The laparoscopic approach to colorectal surgery has several advantages compared to traditional open surgery including reduced length-of-stay (LOS). While the effectiveness of Enhanced Recovery After Surgery (ERAS) on reducing LOS is well-documented, it remains unclear whether ERAS equilibrates the LOS differences between open and laparoscopic surgery. We hypothesized that ERAS would reduce the LOS for both open and laparoscopic surgery and eliminate these differences.

Methods: A single-institution retrospective review of patients undergoing both laparoscopic and open colorectal surgery before and after the implementation of ERAS was conducted. Patient and procedure-specific variables were recorded. Primary outcome was post-operative LOS. Univariate and bivariate comparison were made. Chi-square and Wilcoxon Rank Sums tests were used to determine differences among categorical and continuous variables, respectively.

Results: Four hundred and twenty patients were included in this study. The pre-ERAS (n=210) patient groups included laparoscopic (n=68) and open (n=142) surgical approaches for both benign and malignant disease. The post-ERAS (n=210) group included laparoscopic (n=92) and open (n=118) surgeries. Patient gender, race, ASA class, smoking and insurance status did not differ by surgical approach among Pre-ERAS and ERAS patients (p>0.05). However, age, indication for surgery, procedure type, and operative time were significantly different by surgical approach in both pre-ERAS and ERAS groups (p<0.05). Prior to the initiation of ERAS, laparoscopic surgery exhibited an advantage in shorter LOS compared to open surgery (5 v. 6 days, p= 0.049). With ERAS, the LOS advantage of laparoscopic surgery was eliminated and LOS was similar between laparoscopic and open surgeries (4 v. 4 days, p= 0.12) (Figure 1).

Conclusion: ERAS reduces LOS for both laparoscopic and open colorectal surgery. Importantly, the LOS advantage of laparoscopic surgery was eliminated with ERAS. These data suggest that ERAS has positive effects on all approaches to colorectal surgery and should be used widely.

 

14.14 Concurrent PEH/Bariatric Surgery: Improved Outcomes of Sleeve Gastrectomy Compared to Gastric Bypass

Posted on January 25, 2017

A. Shada1, M. Stem2, L. Funk1, D. C. Jackson1, J. Greenberg1, A. Stroud1, A. O. Lidor1  1University Of Wisconsin,General Surgery, School Of Medicine And Public Health,Madison, WI, USA 2Johns Hopkins University School Of Medicine,Surgery,Baltimore, MD, USA

Introduction:
Nearly 200,000 bariatric operations are performed annually in the US. Paraesophageal hernia (PEH) is a relatively rare subset of hiatal hernia, but is associated with morbid obesity and is a fairly common condition that bariatric surgeons encounter. There is no consensus on the management of PEH at the time of bariatric surgery. We sought to examine short term outcomes following concomitant PEH repair at the time of bariatric surgery. We also investigated whether there were differences in 30 day outcomes between those who underwent a PEH repair and either a laparoscopic sleeve gastrectomy (LSG) or gastric roux-en-y bypass (LRGB).

Methods:
Using the American College of Surgeons National Surgical Quality Improvement Program database (2011-2014), patients who underwent bariatric surgery (laparoscopic gastric bypass or laparoscopic sleeve gastrectomy) with or without PEH repair were identified. A propensity score matching analysis was used to compare 30-day outcomes between these two groups. The primary outcome variable was overall morbidity; secondary outcome variables included mortality, readmissions, and reoperations. An additional propensity matched subgroup analysis compared LSG and LRGB in only those patients who received concurrent PEH repair.

Results:

Of the 76,343 bariatric surgery patients included in this study, 7.80%(n=5,958) underwent concurrent PEH repair. The proportion of bariatric cases that involved a concurrent PEH repair increased during the study period (2.14% in 2010 vs. 12.17% in 2014, p<0.001) with rate of concomitant PEH/LSG noted to be over 2.5 times higher than PEH/LRGB in 2014 (8.90% vs. 3.20%). After initial propensity score matching, 5,952 bariatric surgery patients who underwent a PEH repair were matched with 11,904 bariatric surgery patients who did not undergo a PEH repair. There were no significant differences in 30-day outcomes between the cohorts. However, the subgroup analysis demonstrated that among all patients with concurrent PEH repair, LRGB patients experienced greater rates of morbidity (6.20% vs. 2.69%, p<0.001), readmission (6.33% vs. 3.06%, p<0.001), and reoperation (3.00% vs. 1.05%, p<0.001) when compared to LSG patients.

Conclusion:
This study found that paraesophageal hernia repair at the time of bariatric surgery appears to be safe in the short-term and therefore strengthens the argument for a concurrent approach to the morbidly obese patient with PEH. In patients with PEH who are equivalent candidates for gastric bypass or sleeve gastrectomy, a sleeve gastrectomy may be preferable given that it is associated with a lower rate of postoperative morbidity.
 

14.13 Autonomous Detection and Grading of Post-Operative Complications Using Natural Language Processing

Posted on January 25, 2017

L. V. Selby1, W. R. Narain2, A. Russo1, H. McGowan1, V. E. Strong1, P. D. Stetson2  1Memorial Sloan-Kettering Cancer Center,Surgery,New York, NY, USA 2Memorial Sloan-Kettering Cancer Center,Health Informatics,New York, NY, USA

Introduction:  Natural language processing (NLP) is a computer science technique that allows interpretation of narrative text, but is infrequently used to identify surgical complications.  Our institution tracks post-operative complications using both the American College of Surgeons – National Surgical Quality Improvement Program (NSQIP) and our in-house surgical secondary events (SSE) database, which captures and grades complications for all surgical patients, but sub-optimally records lower-grade complications.  We attempted to use NLP to improve the entry of lower extremity deep venous thrombosis (DVT) and pulmonary embolisms (PE) (collectively: venous thromboembolism [VTE]) in the SSE database.

Methods:  In our 2011 – 2014 cohort of NSQIP patients all lower extremity duplex ultrasounds and computerized tomography angiographies (CTA) of the chest performed within 30 days of surgery were divided into training and validation datasets.   These studies were chosen as they represent the most frequent methods of detecting DVT and PE at our institution, and a bag-of-words-approach with a support vector machine (SVM) model was used for training.  Electronic health record data was used to classify the severity of the VTE according to our modification of the Clavien-Dindo classification.  Due to definition differences between NSQIP and the SSE database, we excluded cephalic and portal vein thromboses identified in NSQIP and compared NLP identified VTEs to VTEs identified by both NSQIP and our SSE database, and undertook a blinded review of all instances of discordance.

Results: Of the 10,295 NSQIP patients, 251 were used in our DVT validation cohort (273 total ultrasounds) and 506 in our PE cohort (552 total CTAs).  The SVM DVT model had a sensitivity of 85.1% and a specificity of 94.6%, while the PE model had a sensitivity of 90.0% and a specificity of 98.7% (Table 1).  The majority of discordances were due to identification of a VTE in studies other than duplex ultrasound or CTA of the chest (9/13; 69.2%), studies not in our original NLP dataset. The majority of DVTs (23 patients, 57.5%) and PEs (20 patients, 69.0%) in the validation set were grade 2 on our modified Clavien-Dindo classification, meaning they required administration of therapeutic intravenous or subcutaneous anticoagulation.  

Conclusion: NLP can reliably detect the presence and severity of post-operative lower extremity DVTs and PEs without requiring manual chart review from trained NSQIP surgical case reviewers. We are extending our NLP pilot to real-time identification and grading of all VTEs and to the detection of other post-operative complications, including wound infections.

 

14.12 A Nationwide Comparison of Laparoscopic Versus Open Appendectomy in Geriatric Patients

Posted on January 25, 2017

B. Zangbar1, L. Boudourakis2, V. Roudnitsky2, L. Dresner1  1State University Of New York Downstate Medical Center,General Surgery / Surgery / Medicine,Brooklyn, NY, USA 2Kings County Hospital Center,Trauma and Acute Care Surgery / Surgery / Medicine,Brooklyn, NY, USA

Introduction:  Acute appendicitis in elderly patients carries an increased risk of complications and mortality. The benefits of laparoscopic appendectomy (LA) remain undefined as compared to open appendectomy (OA) in elderly patients, particularly in cases of perforated appendicitis. The aim of our study was to evaluate the outcomes of LA versus OA in perforated and non-perforated appendicitis in elderly.

Methods:  Nationwide Inpatient Sample database was used to evaluate the clinical data of elderly patients (>65 years old) who underwent LA and OA over an 8-year period (2004-2011). Incidental and elective appendectomies were excluded. Univariate and Multivariate analysis was used.

Results: A total of 42,678 Elderly patients underwent urgent appendectomy in the United States during these years. The overall rate of perforated appendicitis was 28.4%, and 44.3% of all cases were performed laparoscopically. 21.4% of cases managed non-operatively. In non-perforated cases, LA was associated with lower overall complication rate (LA: 3.2% vs. OA: 7.9%; p < 0.001), shorter length of hospital stay (LOS, LA: 3.5 vs. OA: 6.7 days; p < 0.001), lower mortality (LA: 0.05% vs. OA: 1.9%; p < 0.001); and lower hospital charges (LA: $35,712 vs. OA: $46,345; p < 0.001) compared to OA. In perforated cases, LA had a lower overall complication rate (LA: 4.9% vs. OA: 10.4%; p < 0.001), shorter LOS (LA: 5.4 vs. OA: 7.7 days; p < 0.001), lower mortality (LA: 1.1% versus OA: 2.3%; p < 0.001), and lower hospital charges (LA: $42,823 versus OA: $51,393; p < 0.001) compared to OA.

Conclusion: LA has clear mortality and morbidity advantage in elderly patients with acute perforated and non-perforated appendicitis, and is associated with shorter hospital stay and lower hospital charges. Given their higher complication tendency and reduced physiological reserve, in elderly patients, laparoscopic appendectomy should be the standard of care.

 

14.11 Readability, Complexity, and Suitability of Online Resources for Mastectomy and Lumpectomy

Posted on January 25, 2017

B. N. Tran1, M. Singh1, B. T. Lee1  1Beth Israel Deaconess Medical Center,Plastic And Reconstructive Surgery/Surgery,Boston, MA, USA

Introduction:  Nearly half of American adults have low or marginal health literacy. This negatively affects patients’ participation, decision-making, satisfaction, and overall outcomes.   Previous studies in this area focus primarily on readability of online health information. This study compares online resources for mastectomy versus lumpectomy using expanded metrics including readability, complexity, and suitability.

Methods:  Ten most popular websites for mastectomy and lumpectomy were identified using the largest Internet engine (Google). Each website was assessed for readability (SMOG, Simple Measure of Gobbledygook), complexity (pMOSE/iKIRSCH), and suitability (SAM, Suitability Assessment of Materials). Scores were analyzed by each website and overall.

Results: Readability analysis showed average reading grade level of 15.38 and 13.80  (p=0.04) for mastectomy and lumpectomy literacy respectively.  Both exceeded the recommended sixth grade level. Complexity analysis via PMOSE/iKIRSCH revealed a mean score of 6.5 for mastectomy corresponding to a “low” complexity  and 8th-12th grade education. Lumpectomy literature had a lower PMOSE/iKIRSCH score of 5.8 corresponding to a “very low” complexity and 4th-8th grade education (p=0.05). Suitability assessment showed mean values of 41% and 46% (p=0.83) for mastectomy and lumpectomy literacy respectively, both are interpreted as “adequate” for the intended audience. Inter-rater agreement for PMOSE/iKIRSCH are 92% (k=0.73, p<0.001), and 96% (k=0.87, p<0.001) for mastectomy and lumpectomy literature respectively. Similarly, inter-rater agreement for suitability analysis are 94% (k-0.84, p<0.001), and 90% (k=0.73, p<0.001).

Conclusion: Online resources for breast cancer overall are above the recommended literacy level. When comparing mastectomy to lumpectomy, online resources for mastectomy have a higher reading grade level and tend to be more complex. 

14.08 The Safety and Feasibility of Early Discharge Following Ileostomy Closure: A NSQIP Analysis

Posted on January 25, 2017

A. I. Elnahas1, F. Quereshy1, R. Kelly2, T. Jackson1, A. Okrainec1, S. Chadi1, E. Le Souder1, U. David3  1Toronto Western Hospital,General Surgery,Toronto, ONTARIO, Canada 2University of Toronto,Surgery,Toronto, ONTARIO, Canada 3Women’s College Hospital,General Surgery,Toronto, ONTARIO, Canada

Introduction: The recent expansion of enhanced recovery programs after surgery has safely permitted early discharge for select patients following routine operations. As a result, more procedures are now being considered appropriate for outpatient surgery. The objective of this study is to determine if early discharge (i.e. less than 24 hours) following ileostomy closure is comparable to standard discharge (i.e. discharge on postoperative day 2 or 3) with respect to 30-day clinical outcomes.

 

Methods: Data was obtained from the American College of Surgeons’ National Surgery Quality Improvement Program participant use file to perform a retrospective cohort analysis. The study population consisted of patients discharged on postoperative day (POD) 0, 1, 2, or 3 who underwent elective ileostomy closure from 2005-2014. Patients were excluded if they had any concurrent procedure(s) or documented complications during admission. The primary outcome was the 30-day adverse event rate and the secondary outcome was the 30-day readmission rate. A multiple logistic regression analysis was performed to determine the adjusted effect of early discharge as well as the predictors of adverse events and readmissions.

 

Results: The study population consisted of 355 and 5798 patients in the early and standard discharge groups, respectively. There were no relevant clinical differences between the two groups. There were 19 (5.4%) 30-day adverse events in the early group and 341 (5.8%) in the standard group. The early group had 17 (4.8%) 30-day readmissions and the standard group had 294 (5.1%). Using a multiple logistic regression, an adjusted odds ratio (OR) estimate for 30-day adverse events and readmissions was determined for early discharge. The adjusted OR for 30-day adverse events was 0.95 (p=0.83) and for 30-day readmissions was 1.01 (p=0.96). Higher body mass index, longer operative time, ASA≥3, chronic steroid use along with a history of bleeding disorder and diabetes were significant predictors for adverse events and readmissions.

 

Conclusion: Using this large national surgical database, select patients discharged within 24 hours of ileostomy closure did not have a significantly higher rate of adverse events or readmissions compared to patients discharged on POD 2 or 3 following uncomplicated surgery. Predictors of adverse events and readmissions can help guide the selection of patients suitable for early discharge.

 

« Previous Page
Next Page »