08.01 Prolonged Hospital Stays for Patients Discharged to Post-Acute Care after Colorectal Cancer Surgery

Posted on January 18, 2016

E. A. Bailey1, G. C. Karakousis1, R. Hoffman1, M. Neuwirth1, R. R. Kelz1 1Hospital Of The University Of Pennsylvania,Department Of Surgery,Philadelphia, PA, USA

Introduction: As payment models move toward bundled reimbursement for an entire episode of care, including services provided after discharge, it is important to reduce excess utilization of costly inpatient care. The aim of this study was to compare inpatient length of stay (LOS) and cost differences for colorectal cancer surgery patients who did not experience complications and were discharged to home versus a post-acute care facility. We also sought to identify risk factors associated with discharge to a facility.

Methods: All patients who underwent surgical resection for colorectal cancer in New York and California in 2009-2010 and were discharged to home or to post-acute care (skilled nursing facility or inpatient rehab) were identified.. Patients were excluded if they had a surgical complication or a LOS greater than the 75th percentile for a given procedure. Median LOS and median cost per hospitalization were calculated. A multivariate logistic regression analysis was performed to assess preoperative risk factors associated with discharge to post-acute care.

Results: Of the 35,807 patients initially identified, 11,946 (33.3%) were excluded for procedure-specific prolonged LOS or surgical complication. 23,861 patients were included in the study. Most patients were discharged to home (n=22,567, 94.6%) while 1,285 (5.4%) patients were discharged to a post-acute care facility. Patients discharged to home had a shorter median LOS (5 days [IQR 4,7]) than those discharged to a facility (7 days [IQR 6,9], p<0.001). Cost per hospitalization differed significantly between the two groups with a median of $16,882 (IQR $11,876, $24,146) per patient visit in the group discharged to home versus $20,705 (IQR $14,372, $28,208) in the group discharged to post-acute care (p<0.001). Factors associated with discharge to post-acute care by multivariate analysis included advanced age, female sex, race, 3 or more co-morbid conditions, presence of metastasis, emergent admission, at least 1 admission in the previous year, and being a Medicare patient (Table 1).

Conclusion: Even in the absence of a surgical complication, LOS following colorectal surgery was 2 days longer and $3,823 more costly for patients discharged to a post-acute care facility. Pre-operative factors such as age, sex, number of medical co-morbidities, and emergent nature of admission were significantly associated with a lower rate of being discharged to home. Early consultation with patient advocates and social services for discharge planning should be initiated for at risk patients to ensure a timely discharge and commencement of the true recovery process.

08.02 Improved Compliance with Foley Care Guidelines Following the Implementation of a Patient Dashboard

Posted on January 18, 2016

M. Scerbo1, J. Holcomb1 1University Of Texas Health Science Center At Houston,Houston, TX, USA

Introduction: Compliance with recommended guidelines is associated with improved patient outcomes, however achieving high levels of compliance has proven difficult in clinical practice. The creation of automated clinical decision support systems (CDSS) to assist with guideline compliance has been met with the criticism due to failure to follow human factors principles. A patient dashboard that displays integrated information automatically extracted from the electronic health record has recently been implemented in our surgical intensive care unit (ICU). The status of compliance with certain bundles of care is displayed in a color-coded fashion (green-good, red-bad); the contribution to the cognitive load of the user and alarm fatigue is minimal as it does not generate any noise, page or text. We hypothesized that the presence of a real-time, color-coded passive display of compliance status would improve compliance rates with Foley care guidelines in our surgical ICU.

Methods: This observational study was conducted in a 23-bed surgical ICU. Compliance with Foley care guidelines were evaluated pre-implementation (2 months) and post-implementation (1 month) of a real-time display of compliance status. A compliance score was calculated by dividing the total number of interventions achieved by the total number of opportunities for interventions and is expressed as a percentage. For this study, the numerator is the total number of nursing shifts in which patients with a Foley received the proper Foley care. The denominator is the total number of nursing shifts in which a patient had a Foley. Foleys that were inserted and discontinued within one 12-hour shift and in and out catheters were excluded from the analysis. Compliance rates were evaluated overall, as well as day/night shifts and weekend/weekday shifts. Categorical variables were analyzed using the Pearson’s χ2 test and with simple logistic regression.

Results: The overall pre-implementation compliance rate was 90% (1631/1810), compared with post-implementation compliance rate of 94% (792/843). The implementation of a passive display of compliance status via the patient dashboard was associated with increased compliance with Foley care guidelines (OR 1.7, 95% CI 1.2-2.4, p = 0.001).

Conclusion: The introduction of a color-coded, consistently present display reporting the compliance status improved compliance rates with Foley care guidelines. This assistance occurred without changing the workflow, requiring additional action from the user or increasing physician alerts, phone calls/pages, or noise. Future analysis will include impact on patient outcomes.

08.03 Unplanned 30-Day Readmissions in Orthopedic Trauma

Posted on January 18, 2016

D. Metcalfe1, O. A. Olufajo1, C. Zogg1, A. J. Rios Diaz1, A. H. Haider1, M. B. Harris2, M. J. Weaver2, A. Salim1 1Harvard Medical School,Center For Surgery And Public Health,Boston, MA, USA 2Brigham & Women’s Hospital,Department Of Orthopedic Surgery,Boston, MA, USA

Introduction:

30-day hospital readmission is used as a quality metric in some pay-for-performance frameworks, such as the CMS Readmissions Reduction Program (CRRP). There are plans to extend the CRRP to selected surgical populations. However, the odds of unplanned readmission have been shown, in some surgical settings, to be associated with lack of insurance and Black race. These characteristics are also associated with greater odds of injury, which raises the possibility that trauma centers will be unfairly penalized by extension of the CRRP to include injured patients.

This study characterized the reasons for, and factors associated with, unplanned 30-day readmission of orthopedic trauma patients. We also sought to understand whether it is sufficient to limit measurement of readmissions to the hospital at which patients were initially treated by exploring the proportion that were readmitted to other hospitals.

Methods:

Hospital admissions for fracture and/or dislocation (ICD-9-CM 800-839) were extracted from the California State Inpatient Database (SID) 2007-2011, which is an all-payer dataset that captures 98% of hospital admissions. Isolated rib, skull, and facial fractures were excluded. Unplanned readmissions to any hospital in California were tracked using a unique statewide identifier. Multivariable logistic and generalized linear regression models were used to identify independent associations with readmission. The covariates within these models were age, sex, race, payer status, admission source, weekend admission, Injury Severity Score (ISS), Charlson Comorbidity Index (CCI), hospital bed size, trauma center designation, and teaching hospital status.

Results:

There were 416,568 orthopedic trauma admissions to 391 different hospitals. The population was predominantly older (mean age 63.9, SD 23.6), white (71.5%), male (59.5%), and funded by public insurance (63.9%). Severely injured patients (≥15) accounted for only 3.3% cases. 27,008 (6.5%) were readmitted within 30 days, 27.6% of which to a different hospital. Factors significantly associated with 30-day readmission were older age (>65 adjusted odds ratio 1.32, 95% CI 1.24-1.41), Black race (aOR 1.18, 1.10-1.26), public insurance (aOR 1.40, 1.27-1.54), greater comorbidity burden (CCI >2 aOR 1.09, 1.77-1.89), and spine fracture (aOR 1.38, 1.31-1.45). Major reasons for readmission included cardiopulmonary disease (25.9%), procedural complications (12.8%), and musculoskeletal problems (8.5%).

Conclusion:

Many orthopedic trauma readmissions are for cardiopulmonary disease and potentially unrelated to the quality of their index hospitalization. Penalties for unplanned readmissions risk unfairly penalizing hospitals that serve disadvantaged communities and treat a high proportion of trauma patients. Future work should aim to determine the proportion of readmissions that are truly avoidable given optimal trauma and medical care.

08.04 Bundled Payments for Acute Care: Potential Savings or Need for Further Reform?

Posted on January 18, 2016

F. Gani1, J. E. Efron1, E. C. Wick1, S. H. Fang1, B. Safar1, J. Hundt1, T. M. Pawlik1 1Johns Hopkins University School Of Medicine,Surgery,Baltimore, MD, USA

Introduction: The Bundled Payments for Care Improvement initiative (BPCI) was proposed by the Centers for Medicare & Medicaid Services (CMS) in an effort to obtain and reward a greater value of care. In contrast to the traditional fee-for-service paradigm, under the BPCI hospitals are reimbursed a DRG-specific bundled payment including physician and inpatient fees for a single episode of care. Still in its infancy, little is known regarding the potential effects of the BPCI on hospital payments and net margins. Therefore, this study aimed to investigate the potential effects of the BPCI on net margins among Medicare patients undergoing colectomy at a tertiary care hospital.

Methods: Medicare enrollees undergoing elective colectomy between 2009 and 2013 were identified using DRG and ICD-9-CM diagnosis codes. Total hospital costs and net payments were inflation adjusted and reported as 2013 dollars. Multivariable linear regression analysis was performed to calculate risk-adjusted, DRG-specific hospital costs and payments for each patient. Net margins were calculated as the difference between total hospital costs and total payments received.

Results: A total of 821 Medicare enrollees underwent an elective colectomy and met inclusion criteria. The median age of patients was 69 years (IQR 65-74) with 48.7% being male and 67.6% presenting with one or more comorbidity. Postoperative complications were observed among 27.5% of patients. The median length of stay (LOS) was 8 days (IQR 5-14) and over a fourth of patients were discharged with additional care (27.5%). The median risk-adjusted cost among all patients was $25,202 (IQR $17,164-$42,161). Risk-adjusted costs were higher among patients who developed a postoperative complication ($43,489 [IQR $30,325-$70,885] vs. $21,179 [IQR $15,368-$32,307], p<0.001) and among patients with an Observed: Expected LOS>1 ($38,361 [IQR $26,187-$58,631] vs. $17,697 [IQR $13,414-$24,129], p<0.001). The median payment under the fee-for-service structure was $29,684 (IQR $19,609-$47,874) resulting in an overall net margin of $4,557 (IQR $1,254-$8,805), with 18.2% of patients contributing to an overall negative margin. In contrast, under the bundled payment paradigm, the net margin per patient decreased to $3,777 (IQR $-12,219-$11,825, p<0.001) with over 41% of patients contributing to a net negative margin (Figure).

Conclusions: Postoperative complications, length of stay and total hospital costs strongly correlated with hospital costs. Payments under the bundled payments system were lower and the proportion of patients contributing to a net negative margin increased. Further study is warranted to define the impact of bundled payments on quality of care, as well as hospital finances.

08.19 The Association Between Pre-Discharge Complications and Readmissions

Posted on January 18, 2016

M. Morris1,3, L. Graham1,3, J. Richman1,3, R. Hollis1,3, C. Jones1,3, M. Hawn2 1University Of Alabama,Surgery,Birmingham, Alabama, USA 2Stanford University,Surgery,Palo Alto, CA, USA 3Birmingham Veterans Affairs Hospital,Surgery,Birmingham, AL, USA

Introduction:
Post-operative hospital readmission rates are now publically reported and targeted for quality improvement measures. Identifying which readmissions are preventable or unavoidable is challenging. The ability to predict readmission rates at the time of discharge would potentially change clinical practice. We hypothesize that patients experiencing a pre-discharge complication would have increased rates of readmission.

Methods:
We examined all gastrointestinal surgery cases at 120 VA facilities from 2008-2014 with a total hospital length of stay of at least 2 days. Our independent variable of interest was the occurrence of any pre-discharge complication as assessed by the VA Surgical Quality Improvement Program. Our outcome of interest was inpatient readmission within 30 days following hospital discharge. Chi-square tests statistics and t-tests were used to examine differences in bivariate statistics and a multivariate logistic model was used to model 30-day readmission in the population.

Results:
Our sample includes 86,820 procedures among 77,451 patients at 120 VA facilities with an overall readmission rate of (12.3%). In our cohort, 7,927 patients (9.1%) experienced a pre-discharge complication. Superficial wound infection was the most common pre-discharge complication (2.1%), followed by urinary tract infection (1.2%) and reintubation (1.2%). Readmission rates were significantly higher in patients experiencing a pre-discharge complication when compared to those not experiencing an in-hospital complication (17.5% vs 11.8%, p<0.01). Patients who experienced a pre-discharge complication had a longer mean length of post-operative hospital stay (13.5 days) as compared to those without a pre-discharge complication (5.7 days, p<0.01), but were no more likely to experience a post-discharge complication (7.1% vs. 6.6%, p=0.07). In the final adjusted model, pre-discharge complications were only associated with readmission when postoperative hospital stay was less than 7 days (OR=1.4, 95%CI=1.2-1.7), however only 11.5% of patients experiencing an in-hospital complication had a post-operative stay of less than 7 days.

Conclusion:
Occurrence of pre-discharge complications in patients with postoperative length of stay greater than seven days was not associated with readmission suggesting that sequelae of the complication resolve during the prolonged index hospitalization for the majority of admissions.

08.20 Early learners as health coaches for high-risk surgical patients: a pilot study

Posted on January 18, 2016

J. Kaplan1, R. Hofer1, Z. Brinson1, P. Chung2, C. Lucas2, D. Teng3, V. Tang5, J. Broering4, A. Chang2,5, E. Finlayson1 1University Of California – San Francisco,Surgery,San Francisco, CA, USA 2University Of California – San Francisco,School Of Medicine,San Francisco, CA, USA 3University Of California – San Francisco,School Of Nursing,San Francisco, CA, USA 4University Of California – San Francisco,Urology,San Francisco, CA, USA 5University Of California – San Francisco,Geriatrics,San Francisco, CA, USA

Introduction: There are few opportunities for early learners to engage in authentic roles on health care teams. In a new geriatric optimization clinic for frail high-risk surgical patients, first year medical and nurse practitioner students were integrated into an interprofessional team as health coaches.

Methods: Frail surgical patients (those over the age of 80 or over the age of 60 with a geriatric syndrome) with planned elective operations were referred to a new geriatric pre-operative clinic to see a geriatrician, occupational therapist, physical therapist, and nutritionist. Students attended the entire clinic visit with the patient, reviewed the interdisciplinary care plan, attempted to call patients twice weekly preoperatively to promote adherence to care plans, and called weekly in the first month after discharge. They observed surgery and practiced physical exam and presentation skills during the inpatient stay. Students logged all calls and patients filled out health coach satisfaction surveys one week prior to surgery.

Results: In the 6 month pilot period, two first year medical students, one nurse practitioner student, and two research assistants served as health coaches for 32 patients. On average, there were 30 days between the clinic visit and surgery. Patient received an average of 1 call per week. There were 39 instances when health coaches contacted members of the care team, 12 of which were to contact a doctor. These follow up communications were for medication questions, concerning symptoms, or remaining questions from clinic visits. Overall patients were extremely satisfied with the health-coach experience, they felt better prepared for surgery, and would recommend the program to others.

Conclusions: Early medical and nurse practitioner students can serve the important function of health coaches for frail patients preparing for surgery. Students benefited from a unique longitudinal experience and gained skills in communication and care coordination.

08.21 A Survey of the Preferences and Practice of Surgical Hand Antisepsis

Posted on January 18, 2016

B. S. Oriel1,2, K. M. Itani1,3,4 1VA Boston Healthcare System,Surgical Service,West Roxbury, MA, USA 2Tufts Medical Center,Department Of Surgery,Boston, MA, USA 3Boston University School Of Medicine,Department Of Surgery,Boston, MA, USA 4Harvard Medical School,Department Of Surgery,Boston, MA, USA

Introduction:
Products available for surgical hand antisepsis include traditional aqueous scrubs (TAS) and alcohol-based rubs (ABR). As part of a quality improvement initiative, we sought to better understand best practice techniques and preferences among operating room (OR) staff.

Methods:
One-hundred thirty-one anonymous paper surveys were distributed to all surgical and OR personnel who perform surgical hand antisepsis at either one or both of two hospital campuses. Participants represented all surgical specialties. The survey assessed product preference, product-specific patterns of use, hand antisepsis education and product knowledge.

Results:
A total of 97 surveys (74%) were returned; two were incomplete and excluded. The remaining 95 surveys included 44 women (47%), 37 staff surgeons (39%), 29 residents and fellows (31%), and 29 nursing staff (30%). Ninety-five percent confirmed prior use and proper application of both product types, however application times of less than 1 minute for ABR and less than 3-6 minutes for TAS were reported by 10% and 18% of participants, respectively. The application time of each product was most influenced by case-specific wound classification and hand soiling. Among all users, 64% supported having an ABR option available. ABR was favored (52%) and did not differ between physicians and nursing staff (p = 0.67). Respondents preferred ABR for its ease of use (79%) and TAS for its minimal to no side effects (54%). Forty-eight percent had not read the manufacturer's instructions for either product, and over 80% had never received formal in-person training. While 73% of physicians reported familiarity with each product's active ingredients, less than one-third correctly identified alcohol’s concentration of greatest anti-microbial activity or its mechanism of action, and only two-thirds correctly identified chlorhexidine's mechanism of action. Only 3 of 64 physician-respondents reported any familiarity with the Food and Drug Administration's Tentative Final Monograph.

Conclusion:
There were no strong preferences among staff regarding ABR or TAS but ABR was a supported option for surgical hand antisepsis. Preferred product features included ease of use and an acceptable side-effect profile. A definite lack of understanding and education exists regarding product mechanism of action and application. Perceptions of proper use and reported practice were discordant among some staff. The effect of reported practice on surgical site infections should be investigated.

09.01 Racial Disparities in the Use of Laparoscopic Colon Resection Among Medicare Patients

Posted on January 18, 2016

S. R. Thumma1, J. R. Thumma1, A. Ibrahim1, J. B. Dimick1 1University Of MIchigan, Center For Healthcare Outcomes And Policy,Ann Arbor, MI, USA

Introduction: Racial disparities in the use of minimally invasive technology, such as laparoscopic colectomy, are well documented. However, the underlying reasons for these disparities are not well understood. One potential explanation is that minorities receive care in hospitals that are less likely to use minimally invasive approaches. We sought to examine racial disparities in the use of laparoscopic technology for colectomy and explore the contributing factors, including whether the differences in the use of the minimally invasive approach can be explained by the hospitals where minorities receive their care.

Methods: We studied all patients with colon cancer who underwent colectomy (N=112,183, N=1,901 hospitals) in the national Medicare population over 3 years (2010-2012). Univariate and multivariate logistic regression models were used to examine the relationship between race and the use of the laparoscopic approach. In our analyses, we controlled for patient factors (age, gender, race and Elixhauser comorbidities), neighborhood socioeconomic status (SES), and the year of procedure. We present our results stratified by the degree of hospital segregation (i.e., quintiles of the concentration/proportion of blacks in the hospital). Moreover, we assessed the impact of hospitals on the use of laparoscopy by adjusting directly for them using fixed effects models.

Results: In the national Medicare population, 37% of patients received a laparoscopic approach to colectomy. When examining the racial composition of the hospitals where these patients received care, there appeared to be a large degree of racial segregation. For example, the proportion of black patients in a hospital ranged from 0% in the bottom 20th percentile to 49.2% in the top 20th percentile. Black patients were 23% less likely to receive laparoscopic colectomy when compared to non-black patients (OR, 0.77; 95% CI, 0.74-0.80). When each contributing factor was assessed independently using univariate analysis, 36% of the observed disparity in receiving laparoscopic colectomy was explained by the hospitals in which they received care (OR, 0.85; 95% CI, 0.82-0.89) and 39% was explained by SES (OR, 0.86; 95% CI, 0.83-0.89). After adjusting for differences in the hospitals where they received care, SES, and patient factors in a single model, we were able to explain 63% of the observed disparity, leaving 37% of the remaining disparity unexplained. However, even after adjusting for all these factors, black patients were still found to be 9% less likely to receive a laparoscopic colectomy compared to their non-black counterparts (OR, 0.91; 95% CI, 0.87-0.96).

Conclusion: Black patients are less likely than non-black patients to have access to laparoscopic procedures for colon cancer. These disparities are explained in large part by black patients receiving care in hospitals that are less likely to use the minimally invasive approach.

09.02 Using Community Outreach to Gain Insight into Racial Disparities and Cancer Care

Posted on January 18, 2016

R. L. Hoffman1, K. O’Neill2, K. Collier1, C. B. Aarons1, M. K. Lee1, R. R. Kelz1 1University Of Pennsylvania,Philadelphia, PA, USA 2Yale University School Of Medicine,New Haven, CT, USA

Introduction: Outreach programs aimed at reducing disparities in cancer care need to consider the cultural nuances of each community in order to be effective and sustainable. The aim of this study was to gain insight into the beliefs of the local community regarding colon cancer screening and treatment.

Methods: In partnership with a large urban university cancer center, medical students and surgical residents and faculty participated in two community health outreach events focused on cancer prevention, screening and treatment in the surrounding community. Prior to the main educational event, an 11- item survey that addressed commonly held cancer myths and colon cancer knowledge was distributed to all participants. Myth items were adapted from the National Cancer Institute’s ‘Common Cancer Myths and Misconceptions’ website. The survey also collected demographic information and assessed participant willingness to undergo cancer screening (colonoscopy). Myth items were scored on a 5-point Likert scale from ‘strongly agree to strongly disagree,’ and knowledge items were scored as true/false or ‘I don’t know.’ Descriptive statistics and univariate analyses were performed.

Results: There were a total of 256 participants and 188 completed a survey (71% response rate). Attendees were predominantly insured (93%), female (80%) and African American (96%), with a mean age of 61 years (14.0; range 12-95 years). Seven percent of participants were unwilling to undergo a screening colonoscopy; 3% of those age ≤60 years compared to 9% >60 years. Of those unwilling to get a colonoscopy, 33% were male, 92% Black. Of the myths, 25% felt that surgery could cause cancer to spread. Patients ≤60 years old were more likely to believe that surgery could cause cancer spread (64% vs. 35% ≤60yrs; p=0.004). 25% endorsed mistrust in physicians (43% of those ≤60 years, 26% for those >60 years (p=0.03). 18% felt that cancer treatment was worse than the disease. Belief in these myths did not differ by gender. 74% of participants agreed that there were some measures people could take to prevent cancer. 48% were aware of the causes of colon cancer, 49% and 46% misunderstood or didn’t know the definitions of malignant and metastatic, respectively. Males were more unsure of these definitions than females (61% vs 55%, p=0.5 and 70% vs 55%; p=.01).

Conclusion: In the surrounding, predominantly African American community, knowledge of and willingness to undergo cancer prevention strategies was high. However, in participants ≤60 years, in which early detection of cancers is most beneficial, physician mistrust and belief in the surgical spread of cancer was also high. Particularly for cancers where surgery is the mainstay of treatment, surgeon participation in community outreach events is essential in order to minimize barriers to treatment, increase cancer-related knowledge and address disparities in care.

09.03 Racial Disparities in the Receipt of Transdisciplinary Evidence-based Breast Cancer Care

Posted on January 18, 2016

R. Yang1, Y. Ma1, I. Wapnir1, K. F. Rhoads1 1Stanford University,Department Of Surgery,Palo Alto, CA, USA

Introduction:
Disparities in breast cancer have persisted over decades. While survival for White patients has improved, survival in select minority groups has worsened, thus widening the pre-existing gap. Biological factors have been well-studied, but it is not yet known if differences in breast cancer care contribute to worsening racial disparities. We aimed to evaluate differences in the receipt of transdisciplinary evidence-based breast cancer care for racial/ethnic minorities in California (2008-2009).

Methods:
California Cancer Registry (CCR) data were linked to state level inpatient and ambulatory surgery data and used to identify all women with an ICD-03 diagnosis of breast cancer who were treated during the years under study. We identified quality measures reflecting evidence-based care in the diagnosis, surgical treatments, and adjuvant treatments of breast cancer based on current literature and National Comprehensive Cancer Network guidelines. We evaluated the receipt of percutaneous versus excisional biopsy, neoadjuvant chemotherapy for tumor stage T4, chemotherapy for nodal stage N2, radiation therapy following lumpectomy; surgical evaluation of the axilla for stage M0; breast conservation for tumor stage T1 or T2; and breast reconstruction following mastectomy. Differences in receipt of evidence-based care by race/ethnicity were determined using the chi-squared test.

Results:
We identified 42,474 patients with stage I-IV breast cancer. We found that a higher proportion of Black patients (3.1%) underwent excisional biopsy compared to White, Hispanic, and API patients (2.2%, 2.5%, 2.6%, respectively, (p=0.02)). There was no significant difference in the receipt of systemic chemotherapy for nodal stage N2 by patient race (p=0.09). API patients had the lowest rates of neoadjuvant chemotherapy for T4 tumors (24.7%) compared to all other races (Black 28.4%, White 29.8%, Hispanic 42.9%, p=0.007). Radiation following lumpectomy was less frequent for Black (54.9%) and Hispanic patients (60.5%) compared to White (64.9%) and API patients (67.4%, p<0.001). Evaluation of the axilla (either SLNB or axillary dissection) for M0 disease was lower for Black (39.1%), Hispanic (40.6%), and API patients (44.4%) compared to White patients (48.3%, p<0.001). Rates of breast conserving surgery for T1 or T2 tumors were lower among API (45.9%), Hispanic (47.6%), and Black patients (50.1%) compared to White patients (53.4%, p<0.001). Rates of breast reconstruction following mastectomy were lower among Black (34.9%), Hispanic (34.3%), and API patients (36.0%) compared to White patients (40.8%, p<0.001).

Conclusion:

Racial disparities in the quality of care exist across diagnostic, surgical and adjuvant therapies. However, the disparities appear most pronounced within the surgical care. Further studies are needed to determine the drivers of this disparity.

09.04 High Volume Hospitals: Limited Access for Patients with a Lower Income?

Posted on January 18, 2016

N. R. Changoor1,2, J. W. Scott1, G. Ortega2, C. K. Zogg1, L. L. Wolf1, G. Reznor1, E. B. Schneider1, E. E. Cornwell2, A. H. Haider1 1Brigham And Women’s Hospital,Center For Surgery And Public Health,Boston, MA, USA 2Howard University College Of Medicine,Department Of Surgery,Washington, DC, USA

Introduction:
Studies have demonstrated that high volume centers are associated with superior outcomes for multiple cardiac, vascular and oncologic procedures. Racial/ethnic minority patients receive fewer of these procedures at high volume centers, the cause for which remains undetermined. Our study aims to investigate whether disparities in access to high volume hospitals are also associated with patient income.

Methods:
The 2005-2011 NIS was queried for patients with ICD-9-CM procedure/diagnostic codes for oncologic resection of the breast (BrCa), lung (LungCa), esophagus (EsophCa), stomach (GastCa) or colon (ColonCa) as well as for patients who underwent (procedure codes) CABG, AAA repair (AAAR), CEA, total hip replacement (THA), or total knee replacement (TKA). Hospitals treating patients within each operative group were dichotomized into centers that performed greater than 75th percentile (high volume centers) or less than the 25th percentile (low volume centers) of corresponding annual procedures. Median household income for patient's ZIP code was used to stratify patients into income quartile groups. Differences between the highest and lowest income quartiles were used to assess for associated differences in the risk-adjusted odds of presenting to a high-volume center for a given operation. Nationally-weighted, multivariable logistic regression accounted for differences in patient/hospital-level factors and clustering of patients within hospitals.

Results:
A combined total of 2,529,352 patients underwent included operations; most received surgery for BrCa (29.9%), TKA (21.3%) and THA (12.4%). The procedures with the highest proportion of high-income patients presenting to high-volume centers were ColonCa (90.6%), LungCa (74.7%), and BrCa (73.7%), while the procedures with the highest proportion of low-income patients presenting to high-volume centers included CABG (54.5%), AAAR (53.6%) and CEA (53.6%). Risk-adjusted analysis revealed that highest income patients were more likely to receive care at high-volume centers for THA (OR=1.96, CI:1.36-2.82), BrCa (OR=1.69, CI:1.07-2.67) and GastCa (OR=2.28, CI:1.05-4.94) but were less likely to receive care at high-volume centers for AAAR (OR=0.62, CI:0.40-0.95) when compared to lowest income patients. Differences for other procedures were not significant.

Conclusion:
The results indicate significant differences in access to high-volume centers associated with income for certain procedures. Varied associations demonstrated might imply important differences in respective patients’ ability and desire to receive care from high-volume centers, which may be dependent on the procedure type. Strengthening referral systems and re-evaluation of in-network referrals may help to increase access to high volume centers by low-income patients.

09.05 Trauma Recidivism in Urban Youths vs Adults

Posted on January 18, 2016

K. Barrera2, L. Lajoie2, A. Chudner2, D. Solomon2, V. Roudnitsky1, T. Schwartz1 1Kings County Hospital Center,Acute Care Surgery,Brooklyn, NY, USA 2SUNY Downstate,Department Of Surgery,Brooklyn, NY, USA

Introduction:
Recurrent violent injury (RVI) has been a topic of interest in trauma with recent reported rates greater than 30%. There has been specific focus on prevention, with many trauma centers implementing violence prevention programs aimed at youth. In these programs, patients are followed by case workers after discharge. Outcomes have been favorable with a trend towards reduction of recurrent injury as well as crime. Although there are several programs that include victims of injury of all ages, several programs target ages 10-30 years as recidivism is more prevalent in youths. The purpose of this study is to examine the differences between younger versus older recidivists to determine if adult recidivists may also benefit from efforts to reduce recidivism.

Methods:
This is a retrospective cohort study of all patients ages 15-60 seen by the trauma team at a single urban trauma center for violent injuries over a six year period. Patients were followed for recurrent injury for a minimum 3 years. Patients that died after first violent injury were excluded. Data collection also included mechanism of injury. Chi-square test, and T Test were used to determine differences in characteristics of recidivism between 15-25 and 25-60 age groups.

Results:
A total of 2642 patients were included in the analysis. This group was further subdivided into two age groups, 15-25 years (n=1300) and 26-60 years (1342). Overall recidivism rate was 11%. In the youth group, recidivism rate was 14% versus 9% in adults (p<0.001). Amongst the youth group, injuries inflicted by gunshot wounds, stab wounds and blunt assault comprised of 35%, 36%, and 29% of injuries respectively. In the adult group, this was 28%, 34%, and 38% respectively. There were no statistically significant differences between mechanisms of injury between the youth and adult groups. There was no significant difference in mean time to recurrence which was 66 months in youths and 62 months in adults (p=0.8). Mean abbreivated injury severity score was 4.8 in youths and 6.1 in adults and was not statistically significant.

Conclusion:
Violent injury prevention programs have focused on at risk youth as typically recidivism is significantly more prominent in this population. Despite a lower recidivism rate in an adult population, the prevalence is substantial. Adults experience similar current injuries when compared to youth in a similar time frame with a similar degree of injury. In urban trauma centers, violent injury prevention programs addressing an adult population may be beneficial in reduction of recurrent violent injury.

09.06 Health literacy and readability analysis of online resources related to lung cancer

Posted on January 18, 2016

K. D. Weiss1, C. R. Vargas1,2, O. Ho2, D. J. Chuang2, B. T. Lee1,2 1Beth Israel Deaconess Medical Center,Department Of Surgery,Boston, MA, USA 2Beth Israel Deaconess Medical Center,Division Of Plastic And Reconstructive Surgery,Boston, MA, USA

Introduction:

The Internet is a medium commonly utilized by patients seeking resources for health information; for many patients it has become the first source of health care information. Studies have shown that understanding supplemental educational material contributes to increased patient involvement and satisfaction. Limited functional health literacy is a potential barrier for access to these materials. Over 30% of the US population has basic to below basic literary skills which can contribute to overall health disparities, negatively impacts outcomes, and is even associated with increased mortality. As such, the National Institutes of Health and American Medical Association have recommended that patient-directed health resources should be written at a sixth-grade reading level. This study evaluates the readability of the most commonly accessed Internet resources on the subject of lung cancer.

Methods:

A search for ‘lung cancer’ was performed using Google and Bing, and the top 10 websites were identified. Location services were disabled and sponsored sites excluded. Relevant articles (n = 109) with patient-directed content available directly from the main sites were downloaded, and their readability was assessed using 10 established methods (Coleman-Liau, Flesch-Kincaid, Flesch Reading Ease, FORCAST, Fry Graph, Gunning Fog, New Dale-Chall, New Fog Count, Raygor Estimate, and SMOG). Readability scores were also grouped by parent website and by topic for comparison.

Results:

The average reading grade level across all sites was 11.2, with a range from 8.8 (New Fog Count) to 12.2 (SMOG). The readability of the material varied when compared by individual website, with a range from 9.2 to 15.2 (Cancer.gov 9.2, Mayoclinic.org 9.4, Cancer.org 9.6, Lung.org 10.1, Cancer.net 10.8, WebMD.com 11, Medicalnewstoday.com 12.2, Lungcancer.org 12.2, Medicinenet.com 13, Wikipedia.org 15.2). Only 10 articles (9%) were written at a 6th grade level or lower by any test and most of these were on the topics of prevention, risk factors, complications, and making treatment decisions.

Conclusion:

The most commonly web accessed patient-directed information about lung cancer consistently exceeds the recommended sixth grade reading level. Readability varies significantly between individual websites allowing physicians to identify and recommend sites according to a patient’s level of health literacy. Minor modifications to existing materials can significantly improve readability of materials while maintaining content for patients with low health literacy.

09.07 Identifying Predictors of Consent for Solid Organ Donation in Appalachia

Posted on January 18, 2016

J. W. Harris1, J. C. Berger1, R. Gedaly1, M. B. Shah1 1University Of Kentucky,Surgery,Lexington, KY, USA

Introduction: Appalachia is a distinct region characterized by high levels of poverty, poor access to healthcare and a high incidence of end stage organ disease. However, there is little reported regarding the availability and characteristics of potential organ donors. There is a need to understand factors that may impact consent for donation in potential donors residing in Appalachia.

Methods: All Appalachian potential donor (APD) referrals to our OPO from 2007-12 were analyzed. Death certificates for each APD were obtained. Demographics, marital status, education, registry status, cause and manner of death, decoupling, family member approached, understanding of hopelessness by family, and hospital region were collected. The US Census Report was used to obtain median county income and poverty rates. An analysis for donation consent was performed.

Results: 493 APD referrals were included. 207 consented for donation (CD) and 286 did not consent for donation (DNC). Those who CD were significantly younger (38 vs. 49 years, p<0.001). There were no differences in income. On univariate analysis, APD has lower CD if: age >40 vs <40 (32 vs 60%, p<0.001), high school or less education vs college (41 vs 58%, p<0.05), not registered vs on registry (38 vs 82%, p<0.001), OPO vs local provider vs family mentioned donation (37 vs 41 vs 73%, p<0.001), spouse/siblings/children vs parents approached for donation (35 vs 63%, p<0.001), if family did not understand hopelessness (20 vs 45%, p=0.009), declaration of death and approach for consent not decoupled (22 vs 47%, p<0.001), other cause of death vs trauma (36 vs 57%, p<0.001), medical vs non-medical mechanism of death (34 vs 54, p<0.001). There were no significant differences between race, APD gender and marital status, hospital region, miles from home, gender of family member approached, and poverty levels of APD compared to US and state poverty rates.

On multivariate analysis, predictors of CD were: age <18 years (OR 5.6 p<0.001), age 18-39 years (OR 6.5, p<0.001), college education (OR 3.3, p=0.008), graduate school (OR 4.3, p<0.001), on donor registry (OR 8.3, p<0.001), family mentioned donation (OR 5.6, p<0.001) and decoupling declaration of death from approach for donation (OR 5.8, p<0.001).

Conclusion: It is not surprising that poverty rates do not correlate with donation since Appalachia is generally poorer than numerous other areas. However, several observed factors can help target at-risk groups for poor consent in APD. Higher levels of education, younger age and family awareness of organ donation are independently associated with higher rates of consent. While DMV collaborations are paramount, targeting groups with poor consent rates more directly in their local communities may help increase awareness and donor registration. Additionally, OPO and local provider collaborations may maximize donation during the hospital phase of donation in non-registered APD.

09.08 Racial Disparities in Length of Stay Among Patients Undergoing Lower Extremity Revascularization

Posted on January 18, 2016

F. C. Patel1, M. D. Giglia1, A. Gullick1, M. S. Morris1, B. J. Pearce1, D. I. Chu1 1University of Alabama Birmingham,Department Of Surgery,Birmingham, ALABAMA, USA

Introduction: Racial disparities in surgical outcomes, such as readmissions, have been demonstrated in minority populations. Few studies have examined disparities in length-of-stay (LOS) for vascular procedures. We aim to investigate the role of race in determining LOS for patients undergoing lower extremity revascularization (LER) using a national, surgical outcomes registry.

Methods: We queried the 2012-2013 American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) to identify all patients who underwent elective LER and stratified patients by approach (open versus endovascular) and race. Patients were excluded if they had any in-hospital, post-operative complications during their index admission or a 30-day mortality. Chi-square and Wilcoxon Rank Sums tests were used to determine the differences among categorical and continuous variables, respectively. The primary model outcome was total post-operative LOS. Predictors of LOS were identified with multivariate regression using a negative binomial model.

Results: Of 6,843 patients who underwent lower extremity revascularization, 76.2% and 23.8% of patients underwent an open or endovascular approach, respectively. Black patients represented 16.3% of the overall cohort, with women representing 47.5% and 34.3% of black and white patients, respectively. Black and white patients were similar with respect to BMI (27.4 vs. 27.3), ASA class distribution, and functional health. Compared to white patients, black patients were younger (65 vs. 74 years, p<0.05) and had significantly higher rates of smoking, hypertension, dialysis, insulin-dependent diabetes, and steroid use (p<0.05). On adjusted comparison, black patients who underwent an open LER experienced a longer post-operative LOS (4 vs. 3 days, p<0.001) compared to white patients. With an endovascular approach, no significant difference in LOS existed between races (1 vs. 1 day, p>0.05).

Conclusion: Black patients undergoing open LER have a significantly longer LOS in comparison to white patients even with no in-hospital complications. No racial disparities in LOS were observed for patients undergoing LER by endovascular approach. Further investigations will need to examine non-NSQIP elements such as psychosocial, behavioral, and educational factors that may explain disparities in open LER.

08.05 Quality Metrics after Early Mobilization in the PACU in Colorectal Patients: A 6-Month Analysis

Posted on January 18, 2016

M. W. LaPorta1, E. T. Wills1, A. W. Trickey1, P. Graling1, D. B. Colvin1, J. J. Moynihan1, H. D. Reines1 1Inova Fairfax Hospital,Department Of Surgery,Falls Church, VA, USA

Introduction

The importance of early postoperative ambulation has been recognized for decades. Early ambulation is associated with decreased venous thromboembolism (VTE), atelectasis, urinary retention, and length of stay, as well as increased patient satisfaction, improved pain control and return of bowel function. Patients deserve to be active participants in facilitating a safe and swift postoperative recovery. We focused on an early ambulation initiative to reduce postoperative VTE in the colorectal surgery population.

Methods:

Patients ≥18 years of age undergoing elective procedures from one large private colorectal surgery practice were enrolled from October 2014-March 2015. Starting at the pre-surgical visit and continuing in the pre-operative area, patients were educated on the importance of early ambulation on recovery after surgery. Post-anesthesia care unit (PACU) nursing and staff ambulated patients 100 feet within 1 hour of reaching PACU. An interim analysis using retrospective chart review from the Electronic Medical Record and Premier Quality Database compared patient outcomes 6 months pre-implementation versus 6 months post-implementation; outcomes were compared overall then stratified among patients receiving open, laparoscopic, or robotic-assisted procedures. Chi-square or Fisher’s exact tests were calculated with an intent-to-treat analysis. All patients meeting inclusion criteria in the post-ambulation period were included in the analysis, regardless of ambulation status.

Results:

A total of 72 /279 (25.8%) patients successfully ambulated in the PACU. The mean ambulation distance was 128 feet. Team members overcame skepticism related to the safety of this endeavor and the constraints of limited available time for ‘bedside nursing’ by engaging all key stakeholders including: patients and families, perioperative nursing staff, anesthesia services, and the surgical team. Characteristics of pre-implementation patients (N=279) were similar to post-implementation (N=228), including age, body mass index, fall risk scores, surgeons, and procedure types. During the first 6 months of the colorectal mobility initiative, there were no postoperative VTE occurrences, compared to 5 VTE events pre-implementation. Although this reduction does not reach statistical significance, it represents a clinically meaningful improvement and a strong trend (1.8% vs. 0%, p=0.07). Postoperative infections were reduced from 29% to 23% (p=0.09), with a significant infection reduction in laparoscopic procedures (28% vs. 14%, p=0.05).

Conclusion:

Results of the mobility initiative demonstrate that early postoperative ambulation can be accomplished safely. While the observed VTE reduction has not yet achieved statistical significance, meaningful trends have emerged in decreasing VTE and surgical site infections. Large-scale implementation and further evaluation is ongoing at our institution to determine if the trend will continue.

08.06 Do Clinical Pathways Actually Improve Outcomes for Pediatric Appendicitis?

Posted on January 18, 2016

L. R. Putnam1,2, T. G. Ostovar-Kermani1,2, K. T. Anderson1,2, D. H. Pham1,2, M. T. Austin1,2, A. L. Kawaguchi1,2, 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: Standardized clinical care pathways for pediatric appendicitis are widely endorsed among children’s hospitals, yet their influence on patient outcomes is unknown. We aimed to determine the impact of a clinical care pathway for appendicitis, implemented at our children’s hospital in 2011, on length of stay, readmissions, and complications.

Methods: Pre-pathway data were compared to data from prospectively-monitored care pathways for all pediatric patients (<18 years) who underwent an appendectomy for acute appendicitis between June 2009 – December 2010 (pre-pathway) and January 2011 – May 2015 (pathway). Separate pathways were implemented for simple and complicated (gangrenous or perforated) appendicitis patients. Clinical outcomes including surgical site infections (SSI), initial and aggregate 30-day hospital length of stay (LOS), and readmissions within 30 postoperative days were recorded. Chi-squared, Student’s t-tests, Mann-Whitney U tests, and multivariate regression were performed; p-values <0.05 were considered significant.

Results: 2151 patients (pre-pathway: 582, pathway: 1569) underwent appendectomy for acute appendicitis from June 2009 – May 2015. Mean age of pre-pathway vs pathway patients was 9.9 ± 3.7 vs 10.4 ±3.8 (p=0.01). Gender, race/ethnicity, and insurance status were similar between groups (all p>0.05). Simple appendicitis pathway patients experienced decreased initial LOS (1.4 vs 0.9 days, p=0.03) and aggregate 30-day LOS with no difference in SSI, but increased readmissions (1.3% vs 3.4%, p<0.05). There were no significant differences in outcomes for complicated appendicitis pathway patients. On multivariate regression analysis of patients with simple appendicitis, the pathway predicted shorter LOS; younger age and public insurance predicted increased LOS and SSI (Table). For complicated appendicitis patients, laparoscopic appendectomy was associated with decreased LOS; there were no significant predictors of SSI or readmissions.

Conclusion:Clinical pathways do not appear to substantially improve outcomes in pediatric appendicitis. Within our institution, the clinical pathway for simple appendicitis patients was associated with decreased initial and aggregate 30-day LOS, but increased readmissions, whereas the complicated appendicitis pathway was not associated with any improvements. Further efforts should target decreasing readmissions for simple appendicitis patients and determining the impact of pathways on hospital resource utilization. Although there are many benefits to protocolized care, the true efficacy of increasing standardization may lie in process improvement and not directly related to patient outcomes.

08.17 Using Mixed Methods to Comprehensively Assess Processes of Surgical Care Delivery.

Posted on January 18, 2016

O. C. Nwanna-Nzewunwa1, M. Ajiko2, F. Kirya2, J. Epodoi2, F. Kabagenyi2, I. Feldhaus1, C. Juillard1, R. A. Dicker1 1University Of California – San Francisco,Center For Global Surgical Studies,San Francisco, CA, USA 2Soroti Regional Referral Hospital,Surgery,Soroti, SOROTI, Uganda

Introduction: Capacity assessment is an archetypal challenge in surgery. Existing assessment tools use the availability of surgical procedures and resources to identify availability of surgical care. Such methods may fail to capture important information relevant to decision-making and resource allocation among providers, policymakers, and international development partners. This study seeks the relevant information gaps in a tool currently used to assess surgical capacity and explores a mixed methods approach to generate a more complete assessment of hospital surgical capacity.

Methods: In June 2015, quantitative and qualitative research activities were conducted to assess emergency surgical care at a Ugandan Regional Referral Hospital. Infrastructure and human resources at a Regional Referral Hospital were assessed using the Surgeons OverSeas’ Personnel, Infrastructure, Procedures, Equipment and Supplies (PIPES) tool, generating a standardized index and score. Thematic analysis was conducted on four semi-structured focus group discussions with 18 purposively sampled providers involved in the process of surgical care delivery. The process of emergency surgical care was directly observed using time-and-motion methodology over 53 consecutive days to produce a process map.

Results: The PIPES tool identified major deficiencies in workforce and infrastructure; but not the cumulative impact of these deficiencies (e.g. lack of physical space) on the timeliness and process of care. The PIPES tool (see table) lacked indices to capture the in-hospital delays; barriers to accessing available procedures; effect of poor remuneration of providers on their availability at work; upshot of the ‘no user fees policy’ on patient throughput and quality of care. These were revealed in focus group discussions, midnight census and process mapping which also provided sociocultural context and the inimical effect of well-intentioned policies on the process of care. The PIPES tool understated the surgical workforce because it did not include orthopedic officers in its personnel assessment; this may misdirect training and capacity building efforts to only physicians or surgeons. The tool did not show the serious impact of the few unavailable procedures on patient throughput and access to care and may misrepresent the availability of services.

Conclusion: There is a need to augment the information obtained from existing surgical care assessment tools. We recommend the use of direct observation of the process of care in combination with qualitative interviews that capture the providers' experience in order to have a robust assessment of the process of surgical care, especially in developing countries where the systems and processes are less well understood.

08.18 Do Minimum Volume Standards Improve Patient Outcomes with Organ Transplant?

Posted on January 18, 2016

L. H. Nicholas1, S. Stith2 1Johns Hopkins University School Of Medicine,Surgery,Baltimore, MD, USA 2University Of New Mexico,Albequerque, NM, USA

Introduction: A pervasive viewpoint in healthcare is that higher patient volume leads to better outcomes, implying that facility volume can be used to identify high-quality hospitals. Hundreds of studies documenting a positive correlation between hospital volume and patient outcomes (better outcomes at higher volume hospitals) have motivated the use of arbitrary minimum volume standards for elective surgical procedures. There is considerable variation across minimum volume standards used by the Centers for Medicare and Medicaid Services and commercial insurers when contracting with organ transplant centers. It is unknown whether some payers are relying on excessively high standards or whether payers are using thresholds that are too small.

Methods: We used data from the Scientific Registry of Transplant Recipients including all adult heart (54,874 patients in 191 transplant centers) and lung (25,128 patients in 103 transplant centers) transplants in the United States from October 1987 – December 2012 to assess whether transplant outcomes including 1-year mortality and 1-year graft failure vary across centers above and below the minimum volume standards. To account for potential confounding between unobserved patient characteristics and transplant center volume, we used an instrumental variables approach to predict volume of a patient's center with the volume of the center closest to a patient's home zip code.

Results: Minimum volume thresholds affect many transplant centers. Only 57% of kidney transplant centers, 71% of liver, 44% of heart, and 61% of lung transplant centers ever meet the highest minimum volume requirement during our study period. We failed to find a significant relationship between transplant center volume and patient outcomes for either patient or graft survival for any of the four organs studied.

Conclusion: Despite good intentions, minimum volume standards currently used by public and private payers are not identifying higher quality transplant centers and may create unnecessary barriers for patients who could be served by smaller, closer centers.

08.07 Is ‘Routine’ Trauma Underprioritized in Level 1 Trauma Centers?

Posted on January 18, 2016

D. Metcalfe1, O. A. Olufajo1, C. K. Zogg1, M. B. Harris2, J. D. Gates1, A. J. Rios Diaz1, A. H. Haider1, A. Salim1 1Harvard Medical School,Center For Surgery And Public Health,Boston, MA, USA 2Brigham & Women’s Hospital,Department Of Orthopedic Surgery,Boston, MA, USA

Introduction:

There is strong evidence to show that level 1 trauma centers (L1TCs) improve outcomes for severely injured patients. However, L1TCs typically host many complex services and manage a high volume of critically unwell patients. It is therefore possible that some clinical pathways will be disrupted in L1TCs and that vulnerable patient groups might compete ineffectively for resources with higher priority cases. There is emerging evidence that appendectomy for acute appendicitis is delayed in L1TCs with an associated increased rate of complications. Studies from the UK and the Netherlands have also reported that the care of older adults with hip fractures may be compromised in new trauma centres.

Our study sought to compare hip fracture outcomes between L1TCs and non-trauma hospitals (NTHs) to determine whether "routine" trauma is underprioritized in mature higher level trauma centers.

Methods:

Hip fracture cases were identified from the California State Inpatient Database (SID) 2007-2011. The California SID captures 98% of hospital admissions and a unique patient identifier permits admissions to be tracked across all hospitals in California. The inclusion criteria were age ≥65 and an operatively treated hip fracture. To minimize selection bias, patients were excluded if they had any other injuries or were transferred between hospitals. Outcomes were analyzed using multivariable logistic regression and generalized linear models for non-normally distributed data, adjusting for patient- (age, sex, race, payer status, Charlson Co-morbidity Index, weekend admission, admission source) and hospital-level (hospital bed size, teaching status) characteristics.

Results:

91,401 hip fracture admissions were identified, 6,468 (7.1%) of which were admitted to L1TCs and 61,896 (67.7%) to NTHs. The remaining patients (25.2%) were treated in lower level trauma centers.

The delay between admission and operation was longer in L1TCs (median 1 day; 90th percentile 3 days) than in NTHs (median 1 day; 90th percentile 2 days, Kruskall-Wallis p<0.001). Within a generalized linear model, operative delay was 0.29 (95% CI 0.08-0.51) days longer in L1TCs relative to NTHs. Length of stay was also prolonged in L1TCs by 0.85 days (95% CI 0.30-1.40). Both the odds of venous thromboembolism (OR 1.45, 95% CI 1.11-1.88) and unplanned 30-day readmission (OR 1.49, 95% CI 1.24-1.80) were higher in L1TCs. There were no mortality differences between L1TCs and NTHs.

Conclusion:

Operative treatment of hip fractures is delayed in L1TCs, which is associated with prolonged length of stay and increased risk of both venous thromboembolism and 30-day hospital readmission. Further work should aim to understand whether these findings can be explained by intense competition for resources (e.g. operating room time) and how clinical pathways for vulnerable populations can be optimized in L1TCs.

Next Page »