- Email: [email protected]
The role of surgery in stage IIIA non-small cell lung cancer: a decision and cost-effectiveness analysis
e142
Scientific Poster Presentations: 2014 Clinical Congress
J Am Coll Surg
Intermediate (5.9%) > Standard (2.1%). 30-Day Morbidity for VASQIP assessed surgical procedures by surgical complexity designation showed Complex (14.5%) > Intermediate (10.9%) > Standard (4.3%).
INTRODUCTION: Bleeding peptic ulcer disease (bPUD) is a prevalent inpatient condition that is often life threatening. It might be expected that with development of new treatment modalities the management and outcomes of bPUD have changed over time.
CONCLUSIONS: 30-Day Mortality and Morbidity data analysis validates the VHA Surgical Complexity Model.
METHODS: The United States Nationwide Inpatient Sample was used to identify patients hospitalized for bleeding from acute and chronic gastric, duodenal, and peptic ulcers from 1998-2011. Cross-linkage of diagnostic and procedural ICD-9 codes was used for selection. Chi-square and t-tests were used for analysis between groups, multiple logistic regressions for assessment of outcome predictors.
Can we improve prediction models for mortality in severe traumatic brain injury? Aaron Dawes, MD, Greg D Sacks, MD, MPH, H Gill Cryer, MD, PhD, Christy Preston, RN, Deidre Gorospe, RN, J Peter Gruen, MD, Matt Garrett, MD, Melinda A Maggard Gibbons, MD, FACS, Marcia M Russell, MD, FACS, Clifford Y Ko, MD, FACS Los Angeles County Division of Emergency Medical Services and Los Angeles County and University of Southern California Medical Center and University of California-Los Angeles, Los Angeles, CA INTRODUCTION: Traumatic brain injury (TBI) remains a major source of morbidity and mortality. The Trauma Quality Improvement Program (TQIP) developed risk-adjustment models to benchmark hospital performance for patients with TBI. This study explored adding variables specific to TBI to the current TQIP models for mortality. METHODS: Prospectively collected data from all patients with severe TBI presenting to 14 Los Angeles County trauma centers between 2009-2010 were analyzed. Inclusion criteria were Glasgow Coma Scale (GCS) 8, blunt mechanism, and at least one intracranial injury on initial head CT. Two logistic regression models were developed to predict in-hospital mortality: one using only variables included in TQIP (age, gender, race, vital signs, mechanism of injury, injury severity score, GCS, and comorbidities), and a second adding pupil reactivity, initial INR, and specific head CT findings. RESULTS: During the two-year study period, 823 patients met inclusion criteria. Inpatient mortality was 38.8%. INR > 1.4 and head CT findings of cerebral edema and loss of grey/white differential significantly increased the risk of death, while pupil reactivity and the presence of epidural hematoma were found to be protective. The area under the receiver operator curve increased from 0.867 to 0.931 after the inclusion of the TBI-specific variables (p<0.001 under the log likelihood test). CONCLUSIONS: Pupillary response, INR at presentation, and head CT findings improve the accuracy of risk-adjustment models for mortality in TBI. These variables should be added to trauma registries to improve risk-adjusted benchmarking, and to help identify best practices in TBI care. Trends in the management of bleeding peptic ulcers: less traumatic, more successful? Mikhail Kryzhanouski, MD, Kenneth W Bueltmann, MD, Marek Rudnicki, MD, PhD, FACS Advocate Illinois Masonic Medical Center, Chicago, IL
RESULTS: Annual hospitalizations for bPUD decreased from 48.0 to 31.8 discharges per 100,000 people from 1998 to 2011 (p<0.001). In-hospital mortality decreased from 2.59% to 1.86% (p<0.001). The incidence of endoscopic control of bleeding (254 to 351/1,000) and transcatheter embolization (2 to 9.2/1,000) related discharges increased, whereas utilization of surgical procedures decreased (oversewing: 27.3 to 17.4/1,000, vagotomy: 20.7 to 6.2/1,000, gastrectomy: 14.2 to 7.2/1,000 discharges, p<0.001). AIDS, liver disease, coagulopathy, and metastatic cancer were comorbidities associated most strongly with the negative outcome. CONCLUSIONS: There is a decreasing trend in the hospitalization of patients with bPUD over the analyzed 14 years. The roles of endoscopy and transcatheter embolization have increased, while surgical procedures demonstrate decline. AIDS, liver disease, coagulopathy, and metastatic cancer were shown to most negatively affect outcomes of treatment. The role of surgery in stage IIIA non-small cell lung cancer: a decision and cost-effectiveness analysis Pamela P Samson, MD, Aalok Patel, BS, Clifford G Robinson, MD, Daniel Morgensztern, MD, Su-Hsin Chang, PhD, Graham A Colditz, MD, DrPH, Bryan F Meyers, MD, MPH, Traves D Crabtree, MD, Varun Puri, MBBS, FACS Washington University, Saint Louis, MO INTRODUCTION: The purpose of this study was to determine the relative cost-effectiveness of combination chemotherapy, radiation, and surgery (CRS) versus definitive chemotherapy and radiation only (CR) in Stage IIIA non-small cell lung cancer (NSCLC). METHODS: Patients with clinical stage IIIA NSCLC receiving either CR or CRS from 1998-2010 were identified from the National Cancer Data Base (NCDB). Propensity score matching for patient characteristics was performed. The Medicare allowable charges were obtained for the treatment modalities (CR $17,614 and CRS $31,846), and associated mortality risks (CR $12,148, CRS $55,513). The incremental cost effectiveness ratio (ICER) was calculated as the cost per life year gained. RESULTS: 5,265 propensity matched pairs were identified in the two treatment groups. Inputs from the NCDB included a 30-day mortality rate of 0.003% with CR, and 2.2% with CRS, and
Vol. 219, No. 4S, October 2014
one-year intervals of probabilistic survival. From decision modeling, effectiveness (mean survival) after the CR regimen was 2.10 years and after CRS treatment was 2.93 years, with an ICER of $17,617. At a willingness-to-pay of $50,000, the minimum survival difference needed for CRS to be cost-effective is 3.5 months. A two-way sensitivity analysis varied the likelihood of 5-year survival from 0 to 50%, while increasing the cost of surgery, and the CRS regimen dominated the model with a willingness-to-pay of $50,000. CONCLUSIONS: In Stage IIIA NSCLC, surgery in selected patients is associated with a survival benefit that is cost-effective across a variety of cost and survival estimates within a conventional willingness-to-pay of $50,000. Antibiotic choice in thyroidectomy affects outcomes and costs: an analysis of University HealthSystem Consortium data Alexander Langerman, MD, FACS, Sandra A Ham, MS, Jennifer Pisano, MD University of Chicago, Chicago, IL INTRODUCTION: Thyroidectomy is a common procedure, yet data on the consequences of perioperative antibiotic practices are lacking. METHODS: University HealthSystem Consortium inpatient and 30-day readmission billing data on patients undergoing thyroidectomy for thyroid cancer in 2008-2011 for 95 academic and affiliated medical centers were analyzed for antibiotic use, outcomes, and cost. RESULTS: Data from 9035 patients (72.4% women) were included in the study. Antibiotic management over the first 48h revealed 267 unique management strategies. The majority of patients were treated with either no perioperative antibiotics (42.0%) or with cefazolin or cefuroxime (45.1%). Hospitals varied from 0% of cases using no perioperative antibiotics and 41.6% using cefazolin or cefuroxime to 93.6% using no antibiotics and 2.6% using cefazolin or cefuroxime. In adjusted models, use of clindamycin versus no antibiotic had a significant association with reopening the wound (OR¼2.45, 95% CI¼[1.00-5.98]) and use of any antibiotic was associated with antibiotic-related complications (OR¼3.67 [1.87-7.21]). No differences were observed between antibiotic regimens in rates of wound infection. The total cost of hospital admission for patients who were managed with clindamycin ($11,073 [9,677-12,470]) was higher than those managed with no antibiotics ($8,492 [8,319-8,664]; p<0.001). CONCLUSIONS: There is substantial variability in perioperative antibiotic strategies for thyroidectomy. Clindamycin had significantly higher rates of reopening surgical wound and antibioticrelated complications compared to other common regimens and was associated with a higher total hospital cost. These data suggests that avoiding antibiotics for thyroidectomy is associated with lower rates of antibiotic-related complications and lower total hospitalization costs.
Scientific Poster Presentations: 2014 Clinical Congress
e143
Implementing a standardized safe surgery program reduces serious reportable events Terrence J Loftus, MD, FACS, Kathleen Harder, PhD, Deb Dahl, MBA, FACHE, Bridget O’Hare, Karlene Power, Yvette Toledo-Katsenes, Ryan Hutchison, MBA, BSIE, David Jacofsky, MD University of Minnesota, Minneapolis, MN and the Core Institute and Banner Health, Pheonix, AZ INTRODUCTION: Patient safety in the perioperative period is essential for the delivery of quality patient care. Mainstream quality organizations have developed safe surgery recommended practices for ensuring patient safety. Implementing safe surgery practices should result in a reduction in serious reportable events. METHODS: This retrospective cohort study compared the before and after results following the implementation of a standardized safe surgery program across a large healthcare system. Audits were performed to assure adoption of the new process. Serious reportable event (SRE) rates (retained surgical item, wrong site, wrong patient and wrong procedure) were tracked. Statistical analyses were performed on the SRE rate and days between SRE’s. RESULTS: A total of 683,193 cases in the operating room and labor and delivery were evaluated over a 4 year period. The SRE rate prior to implementation was 0.075/1000 cases and after implementation was 0.037/1000 cases. There was a 52% reduction in the SRE rate (p < 0.05). The mean time between SRE’s increased from 27.4 days to 60.6 days (p < 0.05). Robotic and non-robotic cases were affected equally however, a significant difference in SRE rate persisted between robotic and non-robotic cases (p < 0.05). Robotic cases are 7 times more likely to incur an SRE. Audits demonstrated the compliance rates for the program improved to 96% following complete system implementation. CONCLUSIONS: A standardized safe surgery program results in a significant reduction in SRE’s and an increase in time between events. Robotic cases are at high risk for an SRE. Quality of communication during transfer to definitive care in a provincial trauma system: a review of 100 transfers Nori L Bradley, MD, Naisan Garraway, MD, CD, Jennifer Li, BSc, Nasira Lakha, RN, Richard K Simons, MB, BChir, FACS, FRCS, FRCSC, S Morad Hameed, MD, MPH, FRCPS Vancouver Coastal Health and University of British Columbia, Vancouver, British Columbia, Canada INTRODUCTION: Poor communication during trauma patient transfer is linked to adverse events. Our objective was to characterize communication (completeness and quality) of critical clinical data during inter-hospital transfer to definitive care in a provincial trauma system. This work follows up information identified from the BC Trauma Registry.
Scientific Poster Presentations: 2014 Clinical Congress
J Am Coll Surg
Intermediate (5.9%) > Standard (2.1%). 30-Day Morbidity for VASQIP assessed surgical procedures by surgical complexity designation showed Complex (14.5%) > Intermediate (10.9%) > Standard (4.3%).
INTRODUCTION: Bleeding peptic ulcer disease (bPUD) is a prevalent inpatient condition that is often life threatening. It might be expected that with development of new treatment modalities the management and outcomes of bPUD have changed over time.
CONCLUSIONS: 30-Day Mortality and Morbidity data analysis validates the VHA Surgical Complexity Model.
METHODS: The United States Nationwide Inpatient Sample was used to identify patients hospitalized for bleeding from acute and chronic gastric, duodenal, and peptic ulcers from 1998-2011. Cross-linkage of diagnostic and procedural ICD-9 codes was used for selection. Chi-square and t-tests were used for analysis between groups, multiple logistic regressions for assessment of outcome predictors.
Can we improve prediction models for mortality in severe traumatic brain injury? Aaron Dawes, MD, Greg D Sacks, MD, MPH, H Gill Cryer, MD, PhD, Christy Preston, RN, Deidre Gorospe, RN, J Peter Gruen, MD, Matt Garrett, MD, Melinda A Maggard Gibbons, MD, FACS, Marcia M Russell, MD, FACS, Clifford Y Ko, MD, FACS Los Angeles County Division of Emergency Medical Services and Los Angeles County and University of Southern California Medical Center and University of California-Los Angeles, Los Angeles, CA INTRODUCTION: Traumatic brain injury (TBI) remains a major source of morbidity and mortality. The Trauma Quality Improvement Program (TQIP) developed risk-adjustment models to benchmark hospital performance for patients with TBI. This study explored adding variables specific to TBI to the current TQIP models for mortality. METHODS: Prospectively collected data from all patients with severe TBI presenting to 14 Los Angeles County trauma centers between 2009-2010 were analyzed. Inclusion criteria were Glasgow Coma Scale (GCS) 8, blunt mechanism, and at least one intracranial injury on initial head CT. Two logistic regression models were developed to predict in-hospital mortality: one using only variables included in TQIP (age, gender, race, vital signs, mechanism of injury, injury severity score, GCS, and comorbidities), and a second adding pupil reactivity, initial INR, and specific head CT findings. RESULTS: During the two-year study period, 823 patients met inclusion criteria. Inpatient mortality was 38.8%. INR > 1.4 and head CT findings of cerebral edema and loss of grey/white differential significantly increased the risk of death, while pupil reactivity and the presence of epidural hematoma were found to be protective. The area under the receiver operator curve increased from 0.867 to 0.931 after the inclusion of the TBI-specific variables (p<0.001 under the log likelihood test). CONCLUSIONS: Pupillary response, INR at presentation, and head CT findings improve the accuracy of risk-adjustment models for mortality in TBI. These variables should be added to trauma registries to improve risk-adjusted benchmarking, and to help identify best practices in TBI care. Trends in the management of bleeding peptic ulcers: less traumatic, more successful? Mikhail Kryzhanouski, MD, Kenneth W Bueltmann, MD, Marek Rudnicki, MD, PhD, FACS Advocate Illinois Masonic Medical Center, Chicago, IL
RESULTS: Annual hospitalizations for bPUD decreased from 48.0 to 31.8 discharges per 100,000 people from 1998 to 2011 (p<0.001). In-hospital mortality decreased from 2.59% to 1.86% (p<0.001). The incidence of endoscopic control of bleeding (254 to 351/1,000) and transcatheter embolization (2 to 9.2/1,000) related discharges increased, whereas utilization of surgical procedures decreased (oversewing: 27.3 to 17.4/1,000, vagotomy: 20.7 to 6.2/1,000, gastrectomy: 14.2 to 7.2/1,000 discharges, p<0.001). AIDS, liver disease, coagulopathy, and metastatic cancer were comorbidities associated most strongly with the negative outcome. CONCLUSIONS: There is a decreasing trend in the hospitalization of patients with bPUD over the analyzed 14 years. The roles of endoscopy and transcatheter embolization have increased, while surgical procedures demonstrate decline. AIDS, liver disease, coagulopathy, and metastatic cancer were shown to most negatively affect outcomes of treatment. The role of surgery in stage IIIA non-small cell lung cancer: a decision and cost-effectiveness analysis Pamela P Samson, MD, Aalok Patel, BS, Clifford G Robinson, MD, Daniel Morgensztern, MD, Su-Hsin Chang, PhD, Graham A Colditz, MD, DrPH, Bryan F Meyers, MD, MPH, Traves D Crabtree, MD, Varun Puri, MBBS, FACS Washington University, Saint Louis, MO INTRODUCTION: The purpose of this study was to determine the relative cost-effectiveness of combination chemotherapy, radiation, and surgery (CRS) versus definitive chemotherapy and radiation only (CR) in Stage IIIA non-small cell lung cancer (NSCLC). METHODS: Patients with clinical stage IIIA NSCLC receiving either CR or CRS from 1998-2010 were identified from the National Cancer Data Base (NCDB). Propensity score matching for patient characteristics was performed. The Medicare allowable charges were obtained for the treatment modalities (CR $17,614 and CRS $31,846), and associated mortality risks (CR $12,148, CRS $55,513). The incremental cost effectiveness ratio (ICER) was calculated as the cost per life year gained. RESULTS: 5,265 propensity matched pairs were identified in the two treatment groups. Inputs from the NCDB included a 30-day mortality rate of 0.003% with CR, and 2.2% with CRS, and
Vol. 219, No. 4S, October 2014
one-year intervals of probabilistic survival. From decision modeling, effectiveness (mean survival) after the CR regimen was 2.10 years and after CRS treatment was 2.93 years, with an ICER of $17,617. At a willingness-to-pay of $50,000, the minimum survival difference needed for CRS to be cost-effective is 3.5 months. A two-way sensitivity analysis varied the likelihood of 5-year survival from 0 to 50%, while increasing the cost of surgery, and the CRS regimen dominated the model with a willingness-to-pay of $50,000. CONCLUSIONS: In Stage IIIA NSCLC, surgery in selected patients is associated with a survival benefit that is cost-effective across a variety of cost and survival estimates within a conventional willingness-to-pay of $50,000. Antibiotic choice in thyroidectomy affects outcomes and costs: an analysis of University HealthSystem Consortium data Alexander Langerman, MD, FACS, Sandra A Ham, MS, Jennifer Pisano, MD University of Chicago, Chicago, IL INTRODUCTION: Thyroidectomy is a common procedure, yet data on the consequences of perioperative antibiotic practices are lacking. METHODS: University HealthSystem Consortium inpatient and 30-day readmission billing data on patients undergoing thyroidectomy for thyroid cancer in 2008-2011 for 95 academic and affiliated medical centers were analyzed for antibiotic use, outcomes, and cost. RESULTS: Data from 9035 patients (72.4% women) were included in the study. Antibiotic management over the first 48h revealed 267 unique management strategies. The majority of patients were treated with either no perioperative antibiotics (42.0%) or with cefazolin or cefuroxime (45.1%). Hospitals varied from 0% of cases using no perioperative antibiotics and 41.6% using cefazolin or cefuroxime to 93.6% using no antibiotics and 2.6% using cefazolin or cefuroxime. In adjusted models, use of clindamycin versus no antibiotic had a significant association with reopening the wound (OR¼2.45, 95% CI¼[1.00-5.98]) and use of any antibiotic was associated with antibiotic-related complications (OR¼3.67 [1.87-7.21]). No differences were observed between antibiotic regimens in rates of wound infection. The total cost of hospital admission for patients who were managed with clindamycin ($11,073 [9,677-12,470]) was higher than those managed with no antibiotics ($8,492 [8,319-8,664]; p<0.001). CONCLUSIONS: There is substantial variability in perioperative antibiotic strategies for thyroidectomy. Clindamycin had significantly higher rates of reopening surgical wound and antibioticrelated complications compared to other common regimens and was associated with a higher total hospital cost. These data suggests that avoiding antibiotics for thyroidectomy is associated with lower rates of antibiotic-related complications and lower total hospitalization costs.
Scientific Poster Presentations: 2014 Clinical Congress
e143
Implementing a standardized safe surgery program reduces serious reportable events Terrence J Loftus, MD, FACS, Kathleen Harder, PhD, Deb Dahl, MBA, FACHE, Bridget O’Hare, Karlene Power, Yvette Toledo-Katsenes, Ryan Hutchison, MBA, BSIE, David Jacofsky, MD University of Minnesota, Minneapolis, MN and the Core Institute and Banner Health, Pheonix, AZ INTRODUCTION: Patient safety in the perioperative period is essential for the delivery of quality patient care. Mainstream quality organizations have developed safe surgery recommended practices for ensuring patient safety. Implementing safe surgery practices should result in a reduction in serious reportable events. METHODS: This retrospective cohort study compared the before and after results following the implementation of a standardized safe surgery program across a large healthcare system. Audits were performed to assure adoption of the new process. Serious reportable event (SRE) rates (retained surgical item, wrong site, wrong patient and wrong procedure) were tracked. Statistical analyses were performed on the SRE rate and days between SRE’s. RESULTS: A total of 683,193 cases in the operating room and labor and delivery were evaluated over a 4 year period. The SRE rate prior to implementation was 0.075/1000 cases and after implementation was 0.037/1000 cases. There was a 52% reduction in the SRE rate (p < 0.05). The mean time between SRE’s increased from 27.4 days to 60.6 days (p < 0.05). Robotic and non-robotic cases were affected equally however, a significant difference in SRE rate persisted between robotic and non-robotic cases (p < 0.05). Robotic cases are 7 times more likely to incur an SRE. Audits demonstrated the compliance rates for the program improved to 96% following complete system implementation. CONCLUSIONS: A standardized safe surgery program results in a significant reduction in SRE’s and an increase in time between events. Robotic cases are at high risk for an SRE. Quality of communication during transfer to definitive care in a provincial trauma system: a review of 100 transfers Nori L Bradley, MD, Naisan Garraway, MD, CD, Jennifer Li, BSc, Nasira Lakha, RN, Richard K Simons, MB, BChir, FACS, FRCS, FRCSC, S Morad Hameed, MD, MPH, FRCPS Vancouver Coastal Health and University of British Columbia, Vancouver, British Columbia, Canada INTRODUCTION: Poor communication during trauma patient transfer is linked to adverse events. Our objective was to characterize communication (completeness and quality) of critical clinical data during inter-hospital transfer to definitive care in a provincial trauma system. This work follows up information identified from the BC Trauma Registry.