83

ASSOCIATION FOR ACADEMIC SURGERY AND SOCIETY OF UNIVERSITY SURGEONS—ABSTRACTS

81. RANKING HOSPITALS ON SURGICAL QUALITY: DOES RISK-ADJUSTMENT ALWAYS MATTER? J. B. Dimick, J. D. Birkmeyer; University of Michigan, Ann Arbor, MI Introduction: It is a widely held belief that detailed riskadjustment is always necessary in comparative reports of surgical performance. We sought to evaluate the importance of riskadjustment for two cardiac surgery report cards in New York and Pennsylvania. Methods: We abstracted data directly from the two most recent cardiac surgery report cards from New York State and Pennsylvania. We first estimated the correlation between unadjusted and risk-adjusted mortality rates. Based on this first report card, we then divided the hospitals into three groups of historical performance (best, average, and worst) for both unadjusted and risk-adjusted mortality rankings. We then calculated the subsequent risk-adjusted mortality within each of these groups using data from the report card from the subsequent year. Results: Risk-adjusted and unadjusted mortality rates were highly correlated for both New York (Pearson’s r ⫽ 0.95; Spearman’s r⫽0.91) and Pennsylvania (Pearson’s r ⫽ 0.87; Spearman’s r⫽0.89). For both states, riskadjusted and unadjusted rankings were equally good at predicting subsequent mortality. In New York State, mortality for hospitals in the worst group was 50% higher than the best group regardless of whether unadjusted (relative risk [RR], 1.51) or adjusted (RR, 1.49) rankings were used. The same was found in Pennsylvania where the results for unadjusted (RR, 1.53) and adjusted (RR, 1.45) rankings were nearly identical. Conclusions: Risk-adjusted and unadjusted mortality rates from two prominent cardiac surgery report cards are highly correlated and equally good at predicting subsequent performance. In certain contexts, risk-adjustment may not be as important as previously believed. 82. HOSPITAL VOLUME AND OUTCOME AFTER ABDOMINAL AORTIC ANEURYSM REPAIR IN THE ENDOVASCULAR ERA. J. B. Dimick, G. R. Upchurch, Jr.; University of Michigan, Ann Arbor, MI Introduction: The relationship between hospital volume and outcomes after open abdominal aortic aneurysm (AAA) repair is well established. However, it is unclear whether this relationship persists after the introduction of endovascular technology. Methods: We identified all patients undergoing AAA repair in the Nationwide Inpatient Sample to from 2001-2003. We examined the impact of hospital volume on in-hospital mortality for all AAA repairs, and then stratified by the type of repair (open and endovascular). For our analyses, we converted the volume variable into a categorical variable by creating five equal size patient groups (quintiles) for each volume measure (total, open, and endovascular volume).We performed risk-adjusted analysis accounting for differences in baseline patient characteristics using logistic regression. Results: Using the national dataset, we estimate that 158,800 abdominal aortic aneurysm repairs were performed in the United States during the three year period 2001-2003. Of the total repairs performed, 44,800 (27%)

185

were the endovascular approach. Over the study period, endovascular repair assumed a larger proportion of the market-share (from 23% to 33% of total repairs, P⬍.001). Endovascular repair was associated with a much lower in-hospital mortality rate compared with open repair (4.4% vs. 1.2%, P⬍.001). In the risk-adjusted analysis, mortality rates were 40% higher at hospitals in the lowest quintile of total volume when considering all types of repair together (OR, 1.40; 95% CI, 1.14-1.73). A similar relationship between total volume and mortality was found when examining each approach separately: open repair (OR, 1.32; 95% CI, 1.06-1.32) and endovascular repair (OR, 2.55; 95% CI, 1.31-4.97). The relationship between operationspecific volume and mortality was similar to that observed for total volume with both approaches. Conclusions: The introduction of endovascular surgery has not meaningfully altered the observed impact of hospital volume on operative mortality. We examined the relationship between three volume variables (total, open, and endovascular) and found each to be related inversely to operative mortality. We also found that a hospital’s overall or total abdominal aortic aneurysm repair volume is a good proxy for quality with each individual type of repair.

83. THE FINANCIAL IMPLICATIONS OF CESSATION OF MEDICARE PAYMENT FOR SURGICAL SITE INFECTION. J. A. Cohn, Y. M. Ads, M. J. Englesbe, J. L. Paruch, J. C. Magee, D. A. Campbell, Jr.; University of Michigan, Ann Arbor, MI Hypothesis: As part of report CMS-1488-P, the Centers for Medicare and Medicaid Services (CMS) proposed a rule that would result in cessation of payment for care associated with “preventable” hospital-acquired infections, including surgical site infections (SSIs), beginning October 1, 2008. We attempted to quantify the financial implications this rule would have on medical centers. Methods: The National Surgical Quality Improvement Program (NSQIP) perioperative clinical and financial data were recorded on a random selection of patients operated on between July 1, 2003 and December 31, 2005 at the University of Michigan Health System (UMHS). Demographics, pre and post-operative conditions, and financial data of patients with and without surgical site infections were compared. Patients were also stratified by American Society of Anesthesia (ASA) pre-operative score. Variables were evaluated with a student’s t-test or a chi-square test. Differences were considered significant if P⬍0.05. Results: Of the 5409 patients constituting the study population, 320 patients had a wound infection and 5089 patients did not. Those suffering from a wound infection were sicker prior to surgery, including: ASA score of 3 or higher (30.5% no SSI, 50.6% with SSI, P⬍0.001), ascites (0.8%, 2.8%, P⫽0.009), bleeding (4.2%, 9.0%, P⫽0.002), COPD (3.7%, 7.9%, P⫽0.005), dyspnea (9.7%, 17.4%, P⫽0.001), open wound (4.0%, 14.6%, P⬍0.001), sensory loss (0.5%, 2.8%, P⬍0.001), sepsis (4.3%, 7.9%, P⬍0.001), and unclean surgical incision (52.9%, 69.9%, P⬍0.001). In addition, SSIs were associated with more post-operative complications, including: acute renal failure (0.5%, 2.8%, P⬍0.001), DVT (0.5%, 2.2%, P⬍0.001), myocardial infarction (0.4%, 1.9%, P⬍0.001), ventilator 48 hours post-operation (1.6%, 7.2%, P⬍0.001), progressive renal insufficiency (0.3%, 1.6%, P⫽0.001), pulmonary embolism (0.2%, 1.3%, P⫽0.001), systemic inflammatory response syndrome (2.3%, 15.9%, P⬍0.001), unplanned intubation (1.1%, 4.7%, P⬍0.001), other cardiac event (0.4%, 1.6%, P⫽0.004), other respiratory event (0.5%, 3.8%, P⬍0.001), and CNS occurrences (0.4%, 2.2%, P⬍0.001). SSIs was associated with a significantly increased length of stay (9.2 days, P⬍0.001). Average costs, hospital reimbursement, and margin were $21,131, $21,557, and $433 higher, respectively for those patients with a surgical site infection (P⬍0.001, ⬍0.001, and ⫽0.72, respectively). Financial data stratified by ASA score are listed in Table 1. Hospital costs and reimbursements were significantly higher among patients with a surgical site infection for ASA scores 1, 2, and 3 but not for those patients with an ASA score of 4. Conclusion: Surgical site infections

186

ASSOCIATION FOR ACADEMIC SURGERY AND SOCIETY OF UNIVERSITY SURGEONS—ABSTRACTS

are expensive for payers but may actually augment hospital margin. SSIs are associated with patients with a number of pre-operative risk factors. Depending on how the proposed rule on reimbursement for hospital acquired infections in CMS-1488-P is applied, it could have significant effects on hospital finances and lead medical center market pressures away from complex patients at high risk for surgical site infections.

Table 1: Finances of no wound infection vs. wound infection by ASA score ASA Class

1

2

3

4

No wound infection

N

Wound infection

N

Difference

P-Value

$28,580 ⫾ 23,279

10

$17,332

0.01

$15,951

0.03

-$1,372

0.67

Total Hospital Costs

$11,249 ⫾ 20,783

Total Hospital Reimbursement

$13,365 ⫾ 22,699

$29,316 ⫾ 22,596

Hospital Margin

$ 2,101 ⫾ 9,900

$

Total Hospital Costs

$15,491 ⫾ 26,070

$8,912

0.004

Total Hospital Reimbursement

$17,945 ⫾ 27,878

$28,597 ⫾ 31,851

$10,651

0.001

Hospital Margin

$ 2,436 ⫾ 10,995

$ 4,184 ⫾ 15,027

$1,748

Total Hospital Costs

$28,804 ⫾ 63,062

Total Hospital Reimbursement

$32,830 ⫾ 74,996

Hospital Margin

$ 3,976 ⫾ 24,340

Total Hospital Costs

$91,973 ⫾ 201,209

Total Hospital Reimbursement

$83,728 ⫾ 166,135

Hospital Margin

-$ 8,299 ⫾ 65,113

387

1603

768

$24,403 ⫾ 30,080

$51,903 ⫾ 76,232

77

$23,099

0.003

$18,872

0.04

218 ⫾ 24,481

-$4,194

0.16

$82,849 ⫾ 73,788

74

0.18

$51,702 ⫾ 73,145 -$ 96

729 ⫾ 9,962

15

-$9,124

0.86

$93,652 ⫾ 115,334

-$9,924

0.82

$10,762 ⫾ 57,053

$19,061

0.29

84. HOW MUCH MORE COSTLY IS SURGERY IN THE ELDERLY? M. L. McGory, D. S. Zingmond, S. Jain, N. M. Foster, M. J. Leonardi, M. A. Maggard, C. Y. Ko; David Geffen School of Medicine, University of California, Los Angeles, CA Introduction: The population is aging and the elderly (ⱖ 65 years) are increasingly undergoing surgery. While little is known about the most common surgical procedures, even less is known about the outcomes that have important implications regarding resource utilization. The aging population will likely require more healthcare resources (length of stay [LOS], readmissions), including potentially uncompensated surgeon time. Methods: Using the California Patient Discharge Database, all elderly patients ⱖ 65 years of age who underwent a procedure in 2000 at all non-federal California hospitals were identified. Using ICD-9 codes, all procedures were classified by surgical specialty. The most common surgical procedures were identified by surgical specialty and acuity of admission (elective versus emergent). The outcomes of LOS and readmission were compared between the elderly patients ⱖ 65 years and a younger cohort of patients aged 50-64 years. Results: A total of 495,623 elective and 968,698 emergent surgical and non-surgical (e.g. angioplasty) procedures were performed in the elderly during the year 2000. The three most common surgical procedures identified were coronary artery bypass graft (CABG), cholecystectomy, and colectomy. A total of 12,327 CABG, 10,965 cholecystectomy, and 8,990 colectomy operations were performed in elderly California patients in 2000. The surgical workload in the elderly ranges from 160% to 250% higher for these 3 procedures in comparison to the 50-64 age group. On average, the LOS was 2.7 days longer (126% increase) for elderly undergoing emergent cholecystectomy or emergent CABG in comparison to a younger cohort. LOS was also significantly longer for elective CABG (1.7 days, 125% increase), elective colectomy (1.0 day, 114% increase), and elective cholecystectomy (0.8 days, 122% increase). This corresponds to an additional 56,000 postoperative inpatient days costing $56 million (estimated $1,000/day) for elderly in comparison to a younger cohort. Readmission rates within 90 days after surgery were also higher in the elderly population. For example, the readmission rate was 19% for emergent cholecystectomy in patients ⱖ 65 years and 12% in the 50-64 year cohort

(158% increase). The overall cost of these three procedures in the elderly is 125% more compared to the younger group. Conclusions: The aging population has significant implications for healthcare policy including increased costs for the health care system, increased resource utilization of inpatient beds for the hospital, and increased time spent for perioperative surgical care for the surgeon. As Medicare and other payers are trying to curb expenditures, these data suggest utilization and workload are probably increasing. 85. A COST ANALYSIS OF INTRA-OPERATIVE X-RAY SCREENING FOR RETAINED SURGICAL FOREIGN BODIES. L. Devgan, H. Waters, P. J. Pronovost, M. A. Makary; Center For Surgical Outcomes Research, Johns Hopkins University School of Medicine, Baltimore, MD Background: Retained foreign bodies following surgery can be catastrophic for patients, healthcare providers, and hospital systems. To address this increasingly high-profile problem, some institutions have begun routinely screening all surgical patients with intra-operative radiographs. Little is known, however, about the cost-effectiveness of this strategy. Methods: We reviewed the surgical literature to estimate the aggregate probability of a retained foreign body being present, detectible by radiograph, and clinically harmful. In order to determine the financial burden of intraoperative x-ray screening, we used an economic model of cost analysis based on the societal perspective, which considers patients, providers, and health care payers. To approximate costs, we calculated the explicit and implicit physical, personnel, and operating room costs of screening, using estimates in the literature. To predict savings, we used published data to account for the medical, legal, and hospital costs that screening would prevent. Results: The net cost per intra-operative screening x-ray performed is approximately $450. The probability of detecting a clinically harmful foreign body using radiographic screening is approximately 3.9 in 100,000. As such, we determined that routine intra-operative x-ray screening of all surgical patients costs approximately $11.5 million for every clinically significant sequelae detected. Conclusions: The use of routine intra-operative x-rays as a screening tool for surgical retained foreign bodies is extremely costly. Radiographic screening may be more cost-effective if employed preferentially in situations where a high index of suspicion for a retained foreign body exists. 86. A SURGEON REPORT CARD SYSTEM DEVELOPED USING A STANDARDIZED CLASSIFICATION OF ADVERSE OUTCOMES, ERROR PROFILE ANALYSIS AND CONCURRENT MORBIDITY AND MORTALITY REVIEW: A THREE-YEAR EXPERIENCE. A. C. ANTONACCI 1, S. Lam 2, V. Lavarias 3, R. D. Eavey 4; 1Weill Medical College of Cornell University, Nyc, Ny and Christ Hospital, Jersey City, Nj, 2Lenox Hill Hospital, Nyc, Ny, 3Beth Israel Medical Center, Nyc, Ny, 4 Pediatric Otolaryngology Service, Department of Otolaryngology, Massachusetts Eye and Ear Infirmary, Department of Otology and Laryngology, Harvard Medical School, Boston, Ma Introduction: Adverse outcomes traditionally are addressed individually rather than by accumulating the data for actuarial effectiveness. The traditional approach does not allow for systematic analysis, potential predictive capability, and focused resource management. Over three years we optimized the Department of Surgery Quality Improvement and Management (DSQIM) relational database to analyze patient adverse outcomes following surgery. This system was used to refine a standardized taxonomy of relevant adverse outcomes, to produce an error analysis system and to develop a meaningful quality profile for Attending Surgeon performance. Methods: Data was gathered from 29,237 operative procedures to establish a systematic list of adverse outcomes specific to both the hospital and the surgeon. Cases were analyzed at Morbidity and Mortality Conference according to a specific error profiling sys-