Concordance Between Registry and Administrative Data in the Determination of Comorbidity: A Multi-institutional Study.

OBJECTIVE: To characterize agreement between administrative and registry data in the determination of patient-level comorbidities. BACKGROUND: Previous research finds poor agreement between these 2 types of data in the determination of outcomes. We hypothesized that concordance between administrative and registry data would also be poor. METHODS: A cohort of inpatient operations (length of stay 1 day or greater) was obtained from a consortium of 8 hospitals. Within each hospital, National Surgical Quality Improvement Program (NSQIP) data were merged with intra-institutional inpatient administrative data. Twelve different comorbidities (diabetes, hypertension, congestive heart failure, hemodialysis-dependence, cancer diagnosis, chronic obstructive pulmonary disease, ascites, sepsis, smoking, steroid, congestive heart failure, acute renal failure, and dyspnea) were analyzed in terms of agreement between administrative and NSQIP data. RESULTS: Forty-one thousand four hundred thirty-two inpatient surgical hospitalizations were analyzed in this study. Concordance (Cohen Kappa value) between the 2 data sources varied from 0.79 (diabetes) to 0.02 (dyspnea). Hospital variation in concordance (intersite variation) was quantified using a test of homogeneity. This test found significant intersite variation at a level of P < 0.001 for each of the comorbidities except for dialysis (P = 0.07) and acute renal failure (P = 0.19). These findings imply significant differences between hospitals in their generation of comorbidity data. CONCLUSION: This study finds significant differences in how administrative versus registry data assess patient-level comorbidity. These differences are of concern to patients, payers, and providers, each of which had a stake in the integrity of these data. Standardized definitions of comorbidity and periodic audits are necessary to ensure data accuracy and minimize bias.

Gastrointestinal Discontinuity After Emergency Laparotomy.

During laparotomy, patients requiring intestinal resection may be temporarily left in gastrointestinal discontinuity (GID). We performed this study to determine predictors of futility for patients initially left in GID after emergency bowel resection. We divided the patients into 3 groups: never restored continuity and died (group 1), restored continuity and died (group 2), and restored continuity and survived (group 3). We compared the 3 groups for differences in demographics, acuity at presentation, hospital course, laboratory data, comorbidities, and outcomes. From a total of 120 patients, 58 patients died and 62 survived. We identified 31 patients in group 1, 27 patients in group 2, and 62 patients in group 3. On multivariate logistic regression, only lactate (P = .002) and use of vasopressors (P = .014) remained significant to predict survival. The results of this study can be used to identify futile situations which can direct end-of-life decisions.

Examining the Factors Influencing Applicants' Placement on One General Surgery Program's Rank Order List.

PURPOSE: Surgery residency applications include variables that determine an individual's rank on a program's match list. We performed this study to determine which residency application variables are the most impactful in creating our program's rank order list. METHODS: We completed a retrospective examination of all interviewed applicants for the 2019 match. We recorded United States Medical Licensing Examinations (USMLE) step I and II scores, class quartile rank from the Medical Student Performance Evaluation (MSPE), Alpha Omega Alpha (AOA) membership, geographic region, surgery clerkship grade, and grades on other clerkships. The MSPE and letters of recommendation were reviewed by two of the authors and assigned a score of 1 to 3, where 1 was weak and 3 was strong. The same two authors reviewed the assessments from each applicant's interview and assigned a score from 1-5, where 1 was poor and 5 was excellent. Univariate analysis was performed, and the significant variables were used to construct an adjusted multivariate model with significance measured at P < .05. RESULTS: Univariate analysis for all 92 interviewed applicants demonstrated that USMLE step 2 scores (P = .002), class quartile rank (P = .004), AOA status (P = .014), geographic location (P < .001), letters of recommendation (P < .001), and interview rating (P < .001) were significant in predicting an applicant's position on the rank list. On multivariate analysis only USMLE step 2 (P = .018) and interview (P < .001) remained significant. CONCLUSION: USMLE step 2 and an excellent interview were the most important factors in constructing our rank order list. Applicants with a demonstrated strong clinical fund of knowledge that develop a rapport with our faculty and residents receive the highest level of consideration for our program.

Institutional Analysis of Metaplastic Breast Cancer.

Triple-negative breast cancer (TNBC) is typically managed with neoadjuvant chemotherapy (NAC). Metaplastic breast cancer (MBC), a subtype of TNBC, demonstrates different histologic characteristics and less responsiveness to NAC. We performed this study to achieve a better understanding of MBC, including the impact of neoadjuvant chemotherapy. We identified patients diagnosed with MBC from January 2012 to July 1, 2022. A control group of TNBC breast cancer patients from 2020 who did not meet the criteria for MBC was identified. Demographic data, tumor and nodal characteristics, management strategies employed, response to systemic chemotherapy, and treatment outcomes were recorded and compared between groups. A total of 22 patients were included in the MBC group and demonstrated a 20% response to NAC compared to an 85% response rate in the 42 patients in the TNBC group (P = .003). Five patients have recurred (23%) in the MBC group compared to none in the TNBC group (P = .013).

Multidisciplinary Breast Cancer Clinic: A Novel Approach to Decrease Time to Intervention.

Multidisciplinary clinics (MDCs) were created at high-volume surgical oncology centers to optimize breast cancer care, in which patients are seen by multiple subspecialists at one visit. We aim to evaluate our experience with this novel approach. We examined 492 patients with newly diagnosed invasive breast cancer from January 1, 2020, to September 1, 2022. Patients seen at our MDC had a decreased time to intervention across all measured intervals: 3 days faster (10 vs 13 days) from biopsy to clinic visit [t Sat 2.09 > t critical two tail 1.99], 5 days faster (23 vs 28 days) from diagnosis to initiation of neoadjuvant chemotherapy [t Sat 5.12 > t critical two tail 2.019], and 21 days faster (24 vs 45 days) from surgery clinic visit to operation [t Sat 5.12 > t critical two tail 2.019]. Although early in our experience, we have initiated a strategy for improved breast cancer care.

The Surgeon's Risk of SARS-CoV-2 Infection During the Initial Peak of the COVID-19 Pandemic in New Orleans.

Surgeons who care for patients with active SARS-CoV-2 infection represent a unique population of health care providers whose risk of infection has not been elucidated. The objective of this study was to examine SARS-CoV-2 seroprevalence among surgeons who cared for patients with active SARS-CoV-2 infection compared to other employees within our health care system and also the general public of New Orleans. 105 surgeons at our facilities provided direct surgical care to patients with active SARS-CoV-2 infection and underwent voluntary antibody testing. 2/105 (1.9% CI .2%-6.7%) tested positive for SARS-CoV-2 antibodies. 13 343 hospital employees underwent antibody testing and 1066/13 343 (8.0% CI 7.5%-8.5%) tested positive (1.9% vs. 8.0%; P = .03). We saw a significantly lower SARS-CoV-2 seroprevalence among surgeons who directly cared for infected patients versus other hospital employees. When compared to community seroprevalence (6.9% CI 6.0%-8.0%), seroprevalence among our surgeons is also significantly lower (1.9% vs. 6.9%; P = .04).

An Analysis of Outcomes and Management Strategies for Patients With Cholecystostomy Tubes.

BACKGROUND: Percutaneous cholecystostomy tube (PCT) drainage is an effective management strategy for acute cholecystitis in patients medically unfit for surgery. However, little is known about the fate of patients managed by PCT. We conducted this study to determine tube management outcomes for patients with acute cholecystitis managed by PCT. METHODS: The electronic record was queried to identify patients with acute cholecystitis managed by PCT from 2012-2020. Patients were divided into three groups for analysis: 1) ultimately managed by cholecystectomy, 2) eventual confirmation of distal flow of bile from the gallbladder and tube removal, and 3) tubes left in place without further management. RESULTS: A total of 179 patients with acute cholecystitis treated by PCT were included. Sixty-six patients never fully recovered from the medical insult associated with their diagnosis of acute cholecystitis and had their tubes left in situ. Sixty-four of these 66 patients (97%) died during follow-up. The remaining 113 patients recovered from their illness and presented to clinic for evaluation for tube removal and/or cholecystectomy. When distal biliary flow was confirmed, tube removal was favored (n = 70). When cystic duct outflow occlusion persisted, cholecystectomy was planned for patients who became acceptable surgical candidates (n = 43). For patients managed by cholecystectomy, 8 were approached open and 35 laparoscopically, with 12 of 35 (34.3%) converted to open and 23 (65.7%) completed laparoscopically. CONCLUSION: Our study favors PCT removal for patients who recover from their acute illness when distal bile flow from the gallbladder is confirmed. We reserve cholecystectomy for patients who recover from their illness and demonstrate persistent cystic duct outflow obstruction.

An Analysis of the Timing for Closure of a Diverting Loop Ileostomy.

A diverting loop ileostomy (DLI) is used to protect a distal gastrointestinal anastomosis at risk of leakage. While patients typically prefer early DLI closure, surgeons vary in opinion regarding optimal timing. This study evaluated whether the timing of DLI closure impacts outcomes.A retrospective review was performed on patients who underwent DLI creation within one health care system between 2012 and 2020. Patient characteristics and postoperative outcomes were compared across ileostomies closed in ≤2 months, 2-4 months, and >4 months. Outcomes examined included anastomotic leak, other complications, reintervention, and death within 30 days.A total of 500 DLIs were analyzed for the study, 455 of which were closed. The three closure groups were similar in patient characteristics and comorbidities. None of the outcome variables analyzed in this study demonstrated a statistically significant difference between groups, suggesting that in patients otherwise fit for surgery, DLI closure can be safely performed within 2 months of creation.

Lessons Learned From an Early Hotspot During the COVID-19 Pandemic.

The Covid-19 pandemic has provided challenges for surgical residency programs demanding fluid decision making focused on providing care for our patients, maintaining an educational environment, and protecting the well-being of our residents. This brief report summarizes the impact of the impact on our residency programs clinical care and education. We have identified opportunities to improve our program using videoconferencing, managing recruitment, and maintaining a satisfactory caseload to ensure the highest possible quality of surgical education.

History of the Georgia Baptist/Atlanta Medical Center surgical residency.

The Georgia Baptist Hospital established itself as a premier healthcare facility during the first 50 years of the 20th century. The surgical residency started in the 1940s, became accredited in 1958, and has grown into one of the most respected independent programs in the country. The development and growth of the program was a result of the commitment and dedication of the Program Directors in Surgery over the past 50 years. These key leaders included A. Hamblin Letton, John P. Wilson, Paul Stanton, and George Lucas. The hospital's name has changed to Atlanta Medical Center with the sale of the hospital to Tenet in 1997. The same old school approach to surgical training that characterized the residency when it was known as Georgia Baptist persists and provides outstanding training for future surgeons interested in a broadly based surgical education and experience.

Analysis of demographic and tumor characteristics of an inner city breast cancer patient population compared with patients treated in National Surgical Adjuvant Breast and Bowel Project trials.

In June of 2008 we initiated a breast clinic designed to serve patients regardless of funding status. We analyzed age, race,tumor size, nodal status, estrogen, progesterone, and her-2-neu status. We compared our results to NSABP B-06 (nodal status), B-15 (estrogen, progesterone, and Her-2-neu receptor status), B-18, and B-27 (age, race, and tumor size) to determine whether our patient population was similar to patients included in these trials. Forty-nine patients with newly diagnosed breast cancer were treated during our first year (53 total cancers). Eight patients had noninvasive cancer; 45 had invasive disease. The mean age was 52.2 +/- 12.2 years compared to a mean age of 48.4 +/- 9.8 years in the B-06 trial (P = 0.005). Thirty six patients were African American (74%) compared to 10% and 12% in the NSABP B-18 and B-27 trials (P < 0.00001). A total of 23 of our patients with invasive cancer had involved axillary lymph nodes which was statistically more common than the 35.3% of node positive patients in the B-06 trial (P = 0.03). Tumor size (3.6 +/- 3.3 cm), estrogen (54.4%), and progesterone (52.8%) receptor status were similar to NSABP trials. Only 6 (13.3%) of our patients were considered Her-2-neu positive compared to 29.4% in the B-15 trial which was significantly less prevalent (P = 0.02). Significantly different demographic and tumor characteristics were identified in our inner city breast cancer patient population compared to NSABP patients. These results question the validity of using recommendations from large cooperative group trials in the development of treatment plans for our inner city patient population.

Transition from chief residency to specialty training: issues and solutions.

Three fourths of chief residents in general surgery receive further specialty training. The end to start-of-year transition can create administrative conflicts between the residency and the specialty training program. An Internet-based questionnaire surveyed general surgery and surgical specialty program directors to define issues and possible solutions associated with end to start-of-year transitions using a Likert scale. There was an overall response rate of 17.5 per cent, 19.6 per cent among general surgery directors, and 15.8 per cent among specialty directors. Program directors in general surgery felt strongly that the transition is an administrative problem (P < 0.001). They opposed extra days off at the end of the chief resident year or ending in mid-June, which specialty directors favored (P < 0.001). Directors of specialty programs opposed starting the year 1 or 2 weeks after July 1, a solution that general surgery directors favored (P < 0.001). More agreement was reached on whether chief residents should take vacation week(s) at the end of the academic year, having all general surgery levels start in mid-June, and orientation programs in July for specialty trainees. Program directors acknowledge that year-end scheduling transitions create administrative and patient care problems. Advancing the start of the training year in mid-June for all general surgery levels is a potential solution.

Euglycemic Diabetic Ketoacidosis in the Surgical Patient.

Postoperative euglycemic diabetic ketoacidosis (EDKA), a rare cause of acidosis, results from the metabolic derangement of diabetes and is not associated with a surgical complication requiring reoperation. Our acute care surgery service has managed several recent patients who developed postoperative EDKA. Our group was befuddled by the initial case but subsequently quickly recognized and managed the condition. The purpose of this report is to discuss the pathophysiology of EDKA, summarize 3 recent cases, and increase awareness about the condition to permit prompt recognition and treatment.

Factors That Predict Biological Aggressiveness in Estrogen Receptor-Positive / Human Epidermal Growth Factor Receptor 2-Negative / Lymph Node-Negative Breast Cancer.

Background: Traditionally, breast cancer is staged using TNM criteria: tumor size (T), nodal status (N), and metastasis (M). The Oncotype DX assay provides a recurrence score (RS) based on genomics that predicts the likelihood of distant recurrence in estrogen receptor-positive (ER+)/human epidermal growth factor receptor 2-negative (HER2-)/lymph node-negative (LN-) tumors. Methods: We retrospectively reviewed the medical records of patients with ER+/HER2-/LN- breast cancer tumors who were evaluated between 2007 and 2017 with Oncotype DX RS. We compared the RS to tumor size, patient age, progesterone receptor (PR) status, and LN immunohistochemistry to assess for factors that may independently predict recurrence risk. We also compared tumor size to tumor grade. Results: The data set included 296 tumors: 248 ER+/PR-positive (PR+)/HER2- and 48 ER+/PR-negative (PR-)/HER2-. RS ranged from 0 to 66, patient age ranged from 33 to 77 years, and tumor size ranged from 1 to 65 mm. No significant correlation was found between age and RS (r=-0.073, P=0.208). PR- tumors had a significantly higher RS regardless of size (PR- mean RS 30.8 ± 12.7; PR+ mean RS 16.3 ± 7.3; t(53)=7.6, P<0.0001). No significant correlation was seen between tumor size and RS for all tumors (r=-0.028, P=0.635), and this finding remained true for the PR+ tumor subgroup (r=0.114, P=0.072). However, a significant negative correlation was seen between tumor size and RS in the PR- subgroup (r=-0.343, P=0.017). Further analysis to ensure that differences in tumor grade did not account for this correlation showed equal distribution of well differentiated, moderately differentiated, and poorly differentiated tumors with no significant correlation between tumor size and grade. Conclusion: Increasing tumor size may not be associated with increasing biological aggressiveness. Traditionally, smaller tumors are thought to be lower risk and larger tumors higher risk, with a tendency to use chemotherapy with large tumors. However, our data showed a negative correlation between tumor size and RS in the PR- subgroup. A tumor with PR negativity that reaches a large size without metastasizing may suggest a favorable tumor biology. These tumors may not receive as much benefit from chemotherapy as previously thought. Also, the higher RS seen in smaller PR- tumors may demonstrate PR- status as a predictor for higher risk of distant recurrence. We propose that all tumors meeting the ER+/PR-/LN- criteria, regardless of size, should be considered for genotyping, with the RS used to guide chemotherapy benefit.

Use of Postmastectomy Radiation in Patients Treated for Invasive Breast Cancer with Mastectomy and Immediate Reconstruction.

Immediate reconstruction after mastectomy helps women manage the psychological impact of deforming surgery. Postmastectomy radiation therapy (PMRT) can negatively impact the aesthetic result after breast reconstruction. We performed this study to achieve a better understanding of how PMRT is used after reconstruction in our institution. We conducted a retrospective review of a prospectively maintained database of all women who underwent mastectomy for invasive breast cancer followed by immediate reconstruction from 2006 to 2017. Patients were divided into two groups depending on whether PMRT was included in their treatment, and we compared clinical and pathologic characteristics to determine which factors were likely to lead to PMRT. A total of 315 women treated with mastectomy and immediate reconstruction were identified. A total of 96 were treated with PMRT; 219 had mastectomy and immediate reconstruction without radiotherapy. Tumor characteristics, tumor stage, demographics, and comorbidities did not predict the use of PMRT. Neoadjuvant chemotherapy (NAC) was the most powerful predictor for using PMRT. In 47 of 81 (58%) patients treated with NAC, PMRT was used. Whereas 49 of 234 (21%) patients who did not receive NAC were treated with PMRT (P = 0.0001, risk ratio 2.77, 95 per cent confidence interval 2.03-3.77). In our institution, patients treated with NAC followed by mastectomy and immediate reconstruction are significantly more likely to receive PMRT. The increased use of PMRT after NAC should be factored into the preoperative discussion with patients choosing mastectomy and immediate reconstruction.

The role of CT angiography in the diagnosis of blunt traumatic thoracic aortic disruption and unsuspected carotid artery injury.

We have replaced aortography and open thoracic surgery to diagnose and treat blunt traumatic thoracic aortic disruption (TTAD) in favor of CT angiography (CTA) and endovascular repair. The purpose of this study is to review our experience with the management and outcomes of TTAD and associated carotid artery injuries. In January 2003, we initiated a protocol that used CTA to evaluate all patients with suspected TTAD from blunt trauma. When TTAD was diagnosed, patients were managed by endovascular repair using abdominal aortic extension cuffs. Twenty-nine patients with TTAD were managed by endovascular repair. In all patients, abdominal endograft extension cuffs successfully excluded the traumatic disruptions. Six (21%) of these patients had concomitant, unsuspected carotid artery injury diagnosed by CTA. One patient had bilateral carotid artery dissections, sustained irreversible brain injury, and died. Four patients with common carotid dissections were successfully treated by anticoagulation and made uneventful recoveries. One patient with a common carotid-innominate artery dissection and pseudoaneurysm underwent endovascular repair. This study indicates that CTA and endovascular repair provide accurate diagnostic and therapeutic results in the management of blunt TTAD. Furthermore, CTA should include arch and cervical views to detect an unsuspected, concomitant carotid artery injury.

Four-Year Analysis of a Novel Milestone-Based Assessment of Faculty by General Surgical Residents.

OBJECTIVE: In response to our faculty's concerns about the quality and reliability of feedback from general surgery residents, we developed a novel faculty assessment tool. This study was designed as an interim analysis of the tool's effectiveness and discriminatory ability. METHODS: Our department's educational leadership developed milestones in 7 domains that were scored from 1 to 4, with each level representing an educational approach that ranged from ineffective (1) to ideal (4). Each postgraduate year (PGY) class meets annually to develop a consensus regarding each faculty member's effectiveness in each of the 7 domains: (1) operative supervision, (2) operative teaching, (3) clinic and/or hospital supervision, (4) clinic and/or hospital teaching, (5) conference participation, (6) availability, and (7) overall contribution to the training program. We reviewed the results from the initial 4 years of this project. We also analyzed the annual national faculty survey administered by the Accreditation Council for Graduate Medical Education (ACGME) to evaluate faculty satisfaction regarding feedback during the same study period. Data were assessed using the Levene test for homogeneity, analysis of variance, and Wilcoxon-Mann-Whitney tests. RESULTS: Forty-two faculty members were annually evaluated by 29 to 32 residents. Each resident PGY class assigned faculty milestone scores that varied across the 7 domains, demonstrating that faculty scores reflected variable opinions about each specific domain, while avoiding labeling an effective faculty member with all high scores and a less effective member with all poor scores.(p < 0.0001). Milestone scores for a given faculty member differed across PGY classes, indicating that junior residents might evaluate a specific faculty member differently than senior residents (p < 0.0001). Eleven faculty members received low scores of 1 or 2 on the overall contribution to training domain and 8/11 (73%) improved to 3 or 4, the following year. Twenty core faculty members were included on the annual ACGME survey. The results from the study period on the ACGME anonymous faculty survey reflected enhanced satisfaction with resident feedback during the study period, improving from 68% to 88% compliance with ACGME standards and our mean program score improved from 4.1 to 4.4 compared to the national mean of 4.3 (p = 0.02). CONCLUSIONS: This milestone-based faculty assessment tool improves the quality of the feedback from surgical residents when evaluating faculty. When residents assign a negative statement to describe faculty educational effectiveness in a specific domain, performance improves. A milestone-based faculty assessment strategy should be explored on a national level.

Learning preferences of surgery residents: a multi-institutional study.

BACKGROUND: The VARK model categorizes learners by preferences for 4 modalities: visual, aural, read/write, and kinesthetic. Previous single-institution studies found that VARK preferences are associated with academic performance. This multi-institutional study was conducted to test the hypothesis that the VARK learning preferences of residents differ from the general population and that they are associated with performance on the American Board of Surgery In-Training Examination (ABSITE). METHODS: The VARK inventory was administered to residents at 5 general surgery programs. The distribution of the VARK preferences of residents was compared with the general population. ABSITE results were analyzed for associations with VARK preferences. χ2, Analysis of variance, and multiple linear regression were used for statistical analysis. RESULTS: A total of 132 residents completed the VARK inventory. The distribution of the VARK preferences of residents was different than the general population (P < .001). The number of aural responses on the VARK inventory was an independent predictor of ABSITE percentile rank (P = .03), percent of questions correct (P = .01), and standard score (P = .01). CONCLUSION: This study represents the first multi-institutional study to examine VARK preferences among surgery residents. The distribution of preferences among residents was different than that of the general population. Residents with a greater number of aural responses on VARK had greater ABSITE scores. The VARK model may have potential to improve learning efficiency among residents.

Venous Thromboembolism after Inpatient Surgery in Administrative Data vs NSQIP: A Multi-Institutional Study.

BACKGROUND: Previous studies have documented significant differences between administrative data and registry data in the determination of postoperative venous thromboembolism (VTE). The goal of this study was to characterize the discordance between administrative and registry data in the determination of postoperative VTE. STUDY DESIGN: This study was performed using data from the American College of Surgeons NSQIP merged with administrative data from 8 different hospitals (5 different medical centers) between 2013 and 2015. Occurrences of postoperative vein thrombosis (VT) and pulmonary embolism (PE) as ascertained by administrative data and NSQIP data were compared. In each situation where the 2 sources disagreed (discordance), a 2-clinician chart review was performed to characterize the reasons for discordance. RESULTS: The cohort used for analysis included 43,336 patients, of which 53.3% were female and the mean age was 59.5 years. Concordance between administrative and NSQIP data was worse for VT (κ 0.57; 95% CI 0.51 to 0.62) than for PE (κ 0.83; 95% CI 0.78 to 0.89). A total of 136 cases of discordance were noted in the assessment of VT; of these, 50 (37%) were explained by differences in the criteria used by administrative vs NSQIP systems. In the assessment of postoperative PE, administrative data had a higher accuracy than NSQIP data (odds ratio for accuracy 2.86; 95% CI 1.11 to 7.14) when compared with the 2-clinician chart review. CONCLUSIONS: This study identifies significant problems in ability of both NSQIP and administrative data to assess postoperative VT/PE. Administrative data functioned more accurately than NSQIP data in the identification of postoperative PE. The mechanisms used to translate VTE measurement into quality improvement should be standardized and improved.

Postoperative Myocardial Infarction in Administrative Data vs Clinical Registry: A Multi-Institutional Study.

BACKGROUND: Previous studies have documented significant differences between administrative data and registry data in the determination of postoperative MI. The goal of this study was to characterize discordance between administrative and registry data in the determination of postoperative myocardial infarction (MI). STUDY DESIGN: This study was performed using data from the American College of Surgeons NSQIP merged with administrative data from 8 different hospitals, between 2013 and 2015. From each of these sources, the occurrence of a postoperative MI, as ascertained by administrative data and NSQIP data, were compared. In each situation in which the 2 sources disagreed (discordance), a 2-clinician chart review was performed to generate a "gold standard" determination as to the occurrence of postoperative MI. RESULTS: A total of 43,289 operations met our inclusion criteria for analysis. Within this cohort a total of 230 cases of MI were identified by administrative data and/or NSQIP data (administrative rate 0.41%, NSQIP rate 0.42%). A total of 89 discordant ascertainments were identified, of which 42 were admin+/NSQIP- and 47 were admin-/NSQIP+. Accuracy (99.9% for both) and concordance (kappa = 0.89 [95% CI 0.86 to 0.92] for administrative data, kappa = 0.87 [95% CI 0.84 to 0.91] for NSQIP data) of the 2 systems were similar when compared against our gold standard (chart review). The majority of errors were related to false negatives, with sensitivity rates of 81% in both data sources. CONCLUSIONS: In this multi-institutional study, administrative data and NSQIP demonstrated a similar ability to determine the occurrence of postoperative MI. These findings do not demonstrate an advantage of registry data over administrative data in the determination of postoperative MI.