Society Of Thoracic Surgeons Risk Model Research Articles

Outcomes of Isolated Tricuspid Valve Surgery: A Society of Thoracic Surgeons Analysis and Risk Model

BackgroundTo provide patients and surgeons with clinically relevant information, The Society of Thoracic Surgeons (STS) Adult Cardiac Surgery Database was queried to develop a risk model for isolated tricuspid valve (TV) operations. MethodsAll patients in the STS Adult Cardiac Surgery Database who had undergone isolated TV repair or replacement (N = 13,587; age 48.3 ± 18.4 years) were identified (July 2017 to June 2023). Multivariable logistic regression accounting for TV replacement vs repair was used to model 8 operative outcomes: mortality, morbidity or mortality or both, stroke, renal failure, reoperation, prolonged ventilation, short hospital stay, and prolonged hospital stay. Model discrimination (C-statistic) and calibration were assessed using 9-fold cross-validation. ResultsThe isolated TV study population included 41.1% repairs (N = 5,583; age 52.6 ± 18.1 years) and 58.9% replacements (N = 8,004; age 45.3 ± 18.0 years). The overall predicted risk of operative mortality was 5.6%, and it was similar in TV repairs and replacements (5.5% and 5.7%, respectively), as was the predicted risk of composite morbidity and mortality (28.2% and 26.8%). TV replacements were generally performed in younger patients with a higher endocarditis prevalence than TV repairs (45.7% vs 21.1%). The model yielded a C-statistic of 0.81 for mortality and 0.76 for the composite of morbidity and mortality, with excellent observed-to-expected calibration that was comparable in all subcohorts and predicted risk decile groups. ConclusionsAn STS risk model has been developed for isolated TV surgery. The current mortality of isolated TV operations is lower than previously observed. This risk prediction model and these contemporary outcomes provide a new benchmark for current and future isolated TV interventions.

Machine Learning in Cardiac Surgery: Predicting Mortality and Readmission.

Predicting outcomes in open-heart surgery can be challenging. Unexpected readmissions, long hospital stays, and mortality have economic implications. In this study, we investigated machine learning (ML) performance in data visualization and predicting patient outcomes associated with open-heart surgery. We evaluated 8,947 patients who underwent cardiac surgery from April 2006 to January 2018. Data visualization and classification were performed at cohort-level and patient-level using clustering, correlation matrix, and seven different predictive models for predicting three outcomes ("Discharged," "Died," and "Readmitted") at binary level. Cross-validation was used to train and test each dataset with the application of hyperparameter optimization and data imputation techniques. Machine learning showed promising performance for predicting mortality (AUC 0.83 ± 0.03) and readmission (AUC 0.75 ± 0.035). The cohort-level analysis revealed that ML performance is comparable to the Society of Thoracic Surgeons (STS) risk model even with limited number of samples ( e.g. , less than 3,000 samples for ML versus more than 100,000 samples for the STS risk models). With all cases (8,947 samples, referred as patient-level analysis), ML showed comparable performance to what has been reported for the STS models. However, we acknowledge that it remains unknown at this stage as to how the model might perform outside the institution and does not in any way constitute a comparison of the performance of the internal model with the STS model. Our study demonstrates a systematic application of ML in analyzing and predicting outcomes after open-heart surgery. The predictive utility of ML in cardiac surgery and clinical implications of the results are highlighted.

Calibration Factors for STS Risk Model Predictions: Why, How and When They Are Used

The Society of Thoracic Surgeons (STS) Adult Cardiac Surgery Database is the world's premier adult cardiac surgery outcomes registry. This tutorial explains the following: how STS updates the risk models that are used to calculate the predicted risks of adverse events in the registry; why STS on a quarterly basis adjusts or "calibrates" the observed-to-expected ratios to equal 1 (O/E= 1), thereby effectively making the annual number of adverse events predicted by the model match the annual number of adverse events observed in the entire registry; the differences between the calibrated and uncalibrated O/E ratios; and how and when to use each.

STS Adult Cardiac Surgery Database: 2021 Update on Outcomes, Quality, and Research

The Society of Thoracic Surgeons Adult Cardiac Surgery Database is the most mature and comprehensive cardiac surgery database. It has been the foundation for quality measurement and improvement activities in cardiac surgery, facilitated the generation of accurate risk adjusted performance benchmarks and serves as a platform for novel research. Recent enhancements have added to the database's functionality, ease of use, and value to multiple stakeholders. This report is the sixth in a series of annual reports that provide updated volumes, outcomes, database-related developments, quality improvement initiatives, and research summaries using the Adult Cardiac Surgery Database in the past year.

The Performance of the Current Risk Prediction Scoring Systems in Patients Undergoing Anaortic Off-pump Coronary Artery Bypass Grafting.

BackgroundThis study aimed to evaluate the performance of the Society of Thoracic Surgeons (STS) risk model and European System for Cardiac Operative Risk Evaluation (EuroSCORE) II for patients undergoing anaortic off-pump coronary artery bypass grafting (OPCAB).MethodsFrom January 2010 to June 2017, 1,048 patients (isolated anaortic OPCAB: 1,043, on-pump conversion:5) undergoing isolated anaortic OPCAB were enrolled. The STS risk score and EuroSCORE II were calculated with dedicated online software. Calibration of the models were performed by the risk-adjusted event ratio that was defined as observed events divided by expected events (O/E ratio) and Hosmer-Lemeshow test. The discrimination powers were evaluated by the area under the receiver operating characteristic curve (AUC).ResultsOperative mortality occurred in 10 patients (0.95%). The predicted mortality rates by the EuroSCORE II and STS risk model were 2.58 ± 4.15% and 1.72 ± 2.92%, respectively. The O/E ratio of the EuroSCORE II was 0.370 with significant overprediction of operative mortality (confidence interval [CI], 0.157–0.652; P = 0.003). The STS score also overpredicted the operative mortality (O/E ratio, 0.556) with marginal significance (CI, 0.266–1.023; P = 0.052). Permanent stroke occurred in 6 patients (0.53%). The predicted permanent stroke occurrence rate was 1.73 ± 1.48%. The O/E ratio was 0.332 with significant overprediction of permanent stroke (CI, 0.121–0.722; P = 0.011). Regarding discrimination power for the STS risk model, the operative mortality was excellent (AUC, 0.876) and permanent stroke was acceptable (AUC, 0.740). The EuroSCORE II showed good discrimination power (AUC, 0.784). There was a significant difference in discrimination power for mortality between STS and EuroSCORE II risk models (P = 0.007).ConclusionPreexisting risk predicting scoring systems, STS risk model and EuroSCORE II, overpredict the risk of mortality and stroke rate for anaortic OPCAB. These findings suggest the possibility that anaortic OPCAB can lower the operative mortality and occurrence of postoperative stroke than conventional coronary artery bypass grafting.

Prediction of perioperative mortality in cardiac surgery – Risk scoring systems: In quest of the holy grail

Prediction of perioperative mortality in cardiac surgery – Risk scoring systems: In quest of the holy grail

Prognostic performance of the Society of Thoracic Surgeons risk score in patients with left main coronary artery disease undergoing revascularisation: a post hoc analysis of the EXCEL trial.

Accurate risk prediction in patients undergoing revascularisation is essential. We aimed to assess the predictive performance of Society of Thoracic Surgeons (STS) risk models in patients with left main coronary artery disease (LMCAD) undergoing coronary artery bypass grafting (CABG) or percutaneous coronary intervention with everolimus-eluting stents (PCI-EES). The predictive performance of STS risk models for perioperative mortality, stroke and renal failure was evaluated for their discriminative ability (C statistic) and calibration (Hosmer-Lemeshow goodness-of-fit-test; χ2 and p-values) among patients with LMCAD undergoing PCI-EES (n=935) and CABG (n=923) from the randomised EXCEL trial. STS risk scores, in CABG patients, showed good discrimination for 30-day mortality and average discrimination for stroke (C statistic 0.730 and 0.629, respectively) with average calibration. For PCI, STS risk scores had no discrimination for mortality (C statistic 0.507), yet good discrimination (C statistic 0.751) and calibration for stroke. The predictive performance for renal failure was good for CABG (C statistic 0.82), yet poor for PCI (C statistic 0.59). In selected patients with LMCAD from the EXCEL trial, STS risk models showed good predictive performance for CABG yet lacked predictive performance for PCI for perioperative mortality and renal failure. The STS stroke risk model was surprisingly more discriminating in PCI compared to CABG. Improved and procedure-specific risk prediction instruments are needed to accurately estimate adverse events after LMCAD revascularisation by CABG and PCI. ClinicalTrials.gov Identifier: NCT01205776.

P5983Pre-operative physical performance as an independent predictor of in-hospital outcomes in older patients undergoing elective cardiac surgery

Abstract Introduction Risk stratification of patients candidate to cardiac surgery is usually based on the Society of Thoracic Surgeons (STS) score or on the Euroscore II. However, these risk scores has limited predictive value in elderly patients. Purpose We conducted a study to determine whether the Short Physical Performance Battery (SPPB), a tool assessing physical performance, predicts hospital death and major morbidity, beyond STS risk score. The outcome was a composite end-point as defined by STS Major Morbidity or Operative Mortality (STS-MM) in STS Risk Model Outcomes: operative mortality, stroke, renal failure, prolonged mechanical ventilation, deep sternal wound infection, and reoperation. Methods In this prospective, single-center, cohort, hospital-based study, conducted at Careggi University Hospital, Florence, Italy, all patients aged 75+ years referred for an elective coronary aortic by-pass grafting, valvular surgery or combined cardiac surgery were evaluated pre-operatively. Assessment included SPPB, cognitive and functional status and evaluation of comorbidity. Patients receiving emergency/urgent cardiac surgery or a procedure not considered in the STS risk score calculator, who reported previous cardiac surgery or were clinically unstable were excluded. Participants were classified according to the STS-Predicted Risk Of Mortality (STS-PROM) as at low (<4%), intermediate (4 to 8%) or high risk (>8%). Results Out of 250 participants (females: 48.4%; mean age: 79.9 years), 148 (59.2%) were at low, 73 (29.2%) at intermediate and 29 (11.6%) at high risk, based on the STS-PROM. Mean±SEM SPPB score was 8.8±0.2, 7.1±0.4 and 6.0±0.7 in participants at low, intermediate, and high risk, respectively (p<0.001). The outcome occurred in 67 subjects (26.8%). SPPB score predicted the outcome in the entire study sample, controlling for STS-MM score, CKD, and anaemia (adjusted OR: 0.89, 95% CI 0.81–0.98 per each point increase; p 0.017). When analyses were repeated separately in participants with STS-PROM indicative of low (<4%) and in those at medium-high risk, the predictive value of SPPB was enhanced in the formers (adjusted OR 0.73, 95% CI 0.62–0.86 per each point increase; p<0.001- controlling for STS score), whereas was lost in the latter. An alternative model was subsequently tested, where STS score was not included and peripheral artery disease and creatinine (variables contributing to the STS score) were individually entered: SPPB was confirmed as a significant predictor also in this model, controlling for anaemia, peripheral artery disease, and creatinine. The corresponding AUC was 0.813 (Figure 1). Figure 1 Conclusions SPPB predicts mortality and major morbidity in older patients undergoing elective cardiac surgery, in particular in those classified as low-risk with the STS risk score. Use of SPPB should therefore be recommended to improve preoperative risk stratification of older patients.

Current quality reporting methods are not adequate for salvage cardiac operations

ObjectivesOutcomes in cardiac surgery are benchmarked against national Society of Thoracic Surgeons (STS) data and include patients undergoing elective, urgent, emergent, and salvage operations. This practice relies on accurate risk adjustment to avoid risk-averse behavior. We hypothesize that the STS risk calculator does not adequately characterize the risk of salvage operations because of their heterogeneity and infrequent occurrence. MethodsData on all cardiac surgery patients with an STS predicted risk score (2002-2017) were extracted from a regional database of 19 cardiac surgery centers. Patients were stratified according to operative status for univariate analysis. Observed-to-expected (O:E) ratios for mortality and composite morbidity/mortality were calculated and compared among elective, urgent, emergent, and salvage patients. ResultsA total of 76,498 patients met inclusion criteria. The O:E mortality ratios for elective, urgent, and emergent cases were 0.96, 0.98, and 0.93, respectively (all P values > .05). However, mortality rate was significantly higher than expected for salvage patients (O:E ratio, 1.41; P = .04). Composite morbidity/mortality rate was lower than expected in elective (O:E ratio, 0.81; P = .0001) and urgent (O:E ratio, 0.93; P = .0001) cases but higher for emergent (O:E ratio, 1.13; P = .0006) and salvage (O:E ratio, 1.24; P = .01). O:E ratios for salvage mortality were highly variable among each of the 19 centers. ConclusionsThe current STS risk models do not adequately predict outcomes for salvage cardiac surgery patients. On the basis of these results, we recommend more detailed reporting of salvage outcomes to avoid risk aversion in these potentially life-saving operations.

The Influence of the Charlson Comorbidity Index on Procedural Characteristics, VARC-2 Endpoints and 30-Day Mortality Among Patients Who Undergo Transcatheter Aortic Valve Implantation

Aortic stenosis (AS) is a common valvular abnormality and transcatheter aortic valve implantation (TAVI) is being increasingly used to treat patients considered too high risk for conventional surgery. We aimed to assess the prevalence of comorbid conditions in patients undergoing TAVI using the Charlson Comorbidity Index (CCI) and to assess their impact on clinical and procedural outcomes. We analysed 158 patients who underwent a TAVI at our institution between June 2009 and September 2015 to define their co-morbid burden as measured with CCI, and study its impact on procedural characteristics and mortality at 30 days. One hundred fifty-eight (158) patients with a mean age of 82±8years and a mean CCI score of 2.67 underwent a TAVI. Only 12/158 patients had a CCI of 0. The commonest cardiovascular comorbidities were previous myocardial infarction (24%), congestive heart failure (15%) and diabetes mellitus (23%) whilst the commonest non-cardiovascular comorbidities were renal disease (46%) and chronic obstructive pulmonary disease (COPD) (29%). After multivariable adjustment, CCI was not independently associated with adverse clinical outcomes. The addition of CCI to scoring systems such as Logistic EuroScore (LES) and Society of Thoracic Surgeons (STS) risk models improved the area under the curve from 0.75 (95%CI: 0.44-1.00) and 0.83 (95%CI: 0.64-1.00) to 0.78 (95%CI: 0.53-1.00) and 0.89 (95%CI: 0.78-1.00) respectively. The burden of comorbid conditions in patients undergoing TAVI is significant. The CCI score was not independently associated with a higher risk of death but can be useful in addition to LES and STS risk models in informing decision making on the selection of patients for TAVI.

The Society of Thoracic Surgeons 2018 Adult Cardiac Surgery Risk Models: Part 2—Statistical Methods and Results

The Society of Thoracic Surgeons (STS) uses statistical models to create risk-adjusted performance metrics for Adult Cardiac Surgery Database (ACSD) participants. Because of temporal changes in patient characteristics and outcomes, evolution of surgical practice, and additional risk factors available in recent ACSD versions, completely new risk models have been developed. Using July 2011 to June 2014 ACSD data,risk models were developed for operative mortality, stroke, renal failure, prolonged ventilation, mediastinitis/deep sternal wound infection, reoperation, major morbidity or mortality composite, prolonged postoperative length of stay, and short postoperative length of stay among patients who underwent isolated coronary artery bypass grafting surgery (n= 439,092), aortic or mitral valve surgery (n= 150,150), or combined valve plus coronary artery bypass grafting surgery (n=81,588). Separate models were developed for each procedure and endpoint except mediastinitis/deep sternal wound infection, which was analyzed in acombined model because of its infrequency. A surgeon panel selected predictors by assessing model performance and clinical face validity of full and progressively more parsimonious models. The ACSD data (July 2014 to December 2016) were used to assess model calibration and to compare discrimination with previous STS risk models. Calibration in the validation sample was excellent for all models except mediastinitis/deepsternal wound infection, which slightly underestimated risk and will be recalibrated in feedback reports. The c-indices of new models exceeded thoseofthe last published STS models for all populations and endpoints except stroke in valve patients. New STS ACSD risk models have generally excellent calibration and discrimination and are well suited for risk adjustment of STS performance metrics.

Invited Commentary

Invited Commentary

The Society of Thoracic Surgeons Mitral Valve Repair/Replacement Plus Coronary Artery Bypass Grafting Composite Score: A Report of The Society of Thoracic Surgeons Quality Measurement Task Force

The Society of Thoracic Surgeons (STS) Quality Measurement Task Force has developed a composite performance measure for mitral repair/replacement (MVRR) with concomitant coronary artery bypass grafting (CABG). Data were acquired from the STS Adult Cardiac Surgery Database for 26,463 patients undergoing MVRR+ CABG operations between July 1, 2011, and June 30, 2014. Established STS risk models were applied, along with modifications enabling the inclusion of patients with concomitant closures of atrial septal defects and patent foramen ovale, surgical ablation for atrial fibrillation, and tricuspid valve repair (TVR). Participants with fewer than 10 eligible cases over 3 years were excluded. The MVRR+ CABG composite consisted of two domains: risk-adjusted mortality and the any-or-none occurrence of major morbidity (prolonged ventilation, deep sternal infection, permanent stroke, renal failure, and reoperation). Composite performance scores were calculated with the use of hierarchic regression models, and high-performing and low-performing outliers were determined with the use of 95% Bayesian credible intervals. There were 24,740 patients at 703 participant sites after exclusions. Two percent (14/703) of programs were classified as 1-star (lower than expected performance), 95% (666/703) were classified as 2-star (as-expected performance), and 3% (23/703) were classified as 3-star (higher than expected performance). The average unadjusted operative mortality was 6.2% (1,532/24,740), and a monotonic decline in both mortality and morbidity was observed as star rating scores increased. An STS composite performance measure was developed for MVRR+ CABG operations. This measure may be useful for outcome assessment, quality improvement, patient counseling, clinical research, and public reporting.

Abstract 12044: East Meets West in the Assessment of Cardiac Risk: A Multicenter Comparative Study of the Influence of Racial, Ethnic and Cultural Risk Factors on Cardiac Surgical Risk Model Performance

Introduction: Numerous risk models have been applied to coronary artery bypass (CABG) patients. Their applicability across racial, ethnic and cultural backgrounds has not been well-described. We therefore compared the performance of the SinoSCORE, derived from 43 Chinese medical centers (c-statistic=0.80), with the Society of Thoracic Surgeons (STS) CABG mortality risk model in a population of CABG patients from a diverse multicenter American registry. Hypothesis: Racial, ethnic and cultural differences will impact the performance of surgical risk models. Methods: Perioperative data on all patients undergoing CABG between May 2006 and December 2015 from each of 8 centers was entered into the STS database. SinoSCORE and STS predicted risk of mortality was calculated for each patient. Logistic regression was used to develop a risk model based on SinoSCORE variables and performance was compared to STS risk model. Results: 7918 patients underwent CABG of whom 1.36% % were Asian. Sino-SCORE successfully categorized surgical mortality into low (0.12%), medium (0.36%) and high (2.42%) risk of mortality. However, the discrimination and calibration was better for STS than for the derived SinoSCORE (c-statistic 0.8087 vs. 0.767; p=0.0007; Figure; Hosemer-Lemeshow χ 2 52.64 vs. 40.05). Moreover, several of the factors which were predictors of mortality in the original SinoSCORE (age 65-70, BMI<18, chronic renal insufficiency, extracardiac arteriopathy, COPD, atrial fibrillation, urgent status) were not predictors in the derived model when applied to this multicenter Western patient population Conclusions: A CABG risk model derived from a from a broad sampling of Chinese patients did not perform as well in a diverse multicenter American population as it did in the population of origin, or as did the STS risk model. Considerable caution should be exercised, even for common operations, in applying surgical risk models across racial, ethnic and cultural divisions.

50th Anniversary Landmark Commentary on Edwards FH, Clark RE, Schwartz M. Coronary artery bypass grafting: The Society of Thoracic Surgeons National Database experience. Ann Thorac Surg 1994;57:12–9

50th Anniversary Landmark Commentary on Edwards FH, Clark RE, Schwartz M. Coronary artery bypass grafting: The Society of Thoracic Surgeons National Database experience. Ann Thorac Surg 1994;57:12–9

Do pulmonary function tests improve risk stratification before cardiothoracic surgery?

ObjectiveTo assess the added value of pulmonary function tests (PFTs) and different classifications of chronic obstructive pulmonary disease (COPD) to the Society of Thoracic Surgeons (STS) risk model using a clinical definition of lung disease for predicting outcomes after cardiothoracic (CT) surgery. MethodsWe evaluated consecutive patients who underwent nonemergency cardiac surgery and underwent PFTs before CT surgery. We used the STS risk model 2.73 to estimate the postoperative risk for respiratory failure (RF; defined as the need for mechanical ventilation for ≥72 hours, or reintubation), prolonged postoperative stay (PPLS; defined as >14 days), and 30-day all-cause mortality. We plotted the receiver operating characteristics curve for STS score for each adverse event, and compared the resulting area under the curve (AUC) with the AUC after adding PFT parameters and COPD classifications. ResultsOf the 1412 patients with a calculated STS score, 751 underwent PFTs. The AUC of the STS score was 0.65 (95% confidence interval [CI], 0.55-0.74) for RF, 0.67 (95% CI, 0.6-0.74) for prolonged postoperative length of stay (PPLS), and 0.74 (95% CI, 0.6-0.87) for death. None of the PFT parameters or COPD classifications added to the predictive ability of STS for RF, PPLS, or 30-day mortality. ConclusionsAdding individual PFT parameters or different COPD classifications to STS score calculated using clinically based classification of lung disease did not improve model discrimination. Thus, routine preoperative PFTS may have limited clinical utility in patients undergoing CT surgery when the STS score is readily available.

Potential Impact of Modifiable Clinical Variables on Length of Stay After First-Time Cardiac Surgery

Recent financial challenges highlight the importance of accurate prediction of length of hospital stay (LOS). We assessed reliability of The Society of Thoracic Surgeons (STS) risk prediction for extended and shorter LOS and examined whether modifiable clinical variables are associated with LOS in first-time cardiac surgery patients. Isolated aortic valve, mitral valve, and coronary artery bypass graft surgery patients since 2008 were included (n = 3,472). Multivariate regression was used to evaluate nonmodifiable and potentially modifiable (preoperative hematocrit, hemoglobin A1c, body mass index, current smoker, major perioperative morbidity, and blood product transfusion) predictors of LOS in days. Mean age was 63.9 ± 11.2 years, 76% were males, and mean STS mortality risk was 1.9% ± 3.2%. Median (interquartile range) LOS was 4 (3 to 6) days. Predicted STS risk was 6.2% ± 7.1% for extended LOS (>14 days) and 48.3% ± 20.2% for short LOS (<6 days). Extended LOS was observed in 5.2% of patients (observed versus expected, 0.84; p = 0.019). Observed short LOS was better than predicted (67.8%; observed versus expected, 1.40; p < 0.001). Inclusion of modifiable variables in the LOS prediction model was significant (p < 0.001). Significant modifiable predictors were lower hematocrit, higher hemoglobin A1c, major morbidity, and transfusion. Longer predicted LOS from the model correlated with longer actual LOS (rs = 0.63; p < 0.001). Applying the prediction equation from the model to a hypothetical average patient, predicted LOS was 4.6 days. The STS risk model was reliably predictive of short and extended LOS but did not allow prediction of exact LOS in days. Accounting for potentially modifiable clinical variables, such as low hematocrit and blood transfusion, especially in elective patients, should lead to shorter LOS, higher satisfaction, and reduced financial burden.

The STS AVR + CABG Composite Score: A Report of the STS Quality Measurement Task Force

The Society of Thoracic Surgeons (STS) is developing a portfolio of composite performance measures for the most commonly performed adult cardiac procedures. This manuscript describes the third composite measure in this series, aortic valve replacement (AVR) combined with coronary artery bypass grafting surgery (CABG). We identified all patients in the STS Adult Cardiac Surgery Database who underwent AVR+CABG during recent 3-year (July 1, 2009, through June 30, 2012) and 5-year (July 1, 2007, through June 30, 2012) periods. Variables from the STS risk model for AVR+CABG were used to adjust morbidity and mortality outcomes. Evidence for internal mammary artery use in AVR+CABG was examined. We compared composite measures constructed using 3 or 5 years of outcomes with Bayesian credible intervals of 90%, 95%, or 98%. The final STS AVR+CABG composite performance measure is based on 3 years of data and 95% credible intervals. It includes risk-adjusted mortality and morbidity but not internal mammary artery use. Median composite score is 91.0% (interquartile range, 89.5% to 92.2%). There were 2.6% (24 of 915) one-star (lower performing) and 6.5% (59 of 915) three-star (higher performing) programs. Morbidity and mortality decrease monotonically as star ratings increase. The percentage of three-star programs increased substantially among programs that performed more than 150 procedures over 3 years compared with those performing 25 to 50 procedures (32.8% versus 1.6 %). Measure reliability was 0.51. The STS has developed a composite performance measure for AVR+CABG based on 3-year data samples and 95% credible intervals. This composite measure identified 9.1% of STS participants as having higher or lower than expected performance.

Comparison of original EuroSCORE, EuroSCORE II and STS risk models in a Turkish cardiac surgical cohort†

The aim of this study was to compare additive and logistic European System for Cardiac Operative Risk Evaluation (EuroSCORE), EuroSCORE II and the Society of Thoracic Surgeons (STS) models in calculating mortality risk in a Turkish cardiac surgical population. The current patient population consisted of 428 patients who underwent isolated coronary artery bypass grafting (CABG) between 2004 and 2012, extracted from the TurkoSCORE database. Observed and predicted mortalities were compared for the additive/logistic EuroSCORE, EuroSCORE II and STS risk calculator. The area under the receiver operating characteristics curve (AUC) values were calculated for these models to compare predictive power. The mean patient age was 74.5 ± 3.9 years at the time of surgery, and 35.0% were female. For the entire cohort, actual hospital mortality was 7.9% (n = 34; 95% confidence interval [CI] 5.4-10.5). However, the additive EuroSCORE-predicted mortality was 6.4% (P = 0.23 vs observed; 95% CI 6.2-6.6), logistic EuroSCORE-predicted mortality was 7.9% (P = 0.98 vs observed; 95% CI 7.3-8.6), EuroSCORE II- predicted mortality was 1.7% (P = 0.00 vs observed; 95% CI 1.6-1.8) and STS predicted mortality was 5.8% (P = 0.10 vs observed; 95% CI 5.4-6.2). The mean predictive performance of the analysed models for the entire cohort was fair, with 0.7 (95% CI 0.60-0.79). AUC values for additive EuroSCORE, logistic EuroSCORE, EuroSCORE II and STS risk calculator were 0.70 (95% CI 0.60-0.79), 0.70 (95% CI 0.59-0.80), 0.72 (95% CI 0.62-0.81) and 0.62 (95% CI 0.51-0.73), respectively. EuroSCORE II significantly underestimated mortality risk for Turkish cardiac patients, whereas additive and logistic EuroSCORE and STS risk calculators were well calibrated.

Prediction of in-hospital mortality following pulmonary resections: improving on current risk models†

Using a large, prospectively collected and independently validated thoracic database, we created a risk-prediction tool for in-hospital mortality with the aim of improving on the accuracy of Thoracoscore. A prospectively collected and independently validated database containing lung resections was utilized, N = 2574. Logistic regression analysis with bootstrapping, and by the use of a random training and test set was utilized. Comparisons against the Thoracoscore, ESOS.01 and the Society of Thoracic Surgeons (STS) models were performed. A logistic model identified age [odds ratio (OR) 1.1, 95% confidence interval (CI) 1.0-1.2, P = 0.0002], sex (OR 0.34, 95% CI 0.14-0.83, P = 0.02), predicted postoperative FEV1 (OR 0.96, 95% CI 0.94-0.99, P = 0.002), emphysema (OR 3.2, 95% CI 1.0-9.9, P = 0.04), excess alcohol consumption (OR 1.0, 95% CI 1.0-1.0, P = 0.04), pre-existing renal disease (OR 4.3, 95% CI 1.1-17.1, P = 0.04), predicted in-hospital mortality with an receiver operating curve (ROC) of 0.81 and a Hosmer-Lemeshow test of 0.9. Bootstrap analysis confirmed the above risk factors (ROC 0.82 and Hosmer-Lemeshow 0.2). Comparisons between Thoracoscore, ESOS.01 and the STS risk models demonstrated that none was very accurate, as all had low ROC values of 0.69, 0.70 and 0.61, respectively. The STS risk model does not apply to our population (ROC 0.61, Hosmer-Lemeshow, P = 0.004), and the ESOS.01 has poor predictive power (Hosmer-Lemeshow, P < 0.0001). Logistic regression based on age, sex, predicted postoperative FEV1, alcohol consumption and pre-existing renal disease predicts in-hospital mortality with improved accuracy compared with the use of Thoracoscore, ESOS.01 and the STS risk model.