Assessment of validity for EuroSCORE risk stratification system

Prognostic scoring helps doctors, patients and their families to weigh the risks and benefits of medical care and clarifies their expectations.

Objective We aimed to analyze the risk stratification performance of the EuroSCORE system because of its common use in Lithuania.

Design EuroSCORE performance is assessed in terms of sensitivity, specificity, accuracy and area under the receiver operating characteristic (ROC) curve (AUC). Logistic regression is carried out for modeling categorical data and odds ratio calculations of being a non-survivor case for each EuroSCORE risk group.

Results The study was completed on 1002 patients. Mean score for EuroSCORE was 4.77±2.8; ROC curve of 0.71; accuracy was 65.5%; 65.4% sensitivity and 67.2% specificity.

Conclusions EuroSCORE created a moderately predictive area under the ROC curve for our patient population. Probability of non-survival by logistic regression model for each EuroSCORE risk group is statistically significantly higher compared to the lower risk group. Predictions available from prognostic scoring systems could be useful in decision making when there is uncertainty in whether to carry out surgery or not.     

EuroSCORE overestimates the cardiac operative risk

INTRODUCTION: It was the purpose of our study to assess the validity of EuroSCORE (European system for cardiac operative risk evaluation) in our patient population. MATERIALS AND METHODS: Between March 1999 and August 2001, information on risk factors and mortality was collected for 1123 consecutive adult patients undergoing heart surgery with cardiopulmonary bypass. EuroSCORE was used for risk stratification. Mean age +/- standard deviation was 58.6 +/- 10.9 and 29.1% of the patients were female. The area under the receiver operating characteristic (ROC) curve was calculated as an index for the predictive value of the scoring system. RESULTS: The area under the ROC curve was 0.824 for all patients and 0.828 for the isolated CABG subgroup which shows an excellent predictive ability.When the scoring system was applied in low, medium, and high risk groups, there was no overlap between 95% confidence intervals of observed and expected mortality in all three groups both for the isolated CABG cases and for all patients. Decreased left ventricular ejection fraction, emergent operation, and preoperative unstable angina requiring i.v. nitrate treatment were significant predictive variables for early mortality. CONCLUSION: EuroSCORE is a simple and objective system for predicting the risk of heart surgery. The predictive power of the EuroSCORE is excellent, however it seems that mortality is considerably overestimated by this score.     

Early mortality in coronary bypass surgery: the EuroSCORE versus The Society of Thoracic Surgeons risk algorithm

BACKGROUND: We compare two widely used risk algorithms for coronary bypass surgery: The European System for Cardiac Operative Risk Evaluation (EuroSCORE) and The Society of Thoracic Surgeons (STS) risk stratification algorithm. METHODS: Risk factors for all adult patients undergoing heart surgery at the University Hospital of Lund between 1996 and 2001 were collected prospectively at preoperative admission. Predictive accuracy for 30-day mortality was assessed by comparing the observed and the expected mortality for equal-sized quintiles of risk by using the Hosmer-Lemeshow goodness-of-fit test. The discriminatory power was evaluated by calculating the areas under receiver operating characteristics (ROC) curves. RESULTS: The study included 4497 coronary artery bypass-only operations. The average age was 66.4 +/- 9.3 years (range 31 to 90 years). Most patients were men (77.0% versus 23.0%). The actual 30-day mortality was 1.89%. The Hosmer-Lemeshow goodness-of-fit test gave a p value of 0.81 (EuroSCORE) and 0.83 (STS), which indicates a good accuracy of both models. The area under the ROC curve was 0.84 (95% confidence interval [CI] 0.80 to 0.88) for EuroSCORE and 0.71 (95% CI 0.66 to 0.77) for STS. The discriminatory power (area under the ROC curve) was significantly larger for EuroSCORE compared with STS (p < 0.00005). CONCLUSIONS: In this large, single institution study the additive EuroSCORE algorithm had a significantly better discriminatory power to predict 30-day mortality than the STS risk algorithm for patients undergoing coronary artery bypass.     

Single centre, single domain validation of the EuroSCORE on a consecutive sample of primary and repeat CABG

Objectives: Intra- and interdepartmental benchmarking require scoring systems with excellent performance on several properties: discrimination (resolution), reliability (calibration) and stability over the complete spectrum of peri-procedural risk. This single centre, single domain study validates the European system for cardiac operative risk evaluation (EuroSCORE) on an independent sample of primary and repeat coronary artery bypass grafting (CABG) patients and will evaluate these different properties. Methods: The study population is a consecutive series of 2051 isolated primary and repeat CABG patients, inclusive of patients in cardiogenic shock or resuscitation, operated on in a single institution from January 1997 to July 2000. The age of the patients was 66±9 years, 77% were males and 7% were repeat procedures. The EuroSCORE was 5.0±3%, with a range from 0 to 22. The studied event was in-hospital death, defined as mortality during hospital stay, which was unlimited in time and included a stay in a secondary hospital without discharge home. Results: The EuroSCORE predicted 102 deaths versus 81 deaths observed (P=0.14, Fisher exact test). The EuroSCORE described only 20% of the variance of in-hospital mortality. The EuroSCORE created an area under the receiver operating characteristic curve of 0.83±0.03. The highest discriminative accuracy was obtained with 8% EuroSCORE risk (only 64% sensitivity and 87% specificity). Further exploration identified an over score in the EuroSCORE range 0–8 (57%, P<0.0001). There was an equal score (−2%, P=1) in the range 9–11, but an under score in the range 12–22 (−133%, P=0.003). Conclusions: On the condition that these single centre results could be extended to any European cardiac surgery centre, it can be concluded that the overall acceptable performance of the EuroSCORE is the result of an over score in the lower risk and insufficient correction in the higher risk spectrum. The EuroSCORE is probably refined enough for improved informed consent versus aggregated results but should only be used for inter-institutional benchmarking with great caution, preferably below the 12% risk pivot.     

Validation of Euroscore model in an Australian patient population

BACKGROUND: The purpose of the present paper was to assess the performance of the European system for cardiac operative risk evaluation (EuroSCORE) model in an Australian adult cardiac surgical population. METHODS: The additive and logistic EuroSCORE models were retrospectively applied to predict operative mortality in 2106 consecutive patients undergoing cardiac surgery at St Vincent's Hospital, Melbourne between June 2001 and August 2003, and at Geelong Hospital between June 2001 and April 2004. The entire cohort and a subset of patients undergoing isolated coronary artery bypass graft (CABG) surgery were analysed. Model discrimination and calibration was tested by determining the area under the receiver operating characteristic (ROC) curve and Hosmer-Lemeshow chi2, respectively. RESULTS: There were significant differences in the prevalence of risk factors between the Australian and European cardiac surgical populations. There were 81 deaths (observed mortality 3.85%) in the entire cohort and 39 deaths in the isolated CABG group (observed mortality 2.60%). The EuroSCORE models overestimated mortality (entire cohort: additive predicted 5.75%, logistic predicted 9.93%; isolated CABG: additive predicted 4.87%, logistic predicted 7.71%). Discriminative power was very good for the entire cohort (area under ROC curve, 0.81 (additive) and 0.82 (logistic)). Calibration of both models was poor. CONCLUSION: The additive and logistic EuroSCORE model of risk prediction was not validated in the present population of cardiac surgical patients. The models may not accurately predict outcomes of patients undergoing cardiac surgery in Australia.     

Risk-stratification in thoracic aortic surgery: should the EuroSCORE be modified?

OBJECTIVES: Risk-stratification in cardiac surgical procedures is of major interest. Recent studies have shown, that the EuroSCORE is a very good and reliable risk-stratification score in CABG and in valve surgery. The aim of the study was to evaluate the EuroSCORE in patients undergoing surgery on the thoracic aorta. METHODS: Three hundred and sixty-seven consecutive patients underwent surgery of the thoracic aorta and were scored, according to the additive and logistic EuroSCORE algorithm. We compared correlation of predicted and observed mortality and evaluated a modification of the EuroSCORE in order to improve the scoring system. Score validity was assessed by calculating the area under the receiver operating characteristic curve (ROC). RESULTS: Overall hospital mortality was 10.1%. Additive EuroSCORE predicted mortality was 2.3% for 3-6% risk, 12.9% for 7-8% risk, 18.4% for 9-12% risk and 27.3% for a risk >12%. The modified score predicted mortality was 1% for 3-6% risk, 8.2% for 7-8% risk, 12.1% for 9-14% risk, 18.6% for 15-24% risk and 28.6% for a risk >24%. Area under the ROC-curve was 0.68 for the EuroScore and 0.91 in the modified score, 0.72 and 0.86 in the logistic model. CONCLUSIONS: The modified score, taking into account aortic dissection (6 points) and preoperative malperfusion (12 points) significantly improves the predictive value of the EuroSCORE in patients undergoing thoracic aortic surgery.     

Preoperative prediction of early mortality and morbidity in coronary bypass surgery

OBJECTIVE: A scoring system to predict early mortality and morbidity in CABG, distinguishing low and high risk patients. METHODS: 563 patients (1998) served as development dataset, 969 patients as validation set. Univariate and logistic regression analysis was used to identify risk factors. RESULTS: Gender, hypertension, pulmonary disease, reoperation, age, operative status and left-ventricular function were predictive variables for early mortality. The area under the ROC curve was 0.81. We identified a low risk, mortality of 1.8% and a high-risk group, mortality of 13.4%. Diabetes, hypertension, kidney and lung disease, reoperation, operative status and left ventricular function were predictive variables for morbidity. The area under the ROC curve was 0.73. We identified a low risk, morbidity of 17%, and a high-risk group, morbidity of 41%. CONCLUSION: This scoring system is a simple system identifying a low and high-risk group for morbidity and early mortality.     

Performance of three preoperative risk indices; CABDEAL, EuroSCORE and Cleveland models in a prospective coronary bypass database.

OBJECTIVES: The aim of the present study was to evaluate the performance of three different preoperative risk models in the prediction of postoperative morbidity and mortality in coronary artery bypass (CAB) surgery. METHODS: Data on 1132 consecutive CAB patients were prospectively collected, including preoperative risk factors and postoperative morbidity and in-hospital mortality. The preoperative risk models CABDEAL, EuroSCORE and Cleveland model were used to predict morbidity and mortality. A C statistic (receiver operating characteristic (ROC) curve) was used to test the discrimination of these models. RESULTS: The area under the ROC curve for morbidity was 0.772 for the CABDEAL, 0.694 for the EuroSCORE and 0.686 for the Cleveland model. Major morbidity due to postoperative complications occurred in 268 patients (23.6%). The mortality rate was 3.4% (n=38 patients). The ROC curve areas for prediction of mortality were 0.711 for the CABDEAL, 0.826 for the EuroSCORE and 0.858 for the Cleveland model. CONCLUSIONS: The CABDEAL model was initially developed for the prediction of major morbidity. Thus, it is not surprising that this model evinced the highest predictive value for increased morbidity in this database. Both the Cleveland and the EuroSCORE models were better predictive of mortality. These results have implications for the selection of risk indices for different purposes. The simple additive CABDEAL model can be used as a hand-held model for preoperative estimation of patients' risk of postoperative morbidity, while the EuroSCORE and Cleveland models are to be preferred for the prediction of mortality in a large patient sample.     

Enhancing Braden pressure ulcer risk assessment in acutely ill adult veterans

This study sought to determine if a parsimonious pressure ulcer (PU) predictive model could be identified specific to acute care to enhance the current PU risk assessment tool (Braden Scale) utilized within veteran facilities. Factors investigated include: diagnosis of gangrene, anemia, diabetes, malnutrition, osteomyelitis, pneumonia/pneumonitis, septicemia, candidiasis, bacterial skin infection, device/implant/graft complications, urinary tract infection, paralysis, senility, respiratory failure, acute renal failure, cerebrovascular accident, or congestive heart failure during hospitalization; patient's age, race, smoking status, history of previous PU, surgery, hours in surgery; length of hospitalization, and intensive care unit days. Retrospective chart review and logistic regression analyses were used to examine Braden scores and other risk factors in 213 acutely ill veterans in North Florida with (n = 100) and without (n = 113) incident PU from January–July 2008. Findings indicate four medical factors (malnutrition, pneumonia/pneumonitis, candidiasis, and surgery) have stronger predictive value (sensitivity 83%, specificity 72%, area under receiver operating characteristic [ROC] curve 0.82) for predicting PUs in acutely ill veterans than Braden Scale total scores alone (sensitivity 65%, specificity 70%, area under ROC curve 0.70). In addition, accounting for four medical factors plus two Braden subscores (activity and friction) demonstrates better overall model performance (sensitivity 80%, specificity 76%, area under ROC curve 0.88).     

Logistic or additive EuroSCORE for high-risk patients?

OBJECTIVES: To assess whether the use of the full logistic European System for Cardiac Operative Risk Evaluation (EuroSCORE) is superior to the standard additive EuroSCORE in predicting mortality in high-risk cardiac surgical patients. METHODS: Both the simple additive EuroSCORE and the full logistic EuroSCORE were applied to 14,799 cardiac surgical patients from across Europe, of whom there were 4293 high-risk patients (additive EuroSCORE of 6 or more). The systems were compared for absolute prediction and discrimination (area under the receiver operating characteristic (ROC) curve). RESULTS: Actual mortality was 4.72%. The logistic model was closer to this than the additive model (4.84% (4.72-4.94) versus 4.21 (4.21-4.26)). Most of this difference was due to high-risk patients where actual mortality was 11.18% and predicted was 7.83% (additive) and 11.23% (logistic). Discrimination was similar in both systems as measured by the area under the ROC curve (additive 0.783, logistic 0.785). CONCLUSIONS: The additive EuroSCORE model remains a simple "gold standard" for risk assessment in European cardiac surgery, usable at the bedside without complex calculations or information technology. The logistic model is a better risk predictor especially in high-risk patients and may be of interest to institutions engaged in the study and development of risk stratification.     

Preoperative risk stratification in infective endocarditis. Does the EuroSCORE model work? Preliminary results.

OBJECTIVE: There is an important role for risk prediction in cardiac surgery. Prediction models are useful in decision making and quality assurance. Patients with infective endocarditis (IE) have a particularly high risk of mortality. The aim was to assess the performance of European System for Cardiac Operative Risk Evaluation (EuroSCORE) in IE. METHODS: The additive and logistic EuroSCORE models were applied to all patients undergoing surgery for IE (Duke criteria) between January 1995 and April 2006 within our prospective institutional database. Observed and predicted mortalities were compared. Model calibration was assessed with the Hosmer-Lemeshow test. Model discrimination was tested by determining the area under the receiver operating characteristic (ROC) curve. RESULTS: One hundred and eighty-one consecutive patients undergoing 191 operations were analyzed. Observed mortality was 28.8%. For the entire cohort the mean additive score was 10.4 (additive predicted mortality of 14.2%). The mean logistic predicted mortality was 27.1%. Discriminative power was good for the additive and the logistic models for the entire series. Area under ROC curve were 0.83 (additive) and 0.84 (logistic) for the entire cohort, 0.81 and 0.81 for the aortic position, 0.91 and 0.92 for the mitral position, 0.81 and 0.81 for the native valve, 0.82 and 0.83 for the prosthetic valves, and 0.81 and 0.51 for the gram-positive microorganisms, respectively. CONCLUSIONS: This initial sample may be small; however, additive and logistic EuroSCORE adequately stratify risk in IE. Logistic EuroSCORE has been calibrated in IE, a special group of very high-risk patients. Further studies with larger sample sizes are required to confirm these initial results.     

The logistic EuroSCORE predicts the hospital mortality of the thoracic aortic surgery in consecutive 327 Japanese patients better than the additive EuroSCORE.

OBJECTIVE: Risk stratification in thoracic aortic surgery is a topic of major interest. Recent studies have shown the European System for Cardiac Operative Risk Evaluation (EuroSCORE) to be an extremely useful and reliable risk stratification score and also a good indicator of quality of care in cardiac surgery. The purpose of this study was to evaluate the significance of the additive and logistic EuroSCOREs in patients undergoing surgery on the thoracic aorta in Japan. METHODS: We calculated the predicted mortality according to the additive and logistic EuroSCORE algorithms in 327 consecutive patients who underwent surgery of the thoracic aorta during a 30-year period (between 1976 and 2005). We compared the score validity between the two algorithms and also evaluated the score validity for the patients who underwent thoracic aortic surgery. The score validity was assessed by calculating the area under the receiver operating characteristic (ROC) curve. RESULTS: The overall in-hospital mortality was 13%. The area under the ROC curve was satisfactorily high for the additive (0.68, 0.73, 0.73) as well as the logistic EuroSCORE (0.69, 0.74, 0.75) in the patients who underwent thoracic aortic surgery during 30-, 20-, and 10-year periods, respectively. The actual mortality was 7% (Group 1; an additive EuroSCORE of 3-6), 16% (Group 2; 7-11), and 37% (Group 3; >12). The mortality expected by the additive and logistic EuroSCORE in the three different risk groups were (5%, 9%, 19%) and (5%, 14%, 43%), respectively. Namely, the mortality expected by the logistic EuroSCORE perfectly matched with the actual mortality in any of the three risk groups. In contrast, the mortality expected by the additive EuroSCORE tended to dissociate when the number of risks increased. Significant difference was observed between the observed mortality and the mortality expected by the additive EuroSCORE algorithm in the high-risk group (p=0.0473). CONCLUSIONS: Although both the additive and the logistic EuroSCORE reliably predicted the overall operative mortality for thoracic aortic surgery in 327 Japanese patients, the logistic EuroSCORE better matched with the actual mortality in the operative risk especially in the high-risk group.     

基于机器学习算法预测甲状腺乳头状癌右喉返神经后方淋巴结转移907例临床研究

目的 构建基于机器学习算法的甲状腺乳头状癌右喉返神经后方淋巴结（LN-prRLN）转移预测模型，并验证其预测效果。方法 回顾性分析2014年3月至2019年7月在浙江大学医学院附属杭州市第一人民医院肿瘤外科接受手术的907例甲状腺乳头状癌病人。分别纳入性别、年龄、肿瘤大小、被膜侵犯、多灶性等临床病理资料。根据时间序列，分为训练组（n=595）和验证组（n=312）。运用Logistic回归及分类树、随机森林、梯度提升法、支持向量机等机器学习算法进行特征变量选择，并构建LN-prRLN转移的预测模型。结果 分类树模型的受试者操作特征曲线（ROC）曲线下面积为0.654，敏感度为33.00%，特异度为97.78%，准确率为86.89%；随机森林模型的ROC曲线下面积为0.753，敏感度为57.00%，特异度为100.00%，准确率为92.77%；支持向量机模型的ROC曲线下面积为0.604，敏感度为27.00%，特异度为83.19%，准确率为86.39%；梯度提升法的ROC曲线下面积为0.873，敏感度为72.00%，特异度为89.49%，准确率为87.90%。结论 LN-prRLN转移预测模型对甲状腺乳头状癌右喉返神经后方淋巴结转移具有良好的预测效果，其中梯度提升法具有较高的诊断效能。     

Validation of the EuroSCORE model in Australia.

OBJECTIVE: There is an important role for accurate risk prediction models in current cardiac surgical practice. Such models enable benchmarking and allow surgeons and institutions to compare outcomes in a meaningful way. They can also be useful in the areas of surgical decision-making, preoperative informed consent, quality assurance and healthcare management. The aim of this study was to assess the performance of the European System for Cardiac Operative Risk Evaluation (EuroSCORE) model on the Australasian Society of Cardiac and Thoracic Surgeons (ASCTS) patient database. METHODS: The additive and logistic EuroSCORE models were applied to all patients undergoing cardiac surgery at six institutions in the state of Victoria between 1st July 2001 and 4th July 2005 within the ASCTS database who have complete data. The entire cohort and a subgroup of patients undergoing coronary artery bypass grafting (CABG) only were analysed. Observed and predicted mortalities were compared. Model discrimination was tested by determining the area under the receiver operating characteristic (ROC) curve. Model calibration was tested by the Hosmer-Lemeshow chi-square test. RESULTS: Eight thousand three hundred and thirty-one patients with complete data were analysed. There were significant differences in the prevalence of risk factors between the ASCTS and European cardiac surgical populations. Observed mortality was 3.20% overall and 2.00% for the CABG only group. The EuroSCORE models over estimated mortality (entire cohort: additive predicted 5.31%, logistic predicted 8.76%; CABG only: additive predicted 4.25%, logistic predicted 6.19%). Discriminative power of both models was very good. Area under ROC curve was 0.83 overall and 0.82 for the CABG only group. Calibration of both models was poor as mortality was over predicted at nearly all risk deciles. Hosmer-Lemeshow chi-square test returned P-values less than 0.05. CONCLUSIONS: The additive and logistic EuroSCORE does not accurately predict outcomes in this group of cardiac surgery patients from six Australian institutions. Hence, the use of the EuroSCORE models for risk prediction may not be appropriate in Australia. A model, which accurately predicts outcomes in Australian cardiac surgical patients, is required.     

EuroSCORE Predicts Short- and Mid-Term Mortality in Combined Aortic Valve Replacement and Coronary Artery Bypass Patients

Abstract   Background and Aim of the Study: European system for cardiac operative risk evaluation (EuroSCORE) has been studied for its effectiveness in predicting operative mortality, and more recently, long-term mortality in a wide variety of cardiac surgical procedures. Combined coronary artery bypass and aortic valve replacement (AVR-CABG) carries increased perioperative risk, and tends to have higher-risk patients. Performance of the EuroSCORE system in patients undergoing concomitant AVR-CABG has not been well established. Thus, we aimed to analyze the accuracy of both additive and logistic EuroSCOREs in predicting operative and mid-term mortality. Methods: We retrospectively reviewed and calculated EuroSCOREs for all patients who underwent AVR-CABG between January 2000 and December 2004. Patients who had previous cardiac surgery and those undergoing any concomitant procedures were excluded. Areas under the receiver operator curves (ROC) were determined to assess EuroSCORE's accuracy in predicting operative mortality. Kaplan-Meier analysis and Cox regression were used to determine mid-term survival, freedom from repeat revascularization, and predictors of these outcomes. Results: There were 233 patients who met study criteria. Mean follow-up period was 2.2 ± 1.7 years with one patient lost to follow-up. Mean additive and logistic EuroSCOREs were 8.77 and 16.1, respectively, with an observed mortality of 9.44%. The area under the ROC curves for additive EuroSCORE was 0.76 and for logistic EuroSCORE was 0.75. Regression analysis revealed additive EuroSCORE, but not logistic EuroSCORE, to be predictive of mid-term mortality. Conclusions: Both additive and logistic EuroSCOREs were accurate in predicting operative morality. Only additive EuroSCORE was predictive of mid-term mortality in AVR-CABG patients. EuroSCORE remains a good and well-validated risk stratification model applicable to patients who undergo concomitant AVR-CABG.     

Role of dynamic contrast-enhanced magnetic resonance (MR) imaging and proton MR spectroscopic imaging in the detection of local recurrence after radical prostatectomy for prostate cancer

OBJECTIVES: To assess the accuracy of magnetic resonance (MR) spectroscopic imaging (1H-MRSI) and dynamic contrast-enhanced MR (DCEMR) in the depiction of local prostate cancer recurrence in patients with biochemical progression after radical prostatectomy (RP). MATERIALS AND METHODS: 1H-MRSI and DCEMR were performed in 70 patients at high risk of local recurrence after RP. The population was divided on the basis of the clinical validation of MR results with the use of a transrectal ultrasound biopsy examination in a group of 50 patients (group A) and the prostate-specific antigen (PSA) serum level restitution after external beam radiotherapy, in a group of 20 patients (group B). RESULTS: In group A, 1H-MRSI analysis alone showed a sensitivity of 84% and a specificity of 88%; the DCEMR analysis alone, a sensitivity of 71% and a specificity of 94%; combined 1HMRSI-DCEMR, a sensitivity of 87% and specificity of 94%. Areas under the receiver operating characteristic (ROC) curve for 1HMRSI, DCEMR, and combined 1HMRSI /DCEMR were 0.942, 0.93,1 and 0.964, respectively. In group B, 1HMRSI alone showed a sensitivity of 71% and a specificity of 83%; DCEMR, a sensitivity of 79% and a specificity of 100%; combined 1HMRSI and DCEMR, a sensitivity of 86% and a specificity of 100%. Areas under the ROC curve for each of these groups were 0.81, 0.923, and 0.94, respectively. CONCLUSION: Our results show that combined 1H-MRSI and DCMRE is an accurate method to identify local prostate cancer recurrence in patients with biochemical progression after RP.     

非体外循环冠状动脉旁路移植术后早期死亡风险预测——SinoSCORE与EuroSCORE评分的对比

目的 比较中国冠状动脉旁路移植手术评分系统(SinoSCORE)和欧洲心脏外科手术风险评分系统(EuroSCORE)对中国人群非体外循环冠状动脉旁路移植(OPCAB)术后早期死亡风险的预测价值.方法 OPCAB病人资料来自中国心血管外科注册登记研究2004-2005年数据库,观察终点为术后院内死亡.分别用SinoSCORE和logistic EuroSCORE两种模型计算病人预计病死率,并与实际病死率比较.校准度采用Hosmer-Lemeshow拟合优度检验,利用 ROC曲线下面积(AUC)评价模型的区分度.结果 4920例病人中73例发生院内死亡,实际病死率1.48%,SinoSCORE模型和EuroSCORE模型预测的病死率分别为2.73%、4.13%.SinoSCORE模型Hosmer-Lemeshow拟合优度检验P=0.636,AUC=0.794;EuroSCORE模型Hosmer-Lemeshowrny拟合优度检验P=0.01,AUC=0.756.SinoSCORE和logistic EuroSCORE两种模型的区分度均较好,但SinoSCORE的校准度明显优于后者,即两种模型均能预测术后死亡,但SinoSCORE对术后病死率的预测更加准确.结论 SinoSCORE模型比EuroSCORE模型更适用于中国非体外循环冠状动脉旁路移植病人术后早期死亡预测.     

Predictive Power and Implication of EuroSCORE,EuroSCORE II and STS Score for Isolated Repeated Aortic Valve Replacement

Objective: We evaluated the predictive power of the EuroSCORE, EuroSCORE II and Society of Thoracic Surgeons (STS) score for isolated redo aortic valve replacement.Materials and Methods: 78 consecutive patients underwent the aforementioned procedure mainly with a stentless valve prosthesis at our institution. Observed mortality was compared to the predicted mortality, Receiver Operating Characteristics (ROC) curves were calculated and the area under the curve (AUC) analyzed.Result: Observed mortality was 11.5%. EuroSCORE and EuroScore II predicted a mortality of 28.2 ± 21.6% (p <0.001) and 10.2 ± 11.8% (p = 0.75), respectively. AUC of the EuroSCORE was 0.74 (95% CI: 0.62–0.83), p = 0.009 and of the EuroSCORE II 0.86 (95% CI: 0.76–0.93), p <0.0001. Optimal Youden index of the EuroSCORE II was 0.59 refering to a predicted mortality of 9.9% (sensitivity: 77.8% and specificity: 81.2%). Predicted mortality of STS score was 17.8 ± 10.6% (p = 0.08) and AUC was 0.64 (95% CI: 0.53–0.75), p = 0.06.Conclusion: EuroSCORE II calculation was not only superior to EuroSCORE and STS score but led to a very realistic mortality prediction for this special procedure at our institution. A EuroSCORE II greater 10 should encourage to consider an alternative treatment.     

APACHE-II scoring system in perforative peritonitis

BACKGROUND: Prospective assessment of the Acute Physiology and Chronic Health Evaluation-II (APACHE-II) scoring system of stratification of disease severity has been shown to provide objective discrimination between low-risk and high-risk groups of patients with intra-abdominal sepsis. The current study was undertaken to evaluate the performance of APACHE-II score in prediction of mortality risk in patients with peritonitis due to hollow viscus perforation. STUDY DESIGN: Fifty patients admitted to a teaching hospital with peritonitis due to hollow viscus perforation were prospectively studied over a 2-year period. APACHE-II points were assigned to all patients in order to calculate their individual risk of mortality before undergoing emergency surgery. The accuracy in outcome prediction of the APACHE-II system was assessed by means of receiver operating characteristic (ROC) curve and the Pearson correlation coefficient and its significance test. RESULTS: Of the 50 patients admitted during the study period, there were 42 (84%) survivors and 8 (16%) nonsurvivors. Mean APACHE-II score of the study population was 11.38 with a range of 1 to 23. The predicted death rate was 23% and the observed death rate was 16%. Mean APACHE-II score in survivors was 9.88, whereas in nonsurvivors it was 19.25. Using ROC analysis, the area under the curve was found to be .984. Correlation of APACHE-II score and predicted death rate showed perfect correlation, with r = .99 and P <.001 [R2 = .9993]. APACHE-II score between 11 and 15 showed a sensitivity and specificity of 100% and 73.8%, respectively, and APACHE-II score of 16 to 20 had a sensitivity and specificity of 87.5% and 100%, respectively. CONCLUSION: APACHE-II score between 11 and 20 was shown to be a better predictor of risk of mortality in patients with peritonitis due to hollow viscus perforation. Predicted mortality did not correlate with observed mortality in patients with APACHE-II scores of 1 to 10 and greater than 20. The APACHE-II scoring system can be used to assess group outcomes in patients with peritonitis due to hollow viscus perforation. However, it does not provide sufficient confidence for outcome prediction in individual patients.     

Receiver operating characteristic curve analysis of clinical risk models

Receiver operating characteristic (ROC) curve analysis is a useful method to measure the ability of a clinical risk model to discriminate between hospital deaths and survivors. Its use in medicine originated as a method for synthesizing the specificity and sensitivity of diagnostic tests across a spectrum of possible cut points. The area under the ROC curve can be interpreted as a probability of correct classification or prediction. We illustrate its use in three steps: first, with a dichotomous variable to introduce specificity and sensitivity; next, with a categorical risk factor (surgical urgency) to produce a primitive ROC curve; and finally, with a continuous risk factor (age) to approximate the usual ROC curve used for clinical risk models.