Inaccuracy of four coronary surgery risk-adjusted models to predict mortality in individual patients.

OBJECTIVES: This study was undertaken to evaluate the accuracy of four different risk-adjusted models in predicting mortality in individual patients who are undergoing coronary artery by-pass graft surgery. In the last decade several models to stratify patients before open heart surgery, according to factors affecting mortality, were developed with the aim of retrospectively comparing outcomes of open heart surgery, based on reliable stratification of case-mix, and of prospectively identifying high risk patients as a basis for a meaningful informed consent for patients counseling. METHODS: The pre-operative risk of death was calculated with four different models in 418 consecutive patients who underwent coronary artery by-pass surgery and then compared with the actual outcome. To discriminate patients with favorable and unfavorable outcome, the logistic regression analysis and the areas under the receiver-operating-characteristic curves were applied. The accuracy score was used to evaluate the reliability of each score to predict the individual outcome. RESULTS: Seven deaths (1.7%) were observed within 30 days from the operation, and the overall incidence was similar to that predicted by all models. Only the NBI score was not able to discriminate survivors from patients who will die, and the areas under the curves were 0.596 for the Parsonnet score, 0.861 for the Cleveland Clinic Foundation score, 0.823 for the French score, and 0.806 for the EuroSCORE. The four models were highly accurate (between 0.97 and 0.98) to predict the overall mortality. In seven patients who died the mean predictive scores were very low and ranged between 2.1 and 4.6, but were significantly higher than those of patients who survived (between 1.1 and 2.2). CONCLUSIONS: The four pre-surgical predictive models were similarly able to discriminate favorable vs. unfavorable outcomes and highly accurate to predict overall mortality, but very inaccurate to predict mortality in individual patients.

Does EuroSCORE work in individual European countries?

OBJECTIVE: Because of national epidemiological differences in adult heart surgery in Europe, the effectiveness and desirability of a pan-European score for the assessment of quality of surgical care remains controversial. We assessed the predictive value of EuroSCORE in national subsets of the EuroSCORE database. METHODS: The EuroSCORE development data set was divided into national subsets of which those with 500 or more patients were selected for analysis. The Hosmer-Lemeshow goodness-of-fit test was applied to assess the calibration of the EuroSCORE model on individual national samples and the areas under the receiver operating characteristic (ROC) curve were measured to analyse the EuroSCORE discriminative power on individual death prediction. RESULTS: There were 18676 patients in the six largest national samples: Germany, United Kingdom, Spain, Finland, France and Italy (mean: 3113 patients; range: Finland 1266 to France 4507). Major differences were observed in national distribution of procedures: coronary artery bypass grafting accounted for 77.7% of procedures in Finland but only 46.2% in Spain. The EuroSCORE model goodness-of-fit was satisfactory in all countries (P-value overall: 0.4; UK: 0.34; Finland: 0.87; no values less than 0.05). Areas under ROC curves were 0.81 in Germany, 0.79 in the UK, 0.74 in Spain, 0.87 in Finland, 0.82 in France and 0.82 in Italy. CONCLUSION: Despite epidemiological differences between European countries, the discriminative power of EuroSCORE was good in Spain and excellent in all other countries. The system, developed from a merged European database, can therefore be used to assess improvement in quality of care achieved by surgeons and institutions as well as for international European comparison in adult heart surgery.

[Models of preoperative risk stratification in cardiac surgery: the prediction of mortality or of survival?]. / Modelli di stratificazione del rischio preoperatorio in cardiochirurgia: predizione di mortalità o di sopravvivenza?

BACKGROUND: The need to assess the quality of heart surgery outcomes stimulated the development of pre-surgical risk stratification models in order to predict outcome on the basis of patient characteristics. The aim of the study was to compare the predictive accuracy of hospital mortality according to the following three models: Parsonnet (NBI Score), Higgins (CCF Score) and Roques (French Score), in a setting totally independent from the one in which the models were derived. METHODS: For each of the 516 patients undergoing heart surgery at our institution between January 1992 and December 1993, we calculated the pre-surgical risk according to the three models. Then we compared the predicted mortality against the observed mortality by means of the Shannon accuracy index, the ROC curve analysis and the overestimation histogram. RESULTS: Overall observed mortality (1.5%) was similar to the predicted mortality by the NBI Score (1.5 +/- 2.1%, p = ns), the CCF Score (1.7 +/- 2.0%, p = ns) and the French Score (1.9 +/- 2.5%, p = ns). The predictive accuracy of global mortality is very high and equal with the three models, and it is very low in the 8 patients who died (NBI Score = 0.06 +/- 0.06; CCF Score = 0.125 +/- 0.067; French Score = 0.102 +/- 0.07, p = ns). The area under the ROC curve is identical in the 3 models. CONCLUSIONS: The predicted mortality obtained by the three models is not significantly different from the observed mortality and therefore, the global accuracy is similar and very high, while it is very low for patients who will die. The models for pre-surgical risk stratification are useful for comparing the results among different institutions or different surgeons, or for monitoring the results over time in the same institution, but they cannot be used to accurately predict the individual risk of hospital mortality.

Risk stratification for open heart operations: comparison of centers regardless of the influence of the surgical team.

BACKGROUND: Risk-adjusted mortality was previously used to compare institutions as a whole or surgeons. Because the same surgical team is working in two different hospitals, the aim of our study was to assess whether the institution can make a difference in surgical mortality. METHODS: Preoperative data of 554 patients in institution A and 500 in institution B were prospectively collected during the same period of time. All patients were operated on by the same surgeon with the same first assistant and anesthesiology staff in both institutions. Patient population was stratified according to Parsonnet's predictive model, in five risk groups, and mortality was adjusted by the direct standardization method. RESULTS: At institution A it was observed that in-hospital mortality was 2.3% (95% confidence interval, 1.3% to 4.0%), and in institution B 4.0% (95% confidence interval, 2.5% to 6.1%). The difference between the two mortality rates (1.7%; 95% confidence interval, -0.5% to 3.8%) is not statistically significant (p = 0.16), nor is the difference within each class. The standardized mortality ratio was 3.6% (95% confidence interval, 2.7% to 4.8%) and 5.8% (95% confidence interval, 4.6% to 7.2%), respectively. The difference of 2.2% (95% confidence interval, 0.5% to 3.8%) is statistically significant (p = 0.01). CONCLUSIONS: The institution can affect mortality of patients undergoing open heart operations, regardless of the influence of the surgical team.

Clinical outcome and emotional-behavioural status after isolated coronary surgery.

In order to evaluate clinical and psychosocial results of isolated coronary artery by-pass graft (CABG) we studied 626 consecutive patients, mean age 61 +/- 8 years (86% men), in a follow-up (median: 58 months) with a complete questionnaire about cardiosurgical problems (post-operative vital status, angina relapse, infarction, heart failure, PTCA, redo, PM) and psychosocial variables (mood, irritableness, job satisfaction, hobby satisfaction, family relations, sexual activity, general well-being and work status). Global evaluation improvement of psychosocial variables was found in 71% of patients without cardiac events (group A) and 11% of patients with cardiac events (group B); worsening was found in 2% of group A and 1% of group B; no referred variations in 13% and 2% respectively (p < or = 0.05. Interests (in work, hobbies and sexual activities) demonstrate an improvement in 20% (group A) and 2% (group B); worsening in 12% (group A) and 4% (group B); no variations in 51% (group A) and 11% (group B) (p < or = 0.005). Patients reported a well-being evaluation improvement about 66% in the group returning to work without restriction, 13% in those with limitation, 6% no further working; worse or unchanged well-being evaluation was found in 9% of patients returning to work without restriction, 3% with limitation, 3% no further working (p < or = 0.001).(ABSTRACT TRUNCATED AT 250 WORDS)

Quality control of laboratory results in real time.

An integrated informatic system may have an important role in identifying and reducing those errors which affect the reliability of laboratory results. The present work reports on the development of a system that, by means of a computer connected with laboratory instrumentation, allows monitoring of global error (bias, random, gross) using classical statistical quality control systems, integrated by auxiliary methods built on a database stored in the computer. These methods succeeded in greatly reducing bias and random error in the most frequent laboratory tests. Further methods to find, quantify, and reduce the gross and extra-analytical errors are under development.

[Survey on the timeliness and prognostic value of the decision to hospitalize in suspected acute myocardial infarct]. / Inchiesta sulla tempestività e sul valore predittivo delle scelte di ricovero del sospetto infarto miocardico acuto.

It was undertaken a survey, using 416 questionnaires returned from a sample of the Physicians on duty in the Emergency Department of Regione Piemonte (11 out of 25) over a period of 3 months and regarding patients coming to the observation for chest pain and suspected acute myocardial infarction/A.M.I., with the purpose of evaluating by univariate analysis the variable of decision-making subjects (General Practitioner, Physician On-Call, Cardiologist, Patient) behaviour and its effects on decision-making subject prevalence, hospitalization delay, positive predictive value/PPV for A.M.I. and final probability of hospitalization of A.M.I. within 3 hours. It was observed that 42% of the patients with suspected A.M.I. was referred by General Practitioner, 15% by Physician On-Call, 7% by a Cardiologist and 36% arrived by self-referral. The PPV for A.M.I. in the self-referred group was significantly lower in comparison with the PPV of the Cardiologist (p less than 0.01) and with the PPV of the Physician On-Call (p less than 0.01), but not in comparison with the PPV of the General Practitioner. Even if the timeliness of arrival in hospital was observed to be better in the self-referred group and in the group referred by the Physician On-Call, the final probability (considering subject intervention prevalence, PPV and timeliness) for the hospitalization of the A.M.I. within 3 hours from chest pain onset did not differ significantly among the groups of patients referred by the General Practitioner, the Physician On-Call, the Cardiologist or self-referred.