A pragmatic approach for mortality prediction after surgery in infective endocarditis: optimizing and refining EuroSCORE

CHARMS and PROBAST at your fingertips: a template for data extraction and risk of bias assessment in systematic reviews of predictive models

BACKGROUND

Systematic reviews of studies of clinical prediction models are becoming increasingly abundant in the literature. Data extraction and risk of bias assessment are critical steps in any systematic review. CHARMS and PROBAST are the standard tools used for these steps in these reviews of clinical prediction models.

RESULTS

We developed an Excel template for data extraction and risk of bias assessment of clinical prediction models including both recommended tools. The template makes it easier for reviewers to extract data, to assess the risk of bias and applicability, and to produce results tables and figures ready for publication.

CONCLUSION

We hope this template will simplify and standardize the process of conducting a systematic review of prediction models, and promote a better and more comprehensive reporting of these systematic reviews.

Strengths and limitations of patient registries in infective endocarditis

BACKGROUND

Infective endocarditis (IE) is a serious albeit relatively infrequent disease. Given the paucity of cases, particularly in non-referral centres, patient registries have progressively gained relevance to inform about the epidemiology, clinical presentation, and natural history of IE in the last two decades. Although they have become key to advancing knowledge of IE, registries also have shortcomings that lead to relevant consequences that are often overlooked.

OBJECTIVES

We aimed to discuss the strengths and limitations of registries in IE.

SOURCES

We conducted a PubMed search of relevant articles published between January 2000 and June 2022.

CONTENT

The backbone of the contemporary knowledge on IE has been built upon data collected in prospective registries, which has allowed us to collect data on relatively unknown aspects of the disease, identify knowledge gaps, and generate new hypotheses, serving as platforms for further research endeavours. Well-exploited registries can provide key information on how IE is distributed across populations and how it differentially impacts patients and subgroups. However, registries face several difficulties, such as the definition of IE, which includes subjective variables and changes over time. Other limitations include difficulty achieving a comprehensive collection of cases (which depends on both project funding and information systems), over-representation of the centres that created the registry, lack of inclusion of variables to assess endpoints that are relevant to patients in terms of quality of life and prognosis, and ethical issues.

IMPLICATIONS

The review of the advantages and disadvantages of registries aims to improve the quality of the information collected, the viability of the registry itself, and the ability to answer questions that are relevant to both researchers and patients.

Prognostic models for mortality after cardiac surgery in patients with infective endocarditis: a systematic review and aggregation of prediction models

BACKGROUND

There are several prognostic models to estimate the risk of mortality after surgery for active infective endocarditis (IE). However, these models incorporate different predictors and their performance is uncertain.

OBJECTIVE

We systematically reviewed and critically appraised all available prediction models of postoperative mortality in patients undergoing surgery for IE, and aggregated them into a meta-model.

DATA SOURCES

We searched Medline and EMBASE databases from inception to June 2020.

STUDY ELIGIBILITY CRITERIA

We included studies that developed or updated a prognostic model of postoperative mortality in patient with IE.

METHODS

We assessed the risk of bias of the models using PROBAST (Prediction model Risk Of Bias ASsessment Tool) and we aggregated them into an aggregate meta-model based on stacked regressions and optimized it for a nationwide registry of IE patients. The meta-model performance was assessed using bootstrap validation methods and adjusted for optimism.

RESULTS

We identified 11 prognostic models for postoperative mortality. Eight models had a high risk of bias. The meta-model included weighted predictors from the remaining three models (EndoSCORE, specific ES-I and specific ES-II), which were not rated as high risk of bias and provided full model equations. Additionally, two variables (age and infectious agent) that had been modelled differently across studies, were estimated based on the nationwide registry. The performance of the meta-model was better than the original three models, with the corresponding performance measures: C-statistics 0.79 (95% CI 0.76-0.82), calibration slope 0.98 (95% CI 0.86-1.13) and calibration-in-the-large -0.05 (95% CI -0.20 to 0.11).

CONCLUSIONS

The meta-model outperformed published models and showed a robust predictive capacity for predicting the individualized risk of postoperative mortality in patients with IE.

PROTOCOL REGISTRATION

PROSPERO (registration number CRD42020192602).

Comparative Analysis of Classification Algorithms in the Analysis of Medical Images From Speckle Tracking Echocardiography Video Data

Background. Machine learning allows applying various intelligent algorithms to produce diagnostic and/or prognostic models. Such models can be used to determine the functional state of the heart, which is diagnosed by speckle-tracking echocardiography. To determine the patient's heart condition in detail, a classification approach is used in machine learning. Each of the classification algorithms has a different performance when applied to certain situations. Therefore, the actual task is to determine the most efficient algorithm for solving a specific task of classifying the patient's heart condition when applying the same speckle-tracking echocardiography data set. Objective. We are aimed to evaluate the effectiveness of the application of prognostic models of logistic regression, the group method of data handling (GMDH), random forest, and adaptive boosting (AdaBoost) in the construction of algorithms to support medical decision-making on the diagnosis of coronary heart disease. Methods. Video data from speckle-tracking echocardiography of 40 patients with coronary heart disease and 16 patients without cardiac pathology were used for the study. Echocardiography was recorded in B-mode in three positions: long axis, 4-chamber, and 2-chamber. Echocardiography frames that reflect the systole and diastole of the heart (308 samples in total) were taken as objects for classification. To obtain informative features of the selected objects, the genetic GMDH approach was applied to identify the best structure of harmonic textural features. We compared the efficiency of the following classification algorithms: logistic regression method, GMDH classifier, random forest method, and AdaBoost method. Results. Four classification models were constructed for each of the three B-mode echocardiography positions. For this purpose, the data samples were divided into 3: training sample (60%), validation sample (20%), and test sample (20%). Objective evaluation of the models on the test sample showed that the best classification method was random forest (90.3% accuracy on the 4-chamber echocardiography position, 74.2% on the 2-chamber, and 77.4% on the long axis). This was also confirmed by ROC analysis, wherein in all cases, the random forest was the most effective in classifying cardiac conditions. Conclusions. The best classification algorithm for cardiac diagnostics by speckle-tracking echocardiography was determined. It turned out to be a random forest, which can be explained by the ensemble approach of begging, which is inherent in this classification method. It will be the mainstay of further research, which is planned to be performed to develop a full-fledged decision support system for cardiac diagnostics.

Prognostic assessment of valvular surgery in active infective endocarditis: multicentric nationwide validation of a new score developed from a meta-analysis

OBJECTIVES

Several risk prediction models have been developed to estimate the risk of mortality after valve surgery for active infective endocarditis (IE), but few external validations have been conducted to assess their accuracy. We previously developed a systematic review and meta-analysis of the impact of IE-specific factors for the in-hospital mortality rate after IE valve surgery, whose obtained pooled estimations were the basis for the development of a new score (APORTEI). The aim of the present study was to assess its prognostic accuracy in a nationwide cohort.

METHODS

We analysed the prognostic utility of the APORTEI score using patient-level data from a multicentric national cohort. Patients who underwent surgery for active IE between 2008 and 2018 were included. Discrimination was evaluated using the area under the receiver operating characteristic curve, and the calibration was assessed using the calibration slope and the Hosmer-Lemeshow test. Agreement between the APORTEI and the EuroSCORE I was also analysed by Lin's concordance correlation coefficient (CCC), the Bland-Altman agreement analysis and a scatterplot graph.

RESULTS

The 11 variables that comprised the APORTEI score were analysed in the sample. The APORTEI score was calculated in 1338 patients. The overall observed surgical mortality rate was 25.56%. The score demonstrated adequate discrimination (area under the receiver operating characteristic curve = 0.75; 95% confidence interval 0.72-0.77) and calibration (calibration slope = 1.03; Hosmer-Lemeshow test P = 0.389). We found a lack of agreement between the APORTEI and EuroSCORE I (concordance correlation coefficient = 0.55).

CONCLUSIONS

The APORTEI score, developed from a systematic review and meta-analysis, showed an adequate estimation of the risk of mortality after IE valve surgery in a nationwide cohort.

Surgical treatment of infective endocarditis: Results in 831 patients from a single center

INTRODUCTION

This study evaluated surgical outcomes of infective endocarditis (IE), with particular attention to the impact of intravenous drug use (IVDU).

METHODS

Adult patients undergoing surgery for IE between 2011 and 2018 at a single center were included and stratified by IVDU. The primary outcome was overall survival. Secondary outcomes included postoperative complications and hospital readmissions. Kaplan-Meier and multivariable Cox regression were utilized for unadjusted and risk-adjusted survival analyses, respectively. Cumulative incidence function curves were compared for hospital readmissions.

RESULTS

A total of 831 patients (mean age 55 years, 34.4% female) were operated on for IE, including 318 (38.3%) with IVDU. Cultures were most commonly positive for streptococcus (25.2%), methicillin-sensitive Staphylococcus aureus (17.7%), enterococcus (14.3%), or methicillin-resistant Staphylococcus aureus (8.4%). The most common procedures included isolated aortic valve repair/replacement (18.8%), aortic root replacement (15.9%), mitral valve repair/replacement (26.7%), aortic and mitral valve replacement (8.4%), and tricuspid valve repair/replacement (7.6%). Mean follow-up was 3.4 ± 2.4 years. Overall 5-year survival was 64% and was similar between IVDU and non-IVDU. Multivariable analysis demonstrated that IVDU was not associated with mortality risk. IVDU patients displayed higher rates of all-cause readmission (61.6% vs 53.9%; P = .03), drug-use readmission (15.4% vs 1.4%; P < .001), and recurrent endocarditis readmission (33.0% vs 13.0%; P < .001).

CONCLUSIONS

The majority of patients undergoing surgical treatment of IE are alive at 5-years although readmission rates are high. IVDU is not a risk factor for longitudinal mortality although patients with IVDU are at higher overall readmission risk, driven largely by greater readmissions for drug-use and recurrent endocarditis.

WITHDRAWN: Application of Echocardiographic Information Technology in the Research of Infective Endocarditis and Its Neurological Complications

This article has been withdrawn at the request of the Editor-in-Chief. The Publisher apologizes for any inconvenience this may cause. The full Elsevier Policy on Article Withdrawal can be found at https://www.elsevier.com/about/our-business/policies/article-withdrawal.

Prediction of surgical risk in patients with endocarditis: Comparison of logistic EuroSCORE, EuroSCORE II and APORTEI score

OBJECTIVES

APORTEI score is a new risk prediction model for patients with infective endocarditis. It has been recently validated on a Spanish multicentric national cohort of patients. The aim of the present study is to compare APORTEI performances with logistic EuroSCORE and EuroSCORE II by testing calibration and discrimination on a local sample population underwent cardiac surgery because of endocarditis.

METHODS

We tested three prediction scores on 111 patients underwent surgery from 2014 to 2020 at our Institution because of infective endocarditis. Area under the curves and Hosmer-Lemeshow test were used to analyze discrimination and calibration respectively of logistic EuroSCORE, EuroSCORE II and APORTEI score.

RESULTS

The overall observed one-month mortality rate was 21.6%. The observed-to-expected ratio was 1.27 for logistic EuroSCORE, 3.27 for EuroSCORE II and 0.94 for APORTEI. The area under the curve (AUC) value of APORTEI (0.88±0.05) was significantly higher than that one of logistic EuroSCORE (AUC 0.77±0.05; p 0.0001) and of EuroSCORE II (AUC 0.74±0.05; p 0.0005). Hosmer-Lemeshow test showed better calibration performance of the APORTEI, (logistic EuroSCORE: p 0.19; EuroSCORE II: p 0.11; APORTEI: p 0.56).

CONCLUSION

APORTEI risk score shows significantly higher performances in term of discrimination and calibration compared with both logistic EuroSCORE and EuroSCORE II.

Age, creatinine and ejection fraction (ACEF) score: a simple risk-stratified method for infective endocarditis

BACKGROUND

Older age, renal dysfunction and low left ventricular ejection fraction are accepted predictors of poor outcome in patients with infective endocarditis (IE). This study aimed to investigate the prognostic significance of the age, creatinine and ejection fraction (ACEF) score in IE.

METHODS

The study involved 1019 IE patients, who were classified into three groups according to the tertiles of ACEF score: low ACEF (<0.6, n = 379), medium ACEF (0.6-0.8, n = 259) and high ACEF (>0.8, n = 381). The ACEF score was calculated as follows: age (years)/ejection fraction (%)+1 (if serum creatinine value was >2 mg/dL). The relationship between ACEF score and adverse events was analyzed.

RESULTS

In-hospital mortality was 8.2%, which increased with the increase of ACEF score (4.2% vs. 5.0% vs. 14.4% for the low-, medium- and high-ACEF groups, respectively; P < 0.001). ACEF score had a good discriminative ability for predicting in-hospital death [areas under the curve (AUC), 0.706, P < 0.001]. The predictive value of ACEF score in surgical treatment was significantly higher than in conservative treatment for predicting in-hospital death (AUC, 0.812 vs. 0.625; P = 0.001). Multivariable analysis revealed that ACEF score was independently associated with in-hospital mortality (adjusted odds ratio, 2.82; P < 0.001) and long-term mortality (adjusted hazard ratio, 2.51; P < 0.001).

CONCLUSION

ACEF was an independent predictor for in-hospital and long-term mortality in IE patients, and it could be considered as a useful tool for risk stratification. ACEF score was more suitable for surgical patients in terms of assessing the risk of in-hospital mortality.

Short- and long-term mortality in patients with left-sided infective endocarditis not undergoing surgery despite indication

INTRODUCTION AND OBJECTIVES

In infective endocarditis (IE), decisions on surgical interventions are challenging and a high percentage of patients with surgical indication do not undergo these procedures. This study aimed to evaluate the short- and long-term prognosis of patients with surgical indication, comparing those who underwent surgery with those who did not.

METHODS

We included 271 patients with left-sided IE treated at our institution from 2003 to 2018 and with an indication for surgery. There were 83 (31%) surgery-indicated not undergoing surgery patients with left-sided infective endocarditis (SINUS-LSIE). The primary outcome was all-cause death by day 60 and the secondary outcome was all-cause death from day 61 to 3 years of follow-up. Multivariable Cox regression and propensity score matching were used for the analysis.

RESULTS

At the 60-day follow-up, 40 (21.3%) surgically-treated patients and 53 (63.9%) SINUS-LSIE patients died (P <.001). Risk of 60-day mortality was higher in SINUS-LSIE patients (HR, 3.59; 95%CI, 2.16-5.96; P <.001). Other independent predictors of the primary endpoint were unknown etiology, heart failure, atrioventricular block, and shock. From day 61 to the 3-year follow-up, there were no significant differences in the risk of death between surgically-treated and SINUS-LSIE patients (HR, 1.89; 95%CI, 0.68-5.19; P=.220). Results were consistent after propensity score matching. Independent variables associated with the secondary endpoint were previous IE, diabetes mellitus, and Charlson index.

CONCLUSIONS

Two-thirds of SINUS-LSIE patients died within 60 days. Among survivors, the long-term mortality depends more on host conditions than on the treatment received during admission.

Prognostic factors of mortality after surgery in infective endocarditis: systematic review and meta-analysis

PURPOSE

There is a lack of consensus about which endocarditis-specific preoperative characteristics have an actual impact over postoperative mortality. Our objective was the identification and quantification of these factors.

METHODS

We performed a systematic review of all the studies which reported factors related to in-hospital mortality after surgery for acute infective endocarditis, conducted according to PRISMA recommendations. A search string was constructed and applied on three different databases. Two investigators independently reviewed the retrieved references. Quality assessment was performed for identification of potential biases. All the variables that were included in at least two validated risk scores were meta-analyzed independently, and the pooled estimates were expressed as odds ratios (OR) with their confidence intervals (CI).

RESULTS

The final sample consisted on 16 studies, comprising a total of 7484 patients. The overall pooled OR were statistically significant (p < 0.05) for: age (OR 1.03, 95% CI 1.00-1.05), female sex (OR 1.56, 95% CI 1.35-1.81), urgent or emergency surgery (OR 2.39 95% CI 1.91-3.00), previous cardiac surgery (OR 2.19, 95% CI 1.84-2.61), NYHA ≥ III (OR 1.84, 95% CI 1.33-2.55), cardiogenic shock (OR 4.15, 95% CI 3.06-5.64), prosthetic valve (OR 1.98, 95% CI 1.68-2.33), multivalvular affection (OR 1.35, 95% CI 1.01-1.82), renal failure (OR 2.57, 95% CI 2.15-3.06), paravalvular abscess (OR 2.39, 95% CI 1.77-3.22) and S. aureus infection (OR 2.27, 95% CI 1.89-2.73).

CONCLUSIONS

After a systematic review, we identified 11 preoperative factors related to an increased postoperative mortality. The meta-analysis of each of these factors showed a significant association with an increased in-hospital mortality after surgery for active infective endocarditis. Graph summary of the Pooled Odds Ratios of the 11 preoperative factors analyzed after the systematic review and meta-analysis.