Sage Journals: Discover world-class research

Abstract

Infective endocarditis is a complex heterogeneous condition involving the infection of the endocardium and heart valves, leading to severe complications, including death. Surgery is often indicated in patients with infective endocarditis but is associated with elevated risk compared with other forms of cardiac surgery. Risk models play an important role in many cardiac surgeries as they can help inform clinicians and patients regarding procedural risk, decision-making to proceed or not, and influence perioperative management; however, they remain under-utilized in the infective endocarditis settings. Another crucial role of such risk models is to assess predicted versus found mortality, thereby allowing an assessment of institutional performance in infective endocarditis surgery. Traditionally, general cardiac surgery risk models such as European System for Cardiac Operative Risk Evaluation (EuroSCORE), EuroSCORE II, and Society of Thoracic Surgeon’s score have been applied to endocarditis surgery. However, there has been the development of many endocarditis surgery-specific scores over the last decade. This review aims to discuss clinical characteristics and applications of all contemporary risk scores in the setting of surgical treatment of infective endocarditis.

Keywords

infective endocarditis operative mortality risk scores

Introduction

Infective endocarditis (IE) was first described in 1885 by Sir William Osler. It is an infection of the endocardial layer that involves the valves, exposing the subendothelial matrix, and promoting inflammation and fibrin exposure. This leads to the activation of endothelial cells, the release of cytokines, and other inflammatory markers, eventually forming biofilm.^1–3 The incidence of IE has been increasing with time, reaching more than 1 million, with 66,320 deaths in 2019 and 12.7 hospitalizations per 100,000 in the United States alone.^4,5 Additionally, IE is associated with excessive length of stay, increased cost of hospitalization, and mortality in 14.5% of admitted patients in the United States.^5,6 On a global level the mortality of IE ranges from 3.7% to 28.6%. Murdoch et al.⁷ demonstrated a mortality of 18% in their International Collaboration on Endocarditis (ICE)–Prospective Cohort Study.⁷ Similarly, a prospective study from the EURO-ENDO registry reflected 17.1% mortality in patients with IE.⁸ However, this varied according to the underlying risk factor; for example, patients with adult congenital heart disease have a 30-day, 1-year survival rate of 95.4% and 92.7%, respectively.⁹ In addition to mortality, IE is also associated with severe complications such as congestive heart failure, heart block, valve destruction, abscesses, embolic cerebrovascular accident, and many others, with mortality reaching as high as 30%.^10–12 Furthermore, common risk factors for IE include a history of valvular heart disease, rheumatic heart disease, valvular surgery, indwelling cardiac devices, intravenous drug users, congenital heart disease, and a history of IE.^3,9,13,14

Management of IE is multi-disciplinary, involving medical and, in certain individuals, surgical management. According to the American Association for Thoracic Surgery consensus guidelines, certain individuals might benefit from surgical management to control the infection.¹⁵ Those include patients with heart failure, prosthetic valve, relapsing infection, recurrent embolization, local complications such as abscess, with varying levels of evidence.^14,15 Most available literature suggests that surgery is potentially lifesaving for left-sided IE.^13,16 Vikram et al.¹⁷ demonstrated in their multicenter cohort study that surgery was associated with lower 6-month mortality in patients with complicated left-sided native valve endocarditis.¹⁷ Kang et al.¹⁸ also demonstrated that early surgery in left-sided IE patients was associated with reduced all-cause mortality and embolic events.¹⁸ However, surgical management is often a high-risk procedure associated with high mortality risk.¹⁹

The decision for surgery remains challenging, and surgery may be refused by patients and relatives, or not recommended by physicians or surgeons. This occurs in as many as 24% of patients with IE, due to high surgical risk (age, cardiogenic shock, myocardial invasion by the infection, preoperative neurologic complications).^20–22 For such reasons, several operative risk scores have been developed, incorporating risk factors such as age, renal failure, type of organism, and presence of local complications, which predict the risk of mortality and morbidity with surgery. These include the Society of Thoracic Surgeons (STS) and the European System for Cardiac Operative Risk Evaluation (EuroSCORE) with the hope to better assist with shared decision-making.^22–26 Table 1 summarizes the characteristics of all the studies developing endocarditis surgery risk scores, and Table 2 lists the main parameters of all the endocarditis surgery risk scores. This article will review the literature for the commonest operative risk scores used during IE, discussing their strengths and limitations, which could increase awareness of such scores.

Table 1.

Characteristics of all the studies developing endocarditis surgery risk scores.

Scores	Author, year	Country	Number of hospitals	Number of patients	Mean age (in years)	Female (%)	Inclusion criteria	Operative mortality (%)
EuroSCORE	Nashef, 1999	United Kingdom	Multi-center	14,799	Not available	28	All cardiac surgery	4.7
EuroSCORE II	Nashef, 2012	United Kingdom	154	22,381	64.6 ± 12.5	30.9	All cardiac surgery	3.9
STS	O’Brien, 2018	USA	Multi-center	150,150	Not available	44.2	All cardiac surgery	3.2
ACEF/AGEF	Wei, 2019	China	1	1019	44.4 ± 0.53	29.9	⩾18 years with IE, no repeated hospital records, no missing admission creatinine values, and LVEF records, not diagnosed with IE after valve replacement surgery during the same hospitalization	8.2
AEPEI	Gatti, 2017	Italy	8	361	59.1 ± 1.2	21	⩾18 years who underwent surgery for IE (definite IE cases defined by MDC)	15.5
ANCLA	Gatti, 2016	Italy	1	138	60.6 ± 8.5	19.6	Patients with definite IE defined by MDC undergoing surgery	20.3
APORTEI	Barca, 2019	Spain	32	1338	63.6 ± 13.1	27.1	>18 years of age with active IE undergoing surgery	25.6
COSTA	Costa, 2007	Brazil	1	186	58 ± 15.4	39.7	Confirmed IE cases	26.3
De Feo-Cotrufo	De Feo, 2012	Italy	1	440	49 ± 16	28	Patients with IE of native valves	9.1
EndoSCORE	Di Mauro, 2017	Italy	26	2715	59.6 ± 15.1	26.9	Patients with native or PVE	11
ICE	Park, 2016	USA	61	5246	Median age: 59 (45–72)	32	Patients with definite IE defined by MDC	20.2
MELD XI	He, 2018	China	1	609	60.6 ± 8.5	29.6	IE patients confirmed by MDC	8.1
Meta-model	Fernandez-Felix, 2021	Spain	Multi-center	19,039	–	–	Original studies with prognostic models, with or without external validation, to predict mortality after IE surgery	8.2–29.2
PALSUSE	Martinez-Selles, 2014	Spain	26	437	61.4 ± 15.5	24.7	IE patients confirmed by MDC	24.3
RISK-E	Olmos, 2017	Spain	3	671	61 ± 14	31	Patients with definite left-sided IE in the active phase of the disease needing surgery	28.6
SHARPEN	Chee, 2015	Singapore	1	233	50 ± 19	–	Patients with definite IE Duke’s criteria	23
STS-IE	Gaca, 2010	USA	Multi-center	13,617	Median: 55 (45–66)	33.23	IE patients who underwent AV, MV, and/or TV at STS-participating hospitals	8.2
STOP	Habertheuer, 2021	USA	10	1181	Median age: 39 (30–54)	32.7	Patients with a history of IDA undergoing IE surgery	5.9
SYSUPMIE	Luo, 2022	China	2	276	43.7 ± 15.3	29.7	Patients with definite MDC (NVE and PVE)	7.2

ACEF, age, creatinine, and ejection fraction; AGEF, age, GFR, and ejection fraction; AV, aortic valve; ICE, International Collaboration on Endocarditis; IDA, illicit drug abuse; IE, infective endocarditis; EuroSCORE, European System for Cardiac Operative Risk Evaluation; LVEF, left ventricular ejection fraction; MDC, modified Duke’s criteria; MELD, model for end-stage liver disease; MV, mitral valve; NVE, native valve endocarditis; PVE, prosthetic valve endocarditis; STOP, STratification risk analysis in OPerative management; STS-IE, Society of Thoracic Surgeons–infective endocarditis; SYSUPMIE, Sun Yat-sen University Prediction Model for IE; ANCLA, Anemia, NYHA class IV, critical state, large intracardiac destruction, surgery on thoracic aorta; APORTEI, Análisis de los factores PROnósticos en el Tratamiento quirúrgico de la Endocarditis Infecciosa; AEPEI, Association pour l’Etude et la Prevention de l’Endocadite Infectieuse; PALSUSE, prosthetic valve, age≥70, large intracardiac destruction, Staphylococcus spp, urgent surgery, sex (female), EuroSCORE ≥10); RISK-E, Risk Endocarditis; SHARPEN, Systolic BP, Heart Failure, Age, Renal Function, Pneumonia, Elevated peak CRP, Non-Intravenous drug abusers; TV, tricuspid valve; USA, United States.

Table 2.

Parameters of all the endocarditis surgery risk scores.

Parameters	EuroSCORE I	EuroSCORE II	STS	ACEF	AGEF	AEPEI	ANCLA	APORTEI	COSTA	De Feo-Cotrufo	EndoSCORE	ICE	MELD XI	Meta-model	PALSUSE	RISK-E	SHARPEN	STS-IE	STOP	SYSUPMIE
Endocarditis	✓	✓	✓	These are endocarditis-specific risk scores utilized in surgery for IE
Age	✓	✓	✓	✓	✓			✓	✓	✓	✓	✓		✓	✓	✓	✓	✓
Gender	✓	✓	✓					✓			✓			✓	✓			✓
Physical parameters		✓ (height and weight)	✓ (height and weight)			✓ (body mass index)												✓ (body surface area)
NYHA class						✓	✓	✓	✓	✓		✓		✓				✓
Critical preoperative state	✓		✓			✓	✓	✓		✓	✓			✓	✓	✓		✓	✓
Microorganism								✓ (Staphylococcus aureus)			✓ (Pseudomonas aeruginosa, Staphylococcus aureus, fungal disease, others)	✓ (Staphylococcus aureus, viridans group streptococci)			✓ (Staphylococcus species)	✓			✓ (organism another than Staphylococcus)
Intra-cardiac complication							✓	✓		✓	✓	✓		✓	✓	✓
Previous cardiac surgery	✓	✓	✓					✓						✓				✓
Diabetes		✓	✓															✓
Peripheral arterial disease/extracardiac arteriopathy	✓	✓	✓
Chronic lung disease/COPD	✓		✓								✓			✓
Prosthetic valve								✓				✓			✓	✓
LV dysfunction	✓	✓	✓	✓	✓						✓			✓
Serum creatinine	✓	✓		✓							✓		✓	✓				✓
GFR/eGFR					✓	✓
AKI/dialysis		✓ (dialysis)	✓ (renal dysfunction)					✓ (renal failure)		✓ (preoperative renal failure)		✓ (history of dialysis)				✓ (AKI)	✓ (renal function)		✓ (dialysis)
Others	Neurological dysfunction, unstable angina, recent MI, pulmonary hypertension, other than isolated CABG, thoracic aorta surgery, post-infarct septal rupture	Pulmonary disease, neurological or musculoskeletal dysfunction, serum albumin, BNP, symptomatic status, recency and size of MI, systolic PA pressure, the urgency of operation, type of procedure performed in detail	CVD, neurological dysfunction, CHD, poor mobility, elective status, type of surgery, number of diseased vessels, valve pathology (aortic and mitral), pulmonary hypertension, aortic aneurysm/dissection, type of operation (isolated or combined)			Systolic PA pressure	Anemia, aortic cross-clamping time >150 min	Multi-valvular	Sepsis nonresponsive to antibiotic therapy, conduction disorders, arrhythmias, echocardiographic classification, vegetation measuring 10 mm or more	Lack of preoperative attainment of negative blood cultures	Number of treated valves/procedures	Nosocomial IE, symptoms >1 month before admission, vegetation (aortic and mitral), stroke, persistent bacteremia, surgical treatment	Total bilirubin	Pulmonary hypertension, valves treated, etiology of infection, valve location	EuroSCORE ⩾ 10	Thrombocytopenia	Systolic blood pressure, heart failure, pneumonia, elevate peak C-reactive protein, non-intravenous drug abusers	Multiple valve procedure, active endocarditis, arrhythmia	Number of valves affected, valve procedure performed (aortic, mitral, tricuspid)	Heart failure, diastolic function, platelet count, tricuspid valve involvement, presence of urine occult blood, serum albumin, vegetation size ⩾10 mm, multiple valve involvement

ACEF, age, creatinine, and ejection fraction; AGEF, age, GFR, and ejection fraction; AKI, acute kidney injury; BNP, brain natriuretic peptide; CHD, congenital heart disease; CABG, coronary artery bypass grafting; COPD, chronic obstructive pulmonary disease; CVD, cerebrovascular disease; EGFR, estimated glomerular filtration rate; EuroSCORE, European System for Cardiac Operative Risk Evaluation; GFR, glomerular filtration rate; ICE, International Collaboration on Endocarditis; IE, Infective endocarditis; LV, left ventricle; MI, myocardial infarction; NYHA, New York Heart Association; PA, pulmonary artery; STOP, STratification risk analysis in OPerative management; STS-IE, Society of Thoracic Surgeons–infective endocarditis; SYSUPMIE, Sun Yat-sen University Prediction Model for IE.

Methods

Search strategy

A comprehensive search was performed to identify all the studies investigating scoring models utilized in predicting mortality after surgery for IE. A combination of indexing terms, keywords, truncation, and adjacency was used to find all potentially relevant concepts (Supplemental Material). The search was performed across Ovid Medline, Ovid Embase, and Cochrane Library (via Wiley). It was limited to articles published in English from the inception of the databases up till the current date.

Inclusion and exclusion criteria

Studies were included if they satisfied all the following requirements: (1) original research articles, (2) focused on IE risk scoring models, (3) offered details on the creation or validation of the models, (4) included adult human subjects, and (5) were published in English. Studies were excluded if they were review articles, case reports, or studies conducted in pediatric populations.

General cardiac surgery scores

EuroSCORE I

The EuroSCORE, constructed in Europe in 1999, is a risk score widely used in cardiovascular surgery that helps predict the risk of perioperative mortality in patients undergoing a range of procedures.²⁷ There are two distinct ways to score the EuroSCORE, which are the additive model and the logistic model. The additive model can be computed through basic arithmetic, while the logistic model is more complex and needs a computer to generate the score.²⁸ Developed by the European Association for Cardiothoracic Surgery, EuroSCORE takes into account various patient factors (17 items), such as age, gender, comorbidities, and the type of procedure.²⁹ One of the key strengths of EuroSCORE is that it has been extensively validated in multiple patient populations and has consistently been shown to be a reliable predictor of perioperative mortality.^30,31 However, one of the limitations of EuroSCORE is that it has a low discrimination ability for valve surgery and transcatheter aortic valve implantation (TAVI).^32,33 In fact, IE is a rare but life-threatening complication after TAVI. A recent systematic review by Hassanin et al.³⁴ demonstrated that most IE events after TAVI occur in the first year following TAVI. Age, sex, comorbid conditions, and procedural factors constitute the risk factors for IE and surgical intervention is seldom performed in this cohort. Another limitation is that it tends to overestimate risk in low-risk patients and underestimate risk in high-risk patients.^32,35,36

In summary, EuroSCORE is a widely used risk score in cardiovascular surgery that predicts perioperative mortality based on patient factors. Still, it has limitations in its discrimination ability and risk estimation accuracy.

EuroSCORE II

In February 2012, a new system was released, which exhibited improved discrimination and calibration compared to the prior model.^35,37,38 EuroSCORE II is an updated version of the original EuroSCORE, developed to improve the accuracy of risk prediction for patients undergoing cardiac surgery. Like its predecessor, EuroSCORE II is a composite score that considers various clinical and demographic factors to predict the risk of mortality in cardiac surgery patients. However, EuroSCORE II incorporates additional variables, such as active endocarditis, chronic pulmonary disease, and a reduced left ventricular ejection fraction (LVEF), which were not included in the original EuroSCORE.³⁷ Additionally, it uses a different statistical model than the original EuroSCORE, which may result in more accurate risk assessments.

Overall, EuroSCORE II is considered a more reliable and valid predictor of mortality risk in cardiac surgery patients than the original EuroSCORE but still underpredicts risk for patients undergoing TAVI.^33,39,40

STS risk model

In adult cardiac surgery, the STS risk model predicts the risk of operative mortality and morbidity.⁴¹ Risk models are used in adult cardiac surgery for various purposes, including quality measurement, clinical practice improvement, voluntary public reporting, and research. Initially, the focus of risk models was on preoperative mortality prediction. However, the ability to predict only operative mortality is insufficient for predicting surgical outcomes. Therefore, The STS score has been expanded to include the calculation of postoperative morbidity.⁴¹ The predictive performance of the STS algorithms remains the most widely used model in the United States. The National Adult Cardiac Surgery Database of the STS (STS NCD) has become the national standard for recording cardiac surgery data. STS NCD was developed in the late 1980s.^42,43 STS started developing risk models in 1994, beginning with isolated coronary artery bypass graft (CABG) mortality, and has undergone periodic revisions⁴³ as several morbidity risk models. The STS has a large and diverse database that allows for the development of multiple outcome models for cardiac surgical procedures.^43,44 Shahian et al., demonstrated the STS 2008 cardiac surgery risk model using the data collected during 2002–2006, which includes multiple outcomes for each of three cardiac procedure groups, such as isolated CABG, isolated valve procedure [aortic valve replacement (AVR), mitral valve replacement or repair (MVR/r)], and isolated valve procedure plus CABG. The outcomes are operative mortality, renal failure, stroke, reoperation for any cause, prolonged ventilation, deep sternal wound infection, composite major morbidity or mortality, and length of stay. The STS has created a set of completely new risk models for adult cardiac surgery to consider the changes in patient characteristics, risk profiles, surgical technique, and outcomes.⁴⁴ In 2018, STS published updated risk models for operative mortality and major morbidity for the isolated CABG, isolated AVR, isolated MVR/r, AVR + CABG, and MVR/r + CABG.⁴⁵ STS risk score can be calculated by STS-approved software and using the online STS calculator for one patient at a time and by applying the published regression coefficients for one or more patients at a time.⁴⁶ The performance of risk models can be assessed from two perspectives. One of these is discrimination which refers to the ability of the model to distinguish who will and who will not experience the given outcome such as operative mortality. This is quantified by calculating the area under the curve (AUC), or C-statistic. The C-statistic typically ranges from 0.5 to 1.0. A model that does not discriminate has a C-statistic of 0.5. In contrast, a model that accurately predicts an outcome has a C-statistic of 1.0 (perfect discrimination). C-statistics in the validation sample of the STS Adult Cardiac Surgery (ACS) 2018 risk models ranged from 0.588 for reoperation in valve + CABG to 0.826 for renal failure in CABG. To assess the calibration of a risk model, the observed outcomes are compared to the predicted outcomes, with the aim of achieving a ratio of observed to expected outcomes equaling 1 (O/E = 1). Calibration in the validation sample was excellent. This model exceeded the previous STS model in terms of calibration and discrimination except for mediastinitis/deep sternal wound infection and stroke in valve patients.^47,48 The recently developed STS 2021 ACS risk models for operative mortality, major morbidity and combined mortality/ morbidity for multiple valve operations include AVR + MVR/r and AVR + MVR/r + CABG.⁴⁹ Although the STS risk score is used to predict the operative mortality and morbidity, it is the strongest predictor of long-term survival following TAVI⁵⁰ It does have limitations, such as that it cannot be used to study the other type of cardiac surgery that have not yet been included such as Maze procedure for surgical ablation of atrial fibrillation.⁵¹

To summarize, widely used in adult cardiac surgery, the STS risk models forecast operative mortality and morbidity while emphasizing quality measurement, practice improvement, public reporting, and research. They also take into account patient characteristics, risk profiles, surgical technique, and outcomes into account.

Endocarditis surgery-specific scores

ACEF and AGEF score

Age, creatinine, and ejection fraction (ACEF) and age, glomerular filtration rate (GFR), and ejection fraction (AGEF) score were designed by Wei et al. in 2019.⁵² ACEF score was first designed for cardiac surgical patients with extrapolation to other conditions. It was also used to predict perioperative mortality in mitral valve surgery.⁵³ ACEF score additionally allowed risk stratification of events after coronary revascularization.^54,55 Capodanno et al.⁵⁶ integrated GFR to the score giving birth to AGEF score and reflected that AGEF score had superior discriminatory capabilities than ACEF score in patient cohort undergoing percutaneous coronary intervention.⁵⁶

A total of 1019 patients were included with a primary endpoint of in-hospital mortality and secondary endpoint of in-hospital major adverse cardiovascular events (MACEs) and long-term all-cause mortality. Overall, in-hospital and long-term mortality were 8.2% and 14.5%, respectively.

The patients were divided into three groups – low ACEF (<0.6, n = 379), medium ACEF (0.6–0.8, n = 259), and high ACEF (>0.8, n = 381). The maximum follow-up duration was 82 months with a median time of 29 months. Long-term mortality was higher in high ACEF groups as compared to low and medium groups. The predictive ability of ACEF and AGEF scores was comparable for both in-hospital and long-term mortality. The AUC for ACEF score and AGEF score was 0.717 (0.671–0.764) and 0.729 (0.685–0.773), respectively. This scoring system had its own limitations. Residual confounding factors might have played a role in influencing the clinical outcomes. ACEF score was calculated during admission of the patients and its prognostic value in the follow-up period is uncertain.

Therefore, risk assessment tools ACEF and AGEF scores exhibit predictive capacities for in-hospital and long-term mortality in cardiac surgery patients, with the AGEF score exhibiting superior discriminatory powers in percutaneous coronary intervention cases.

AEPEI score: The association for the study and prevention of IE study group

The AEPEI score was developed by Gatti et al.⁵⁷ in 2017 to assess the risk of in-hospital mortality following surgery for IE. The study, which was conducted at multiple centers, analyzed the outcomes of 361 patients who underwent surgery for IE. Five variables were found to be independently associated with increased in-hospital mortality: BMI > 27 kg/m² [odds ratio (OR): 1.79; p = 0.049], estimated GFR <50 mL/min (OR: 3.52; p < 0.0001), New York Heart Association (NYHA) class IV (OR: 2.11; p = 0.024), systolic pulmonary artery pressure >55 mm Hg (OR: 1.78; p = 0.032), and critical state (OR: 2.37; p = 0.017). The score ranges from 0 to 7, with a mortality rate of 4.5% for a score of 0 and 72.4% for a score of 7. The score was validated both internally and externally, with results showing satisfactory calibration and discriminatory power. Notably, the AEPEI score was found to be similar to the EuroSCORE II. The primary limitation of the AEPEI score is the small sample size and the presence of only a few variables, which increases the risk of potential confounding variables and unaccounted for factors related to mortality. Noteworthily, Gatuz and Cuenza⁵⁸ evaluated the sensitivity and specificity of the AEPEI score in 66 patients and found a sensitivity of 100% and a specificity of 50%.

To summarize, AEPEI score is a risk assessment tool for in-hospital mortality following surgery for IE, based on five variables, and showing satisfactory calibration and discriminatory power, with similarities to EuroSCORE II, but limited by a small sample size and potential confounding variables.

ANCLA score

The ANCLA score is a scoring system used to predict the risk of in-hospital death after surgery for IE. It was developed by Gatti et al.⁵⁹ in 2017 based on a study of 138 patients who underwent IE surgery. The score was created using multivariate analysis to identify independent predictors of in-hospital mortality, and variables were defined as follows: anemia (OR: 11.0; p = 0.035), NYHA class IV (OR: 2.61; p = 0.09), critical state (OR: 4.97; p = 0.016), large intracardiac destruction (OR: 6.45; p = 0.0014), and surgery of the thoracic aorta (OR: 7.51; p = 0.041). The authors reported that the ANCLA score outperformed three specific scoring systems for in-hospital mortality following EI surgery and was similar to EuroSCORE II despite having fewer parameters.⁶ It was internally validated but not externally, which is a limitation, along with the small sample size. However, in a later analysis, when compared one to one with other risk factor scores (STS, De Feo, PALSUSE, RISK-E, EndoSCORE, and AEPEI) the ANCLA score was found to have higher accuracy of prediction except for the AEPEI score where it was similar (p = 0.077).⁵⁹

To summarize, the ANCLA score is a predictive scoring system for in-hospital mortality after surgery for IE, demonstrating superior accuracy compared to several other risk scores.

APORTEI score

The APORTEI score was developed from the pooled estimates obtained from a systematic review and meta-analysis examining IE-specific risk factors, and their impact on in-hospital mortality after IE valvular surgery.²⁴ The authors then assessed its prognostic utility in a multicentric nationwide cohort of 1338 patients who underwent surgery for active IE from 2008 to 2018. They also analyzed agreement between the APORTEI and EuroSCORE I. The score demonstrated adequate discrimination (AUC: 0.75; 95% CI: 0.72–0.77) and calibration (calibration slope = 1.03; Hosmer–Lemeshow test: p = 0.389), with adequate estimation of the risk of mortality after IE valve surgery. However, there was a lack of agreement between the APORTEI and EuroSCORE I (concordance correlation coefficient 0.55).²⁴

A study by Fernández-Cisneros et al. employed all the common scores in their cohort of left-sided IE patients who underwent cardiac surgery in a 5-year period. The PALSUSE (AUC: 0.87), RISK-E (AUC: 0.89), and APORTEI (AUC: 0.87) scores had the best performance in their cohort.⁶⁰ Urso et al. compared the performances of APORTEI, EuroSCORE, and EuroSCORE II on 111 patients who underwent IE cardiac surgery at a single center. The O/E ratio was 1.27 for logistic EuroSCORE, 3.27 for EuroSCORE II, and 0.94 for APORTEI. APORTEI risk score demonstrated higher discrimination performance, with an AUC of 0.88 ± 0.05, compared to logistic EuroSCORE (AUC: 0.77 ± 0.05; p = 0.0001) and EuroSCORE II (AUC: 0.74 ± 0.05; p = 0.0005). APORTEI also showed better calibration performance (logistic EuroSCORE: p = 0.19; EuroSCORE II: p = 0.11; APORTEI: p = 0.56).²³

In a nutshell, the APORTEI score, created from a systematic review and validated in multicentric populations, outperforms EuroSCORE I but disagree with it when predicting in-hospital mortality following IE valve surgery.

COSTA score: Coronary surgery and transplantation association

The COSTA score is a system designed to assess the risk of in-hospital mortality for patients undergoing surgery for IE. It was developed by da Costa et al. in 2007 and was based on a retrospective study of 186 patients who had undergone surgery for IE.⁶¹ The study involved multivariate analysis to determine the main parameters independently associated with in-hospital mortality. The parameters used in the COSTA score are age ⩾40 years (OR: 4.16; 95% CI: 1.63–10.80 − four points), NYHA class IV or cardiovascular shock (OR: 4.93; 95% CI: 1.86–13.05 – five points), uncontrolled sepsis (OR = 5.97; 95% CI: 1.95–18.35 − six points), conduction disorder (OR: 5.07; 95% CI: 1.67–15.35 − five points), arrhythmia (OR: 8.17; 95% CI: 2.60–25.71 − eight points), valve with extensive damage or abscess or prosthesis (OR: 4.77; 95% CI: 1.44–15.76 − five points), and large mobile vegetation (OR: 4.36; 95% CI: 1.55–12.90 − four points). The study found that patients with a score of 0–10 had a mortality rate of 5.26%, which increased to 78.9% when the score exceeded 20 points. Subsequent studies have pointed out the limitations of the COSTA score. For example, Wang and Pemberton²⁶ evaluated the ability of several risk scores to predict in-hospital mortality in 146 patients undergoing surgery for IE and found that the COSTA score could not predict in-hospital mortality. Varela et al.⁶² also found that the COSTA score performed worse in preoperative risk assessment than three other scores when studying the outcomes of 180 patients who had undergone surgery for IE.⁶²

To recapitulate, the COSTA score was created in 2007 and is a scoring system for predicting in-hospital mortality in patients having surgery for IE. However, further research has questioned its predictive power and revealed limits compared to other scoring systems.

De Feo-Cotrufo score

The De Feo score was devised by De Feo et al.⁶³ in 2011 to estimate mortality risk in patients with IE undergoing surgery. Active and healed endocarditis were seen in 365 and 75 patients, respectively. Out of the total sample, 252 patients underwent surgery for IE with an overall operative mortality of 9.1%. Six independent predictors of mortality were identified – age, NYHA class IV, preoperative renal failure, preoperative mechanical ventilation, peri-valvular involvement (abscess or mycotic aneurysm), and latest blood cultures being positive before surgery. A prognostic scoring model was designed after assigning scores to these individual variables. The total score of 0 or 5 constituted class I or very low risk, 7–13 was class 2 or low risk, 14–19 was class 3 or high risk, and ⩾20 denoted class 4 or very high risk. In comparison with logistic EuroSCORE, the De Feo score had good discrimination power (AUC: 0.91; 95% CI: 0.85–0.97) as opposed to 0.84 (0.77–0.91) for the former. The small sample size was one of the limitations of this score design. Further, a smaller number of patients in healed endocarditis phase also added to this study’s limitation.

In conclusion, the De Feo score is a prognostic scoring model for assessing mortality risk in patients undergoing surgery for IE. It has been shown to have strong discrimination power when compared to the logistic EuroSCORE and has six independent predictors of death discovered.

EndoSCORE

EndoSCORE is a logistic risk model developed in 2017 and designed to predict early mortality risk after surgery for either native valve or prosthesis IE.⁶⁴ It uses some variables used in EuroSCORE but also considers perivalvular lesion along with different pathogens like Pseudomonas aeruginosa and fungal disease.⁶⁴ The score showed high discriminative power and good calibration.⁶⁵ However, a limitation of EndoSCORE is that it was developed and internally validated in a specific patient population and may not be as accurate in other patient groups.^64,66 Additionally, a comparative study by Gatti et al.⁶⁵ showed that it performed poorly in its accuracy of prediction and its goodness-of-fit compared to other similar risk models such as ANCLA and AEPEI.⁶⁵

In summary, EndoSCORE is a risk scoring system for predicting early mortality after surgery for IE. It takes into account factors like perivalvular lesions and particular pathogens, and it exhibits high discriminative power and good calibration.

ICE score

The ICE score was designed and validated by Park et al.⁶⁷ in 2016, using a large multinational sample, to predict 6-month mortality in patients with IE. The ICE is an 18-item score primarily categorized in four groups of weighted variables, including host characteristics (age category and history of dialysis), IE factors (nosocomial IE, prosthetic valve IE, symptoms duration >1 month before admission), Staphylococcus aureus-related factors (viridans type, presence of aortic/mitral vegetation), and IE complications (heart failure, stroke, paravalvular complication, and persistent bacteremia), in addition to surgical treatment. Approximately half of the patients in both derivation and validation samples were treated with surgery, which was found to be independently associated with mortality (similarly to other variables). Thus, the ICE score may be used to predict 6-month mortality following surgery for IE; however, it is only aimed at this one specific outcome.⁶⁷

Overall, the ICE score is a validated 18-item scoring system that takes into account a variety of host features, IE variables, factors associated to Staphylococcus aureus, IE sequelae, and surgical treatment to predict 6-month mortality specifically after surgery for IE.

Meta-model

Meta-model is an aggregation of all the prognostic models available which predict mortality in patients undergoing surgery for IE.⁶⁸ Out of the 11 prognostic models across nine studies, three models (EndoSCORE, Specific EuroSCORE I, Specific EuroSCORE II) were used in the construction of the meta-model because of low/unclear risk of bias. The components of the model included age, gender, renal failure (creatinine >2 mg/dL), chronic pulmonary disease, pulmonary hypertension (systolic pulmonary artery pressure >60 mm Hg), prior cardiac surgery, LVEF, critical preoperative state, NYHA classification, abscess, fistula, urgency of the procedure, valves treated, etiology of the infection, and valve location. The validation was performed in the GAMES (Grupo de Apoyo al Manejo de la Endocarditis infecciosa en Espana) registry. The C-statistics of the meta-model (0.79; 95% CI: 0.76–0.82) was better than all the three individual models [EndoSCORE: 0.759 (0.731–0.788), specific EuroSCORE I: 0.758 (0.731–0.786), specific EuroSCORE II (0.762 (0.735–0.789)]. The sensitivity analysis of the meta-model also demonstrated better performance than the three models. The model had its own limitations. The validation data set had either 30-day or in-hospital mortality as outcome, but the outcomes in all the three models used for construction of the model was 30-day mortality. Out of the 11 prediction models identified, only three could be used because of the low risk of bias.⁶⁸

Overall, age, gender, comorbidities, and procedural parameters are included in the EndoSCORE, specific EuroSCORE I, and specific EuroSCORE II prognostic models that make up the meta-model for predicting mortality in IE surgery, which has greater predictive performance than the individual models.^69–72

PALSUSE score

PALSUSE is an acronym for seven variables, prosthetic valve, age >70, large intracardiac destruction, Staphylococcus species, urgent surgery, female gender, EuroSCORE > 10, and has been validated in its operative risk assessment, predicting in-hospital mortality, and perioperative complications in patients with IE. However, unlike other scores, it does not account for heart failure, sepsis, conduction abnormalities, or kidney failure.⁶⁵ Compared to other models, such as STS and De Feo-Cotrufo, PALSUSE has a fair accuracy. It could not predict long-term prognosis; however, it was able to predict perioperative complications, such as acute kidney injury.²⁶ Furthermore, compared to other scores, such as the EuroSCORE II, the PALSUSE score has a non-satisfactory discriminatory power but with good calibration, which reflects prognostic accuracy⁷³ as shown in the multicenter study in Spain, including 1000 individuals with IE.⁷³ When the PALSUSE was above 3, it was associated with in-hospital mortality of up to 45.4%.

In conclusion, PALSUSE is a helpful score with a prognostic value that could be used to predict surgical mortality in patients with IE and could be used to aid decision-making; however, it is not the gold standard due to the limitations mentioned above.⁷⁴

RISK-E score

Surgical scores are widely used in cardiac surgery to select the best strategy in valve heart disease or coronary revascularization. However, these scores were neither specific nor accurate for IE. The RISK-E score was developed to estimate the surgical risk of in-hospital mortality in patients with active left-sided IE. Olmos et al.⁷⁵ developed and validated a calculator to predict the risk of in-hospital mortality in patients with active left-sided endocarditis undergoing cardiac surgery. The study includes 2299 patients with IE who were analyzed. The RISK-E score for IE considers a variety of patient factors that are known to influence the risk of adverse events, such as age, prosthetic endocarditis, virulent microorganism, septic shock, thrombocytopenia, acute renal insufficiency, cardiogenic shock, and peri annular complications. The RISK-E score of each patient is calculated by adding up the points according to the presence of each risk factor. Therefore, if the patient has no variables in the score, the risk score is 0, and in patient with eight variables in the score, the total score is 68 and the predicted probability of postoperative mortality ranged from 3% for a score of 0 to 97% for a patient with highest score of 68. As a result of this study, the score showed a good discrimination, with an area under the ROC of 0.77 (95% CI: 0.71–0.82). RISK-E score gained good predictive performance, AUC: 0.76 (95% CI: 0.64–0.88) in the external validation. The reason behind the RISK-E score is regarding infectious factors, this includes one of the most important variables that is infection due to microorganisms. Staphylococcus aureus and fungi have been recognized as independent predictors of mortality. Interestingly, microorganisms are included in a few other scores as well such as PALSUSE, EndoSCORE, ICE, APORTEI, and STratification risk analysis in OPerative management (STOP) score and a prognostic score model created by Martins and da Cruz Lamas,⁷⁶ where non-HACEK organisms were associated with mortality. This model, like any other predictive model, has limitations. First, because of the large number of cases studied retrospectively, it has the potential bias inherent in observational studies. Due to lack of data, the retrospective estimation of this score was not done and so STS score was not calculated.^75,77

In conclusion, RISK-E score is a validated scoring system for determining the surgical risk of in-hospital mortality in patients with active left-sided IE that takes into account a number of patient factors.

SHARPEN

SHARPEN is a risk score that comprises variables that can predict in-hospital mortality both in surgical and non-surgical patients with IE. The SHARPEN score includes systolic blood pressure <90 mm Hg at presentation, heart failure, age (with varying score according to the age group: <50, 50–65, and >65 years), renal function (>2.26 mg/dL), pneumonia (⩾48 h after admission), elevated peak CRP (>200 mg/L during hospitalization), and non-intravenous drug abuser.⁷⁸ This score was recently validated, in a retrospective study that included 179 hospitalized patients for IE, with variable presentation and complications. This study found a sensitivity of 70.0% and specificity of 71.2%, positive predictive value of 41.1%, and negative predictive value of 89.2% with a score of 10 as a cut-off. A SHARPEN score >10 was an independent predictor of mortality. Including both surgical and non-surgical cohort, the score had a good performance (AUC: 0.76; 95% CI: 0.67–0.85; p < 0.001).⁷⁸ Another study by Chee et al., through an 11-year study, also independently correlated with mortality. The score was stratified into low risk (2–6), moderate risk (7–10), and high risk (11–20) groups, with a good predictive performance (AUC: 0.86; 95% CI: 0.80–0.91; p < 0.001). All patients with score ⩾15 had died. This needs to be interpreted keeping in mind the small sample size of the cohort.⁷⁹

In summary, with good predictive performance and a cutoff score of 10 or higher indicating increased mortality risk, the SHARPEN score is a validated risk scoring system for predicting in-hospital mortality in patients with IE.

STS score for IE

Gaca et al.²⁵ developed a risk score using the STS database with more specific sub scores including variables unique to IE. The need for a dedicated stratification tool that can be used in both preoperative patient information to identify risk factors that affect operative morbidity and mortality in IE surgery and in bedside decision-making processes. The proposed risk score system is a simplified model and consists of a total of 13 important variables, which include prior CABG, urgent or emergency, no cardiogenic shock, emergency, salvage, or cardiogenic shock, preoperative intra-aortic balloon pump (IABP) or inotropes, multiple valve procedure, prior valve surgery, insulin-dependent diabetes mellitus (IDDM), non-insulin dependent diabetes mellitus, hypertension, chronic lung disease, active endocarditis, renal failure or creatinine >2.0, and arrhythmia. Furthermore, the 14 risk factors for the composite end point of major morbidity and mortality which include female gender, BSA > 1.9, age >60, prior CABG, urgent or emergency, no cardiogenic shock, emergency, salvage, or cardiogenic shock, preoperative IABP or inotropes, multiple valve procedure, prior valve surgery, IDDM, NYHA class IV, active endocarditis, renal failure or creatinine >2.0, arrhythmia. Each risk factor was assigned a point score based on its predictive value. The most important predictor of the composite end point of death or major complication was an operation classified as emergency, emergency/salvage, or cardiogenic shock. The second most important predictor value was renal failure or a preop creatinine level >2.0 mg/dL.²⁵ The results of the simplified model predicting the composite end point of major morbidity and mortality demonstrated good predictive ability compared with observed mortality with a C-statistic of 0.729. Likewise, the simplified model end point estimate for postoperative mortality demonstrated good predictive ability compared with observed mortality with a C-statistic of 0.758.^25,43,47 The STS models are used in patients undergoing isolated CABG or valve surgery and may not fully account for the unique characteristics of IE patients. Most of the studies have reported the relatively small size of the sample as one of the limitations and larger studies are required for external validation.^26,57,63,71

In summary, using the STS database, Gaca et al. created a simplified risk score system for IE surgery that considered 14 major risk factors for morbidity and mortality as well as 13 important variables for postoperative mortality. The system showed good predictive ability but had some drawbacks, such as small sample sizes and the need for larger external validation studies.

STOP score

The STOP score is focused on morbidity and mortality prediction in patients with drug-induced endocarditis. It was developed by Habertheuer et al.⁸⁰ in 2021 in a multicenter study, in response to the opioid epidemic and the lack of predictive tools for this specific, notably younger patient population. Morbidity is defined as reintubation, prolonged ventilation, pneumonia, renal failure, dialysis, stroke, reoperation for bleeding, and the need for a permanent pacemaker. The STOP score considers seven variables: dialysis, IE treatment status, lung disease at baseline, emergency, prosthetic valve IE, aortic valve procedure, and number of valves affected. It is particularly optimized for patients with drug-related IE, which tend to have different demographic characteristics (i.e. younger age), making it a more specialized tool. However, while it does account for factors related to surgery, further analysis is needed to assess whether it would be superior to other scores in predicting post-surgical outcomes in this subset of patients.⁸⁰ Overall, STOP score is a specialized predictive tool designed for drug-induced endocarditis that considers factors related to morbidity and mortality in this particular patient population. However, more research is required to determine whether it is more accurate than other scores at predicting post-surgical outcomes.

Sun Yat-sen University Prediction Model for IE

The Sun Yat-sen University Prediction Model for IE (SYSUPMIE) risk model is a machine learning-based tool for risk stratification in patients with IE who meet the definite modified Duke criteria.⁸¹ It is a quick and practical method for assessing the risk of patients with IE, utilizing eight variables to aid in risk stratification. These variables include heart failure, platelet count, diastolic function, tricuspid valve involvement, presence of urine occult blood, serum albumin, vegetation size greater than or equal to 10 mm, and the involvement of multiple valves. In a study involving 276 patients admitted with IE, SYSUPMIE demonstrated good accuracy and generalizability, with an AUC of 0.812. Compared to other standard scoring systems, it correlated well and was shown to predict early mortality after surgery for IE accurately. One notable advantage of the SYSUPMIE model is its ability to account for left- and right-side involvement, native valve endocarditis, prosthetic valve endocarditis, and acute and non-acute presentations. However, there are a few limitations to the SYSUPMIE model. Age was not included as a variable, which could be a drawback as older adults are at higher risk for poor outcomes after surgery. Additionally, patients receiving albumin and platelet transfusion therapy may experience inaccuracies in the scoring. Calibration of the score was not performed given the sample size of the cohort. Despite these limitations, the study demonstrated that the SYSUPMIE model is useful for decision-making and perioperative management in patients with IE, ultimately leading to improved treatment outcomes. Compared to other scoring systems like EuroSCORE II and STS, it is easier to perform and bypass several limitations in previous scoring systems with having a higher AUC (0.810–0.813).⁸¹

In conclusion, SYSUPMIE risk model is a machine learning-based tool for risk stratification in patients with IE. It uses eight variables to accurately predict early mortality and has shown good accuracy and generalizability. However, it has some restrictions, including the exclusion of age as a variable and potential inaccuracies in patients receiving specific therapies.

Other scores utilized in IE

MELD XI score

The MELD (model for end-stage liver disease) incorporates blood levels of creatinine, total bilirubin, and the international normalized ratio (INR). It was developed by the United Network for Organ Sharing to rank patients’ liver transplants. It has been shown to correlate with a patient’s prognosis in end-stage liver failure, liver transplant, and transjugular intrahepatic portosystemic.⁶⁹ Individuals with heart failure, in certain instances, might suffer from hepatic congestion and dysfunction. Hence the MELD-XI (excludes INR) was thought to be used in these patients, revealing its prognostic value.⁷⁰ Moreover, recently, the MELD-XI score has been studied in an observational study of 858 patients with IE, to assess its usefulness in determining the prognosis, knowing that one of the most common complications of IE includes heart failure, which usually requires surgical management. First, a high MELD-XI score of >10 was associated with higher NYHA class, valvular abnormalities, worse kidney function, and lower surgical utilization in this group. Second, it correlated with MACEs, predicting in-hospital and long-term mortality. Furthermore, adding NYHA class, elevated CRP, and non-surgical management, the score was of a higher power.

In conclusion, this score could be utilized in risk assessment in patients with IE.⁷¹

Clinical application of risk scores

In up to 30% IE patients, medical management alone is insufficient, and cardiac surgery is required.^7,74 While there are clear guidelines for the indications for surgery, in practice, this decision is far more complex and nuanced and requires personalization on a case-by-case basis.¹³ Postoperative mortality rates for IE can be high, reaching up to 28.8% for in-hospital mortality, 26.82% for 30-day mortality, and 17.5% for 6-month mortality post-operatively.^22,62,82 Yet timely surgery has known survival benefit, and up to one-fourth of patients with indications for surgery do not end up receiving it.^13,22,83,84 Risk scores assist in estimating the operative mortality and guides clinicians to determine whether benefits outweigh the risk of surgery. From a patient’s perspective, it helps them understand their risk of surgery in numbers. In high-risk cases, it allows for more thorough and aggressive perioperative management. Additionally, risk scores permit an objective comparison of risks between different patients and different procedures. General scores are unable to incorporate endocarditis-specific parameters as they were not routinely collected in general cardiac surgery databases. The endocarditis-specific risk scores are more likely to include endocarditis parameters like microorganism, intracardiac complication/abscess, and other detailed imaging parameters. Tables 1 and 2 describe the study characteristics describing the scores and all the score parameters used to derive the score, respectively.

The main risk score parameters are common to many of the scores, such as age, gender, NYHA class, critical preoperative state, and left ventricle (LV) dysfunction. The major advantage of endocarditis-specific risk scores is that they are derived from endocarditis surgeries and incorporate endocarditis-specific parameters such as microorganism and intracardiac complications. The general risk scores like EuroSCORE and STS score derivation cohort had only a fraction of patients (<10%) undergoing endocarditis surgery so they may not always be applicable to endocarditis surgery, but they usually have much larger size for derivation samples.

Limitations

While risk scores serve as useful tools for risk stratification, aiding clinical decision-making, and for counseling patients; they have inherent limitations. Each risk score is derived from cardiac surgery in specific demographics and settings. Therefore, it needs to be externally validated. There are many factors that are not included in the current models but also need to be considered such as clinical (embolic events, liver disease, frailty), imaging (right ventricular function, right ventricular systolic pressure, left atrial size, endocarditis complications), laboratory, and perioperative parameters. Risk models should be derived and evaluated in a way that they can also help in predicting other outcomes beyond operative mortality, such as in-hospital morbidities and long-term mortality. Many scores are reported as additive models, rather than logistic such as EuroSCORE and STS. Additive models can only assess discrimination (AUC) but not calibration. Therefore, more logistic score models are needed to provide accurate estimation of operative mortality.

Another limitation of our study is that it did not involve a systematic review. Further, several cross-validation analyses have found that even though the scores have good discriminatory and comparative powers, the expected mortality they predict is higher than the observed mortality in the actual patient cohort. Use of these risk scores without case-by-case individualization by a highly trained Endocarditis Team, has likely led to unwarranted denial of surgery. One set of researchers have hypothesized that the presence of an Endocarditis Team, and undergoing timely cardiac surgery, may be used as correction factors for risk scores predicting mortality, as both strategies may be associated with lower risk of death.

Conclusion

Risk scores are an important objective tool for estimating risk of adverse outcomes in cardiac surgery, and this review has discussed all the contemporary risk scores for endocarditis surgery. The scores have not only critical roles in clinical decision-making, but also in epidemiological analyses, and in counseling patients and their families about prognosis and expectations. This is especially important for high-risk procedures such as endocarditis surgery when often both the risk of operating and not operating are high. Both general and endocarditis-specific risk scores have been reviewed, each with shared but somewhat different parameters. Further external validation and refinement of risk scores are necessary before widespread clinical use, especially for endocarditis-specific scores. Risk scores together with the specialized Endocarditis Team evaluation in an individualized approach are crucial in clinical decision-making, including prevention of both futile interventions and inappropriate denial of surgery.

Supplemental Material

sj-doc-2-tak-10.1177_17539447231193291 – Supplemental material for Contemporary risk models for infective endocarditis surgery: a narrative review

Supplemental material, sj-doc-2-tak-10.1177_17539447231193291 for Contemporary risk models for infective endocarditis surgery: a narrative review by Ankit Agrawal, Aro Daniela Arockiam, Yasser Jamil, Joseph El Dahdah, Bianca Honnekeri, Michel Chedid El Helou, Joseph Kassab and Tom Kai Ming Wang in Therapeutic Advances in Cardiovascular Disease

Supplemental Material

sj-docx-1-tak-10.1177_17539447231193291 – Supplemental material for Contemporary risk models for infective endocarditis surgery: a narrative review

Supplemental material, sj-docx-1-tak-10.1177_17539447231193291 for Contemporary risk models for infective endocarditis surgery: a narrative review by Ankit Agrawal, Aro Daniela Arockiam, Yasser Jamil, Joseph El Dahdah, Bianca Honnekeri, Michel Chedid El Helou, Joseph Kassab and Tom Kai Ming Wang in Therapeutic Advances in Cardiovascular Disease

Footnotes

Acknowledgements

None.

Declarations

ORCID iDs

Ankit Agrawal

Joseph El Dahdah

Michel Chedid El Helou

Supplemental material

Supplemental material for this article is available online.

References

Liesenborghs

Meyers

Lox

, et al. Staphylococcus aureus endocarditis: distinct mechanisms of bacterial adhesion to damaged and inflamed heart valves. Eur Heart J 2019; 40: 3248–3259.

Lerche

Schwartz

Theut

, et al. Anti-biofilm approach in infective endocarditis exposes new treatment strategies for improved outcome. Front Cell Dev Biol 2021; 9: 643335.

Kouijzer

JJP

Noordermeer

van Leeuwen

, et al. Native valve, prosthetic valve, and cardiac device-related infective endocarditis: a review and update on current innovative diagnostic and therapeutic strategies. Front Cell Dev Biol 2022; 10: 995508.

Chen

Zhan

Zhang

, et al. The global, regional, and national burden and trends of infective endocarditis from 1990 to 2019: results from the Global Burden of Disease Study 2019. Front Med 2022; 9: 349.

Bor

Woolhandler

Nardin

, et al. Infective endocarditis in the U.S., 1998–2009: a nationwide study. PLoS One 2013; 8: e60033.

McLaughlin

Doulamis

Briasoulis

, et al. Increasing mortality rates from infective endocarditis among young US residents. J Intern Med 2023; 293: 262–263.

Murdoch

Corey

Hoen

, et al. Clinical presentation, etiology and outcome of infective endocarditis in the 21st century: the International Collaboration on Endocarditis-Prospective Cohort Study. Arch Intern Med 2009; 169: 463.

Habib

Erba

Iung

, et al. Clinical presentation, aetiology and outcome of infective endocarditis. Results of the ESC-EORP EURO-ENDO (European infective endocarditis) registry: a prospective cohort study. Eur Heart J 2019; 40: 3222–3232.

Verzelloni Sef

Jaggar

Trkulja

, et al. Factors associated with long-term outcomes in adult congenital heart disease patients with infective endocarditis: a 16-year tertiary single-centre experience. Eur J Cardio-Thorac Surg 2023; 63: ezad105.

10.

Mocchegiani

Nataloni

. Complications of infective endocarditis. Cardiovasc Hematol Disord Drug Targets 2009; 9: 240-248.

11.

Heiro

Nikoskelainen

Engblom

, et al. Neurologic manifestations of infective endocarditis: a 17-year experience in a teaching hospital in Finland. Arch Intern Med 2000; 160: 2781–2787.

12.

García-Cabrera

Fernández-Hidalgo

Almirante

, et al. Neurological complications of infective endocarditis risk factors, outcome, and impact of cardiac surgery: a multicenter observational study. Circulation 2013; 127: 2272–2284.

13.

Habib

Lancellotti

Antunes

, et al. 2015 ESC guidelines for the management of infective endocarditis: the Task Force for the Management of Infective Endocarditis of the European Society of Cardiology (ESC). Endorsed by: European Association for Cardio-Thoracic Surgery (EACTS), the European Association of Nuclear Medicine (EANM). Eur Heart J 2015; 36: 3075–3123.

14.

Shah

ASV

McAllister

Gallacher

, et al. Incidence, microbiology, and outcomes in patients hospitalized with infective endocarditis. Circulation 2020; 141: 2067–2077.

15.

Pettersson

Coselli

Pettersson

, et al. 2016 the American Association for Thoracic Surgery (AATS) consensus guidelines: surgical treatment of infective endocarditis: executive summary. J Thorac Cardiovasc Surg 2017; 153: 1241–1258.e29.

16.

Baddour

Wilson

Bayer

, et al. Infective endocarditis in adults: diagnosis, antimicrobial therapy, and management of complications: a scientific statement for healthcare professionals from the American Heart Association. Circulation 2015; 132: 1435–1486.

17.

Vikram

Buenconsejo

Hasbun

, et al. Impact of valve surgery on 6-month mortality in adults with complicated, left-sided native valve endocarditis: a propensity analysis. JAMA 2003; 290: 3207–3214.

18.

Kang

Kim

, et al. Early surgery versus conventional treatment for infective endocarditis. New Engl J Med 2012; 366: 2466–2473.

19.

Tornos

Iung

Permanyer-Miralda

, et al. Infective endocarditis in Europe: lessons from the Euro heart survey. Heart 2005; 91: 571–575.

20.

Jault

Gandjbakhch

Rama

, et al. Active native valve endocarditis: determinants of operative death and late mortality. Ann Thorac Surg 1997; 63: 1737–1741.

21.

Alexiou

Langley

Stafford

, et al. Surgery for active culture-positive endocarditis: determinants of early and late outcome. Ann Thorac Surg 2000; 69: 1448–1454.

22.

Chu

Park

Athan

, et al. Association between surgical indications, operative risk, and clinical outcome in infective endocarditis: a prospective study from the International Collaboration on Endocarditis. Circulation 2015; 131: 131–140.

23.

Urso

Tena

MÁ

Horcajada

, et al. Prediction of surgical risk in patients with endocarditis: comparison of logistic EuroSCORE, EuroSCORE II and APORTEI score. Enfermedades Infecciosas y Microbiologia Clinica (English Ed.) 2021; 39: 244–247.

24.

Varela Barca

Fernández-Felix

Navas Elorza

, et al. Prognostic assessment of valvular surgery in active infective endocarditis: multicentric nationwide validation of a new score developed from a meta-analysis. Eur J Cardio-Thorac Surg 2020; 57: 724–731.

25.

Gaca

Sheng

Daneshmand

, et al. Outcomes for endocarditis surgery in North America: a simplified risk scoring system. Internet J Thorac Cardiovasc Surg 2011; 141: 98–106.e2.

26.

Wang

TKM

Pemberton

. Performance of endocarditis-specific risk scores in surgery for infective endocarditis. Thorac Cardiovasc Surg 2018; 66: 333–335.

27.

Wang

TKM

Bin Saeedan

Chan

, et al. Complementary diagnostic and prognostic contributions of cardiac computed tomography for infective endocarditis surgery. Circ Cardiovasc Imaging 2020; 13: e011126.

28.

Michel

Roques

Nashef

. Logistic or additive EuroSCORE for high-risk patients? Eur J Cardio-Thorac Surg 2003; 23: 684–687.

29.

Gogbashian

Sedrakyan

Treasure

. EuroSCORE: a systematic review of international performance. Eur J Cardio-Thorac Surg 2004; 25: 695–700.

30.

Roques

Nashef

SAM

Michel

, et al. Does EuroSCORE work in individual European countries? Eur J Cardio-Thorac Surg 2000; 18: 27–30.

31.

Nashef

SAM

Roques

Hammill

, et al. Validation of European System for Cardiac Operative Risk Evaluation (EuroSCORE) in North American cardiac surgery. Eur J Cardio-Thorac Surg 2002; 22: 101–105.

32.

Parolari

Pesce

Trezzi

, et al. EuroSCORE performance in valve surgery: a meta-analysis. Ann Thorac Surg 2010; 89: 787–793.e2.

33.

Watanabe

Hayashida

Lefèvre

, et al. Is EuroSCORE II better than EuroSCORE in predicting mortality after transcatheter aortic valve implantation? Catheter Cardiovasc Interv 2013; 81: 1053–1060.

34.

Hassanin

Afify

Zook

, et al. Infective endocarditis after transcatheter aortic valve implantation: a systematic review. Cardiol Rev 2023; 31: 93–98.

35.

Holmes

Patel

, et al. Comparison of EuroSCORE II, original EuroSCORE, and the Society of Thoracic Surgeons risk score in cardiac surgery patients. Ann Thorac Surg 2016; 102: 573–579.

36.

Shapira-Daniels

Blumenfeld

Korach

, et al. The American Society of Thoracic Surgery Score versus EuroSCORE I and EuroSCORE II in Israeli patients undergoing cardiac surgery. Isr Med Assoc J 2019; 21: 671–675.

37.

Nashef

SAM

Roques

Sharples

, et al. EuroSCORE II. Eur J Cardio-Thorac Surg 2012; 41: 734–745.

38.

Silverborn

Nielsen

Karlsson

. The performance of EuroSCORE II in CABG patients in relation to sex, age, and surgical risk: a nationwide study in 14,118 patients. J Cardiothorac Surg 2023; 18: 40.

39.

Biancari

Juvonen

Onorati

, et al. Meta-analysis on the performance of the EuroSCORE II and the Society of Thoracic Surgeons scores in patients undergoing aortic valve replacement. J Cardiothorac Vasc Anesth 2014; 28: 1533–1539.

40.

Guida

Mastro

Scrascia

, et al. Performance of the European system for Cardiac Operative Risk Evaluation II: a meta-analysis of 22 studies involving 145,592 cardiac surgery procedures. J Thorac Cardiovasc Surg 2014; 148: 3049–3057.e1.

41.

Prins

De Villiers Jonker

Botes

, et al. Cardiac surgery risk-stratification models. Cardiovasc J Afr 2012; 23: 160–164.

42.

Granton

Cheng

. Risk stratification models for cardiac surgery. Semin Cardiothorac Vasc Anesth 2008; 12: 167–174.

43.

Jin

Furnary

Fine

, et al. Using society of thoracic surgeons risk models for risk-adjusting cardiac surgery results. Ann Thorac Surg 2010; 89: 677–682.

44.

Shahian

O’Brien

Filardo

, et al. The Society of Thoracic Surgeons 2008 cardiac surgery risk models: Part 3—Valve plus coronary artery bypass grafting surgery. Ann Thorac Surg 2009; 88: S43–S62.

45.

Puskas

Kilgo

Thourani

, et al. The Society of Thoracic Surgeons 30-day predicted risk of mortality score also predicts long-term survival. Ann Thorac Surg 2012; 93: 26–35.

46.

Shahian

Jacobs

Badhwar

, et al. The Society of Thoracic Surgeons 2018 adult cardiac surgery risk models: Part 1—Background, design considerations, and model development. Ann Thorac Surg 2018; 105: 1411–1418.

47.

Jacobs

Shahian

O’Brien

, et al. Assessing the performance of risk models with discrimination and calibration. Ann Thorac Surg 2023; 115: 282–286.

48.

Grant

Collins

Nashef

SAM

. Statistical primer: developing and validating a risk prediction model^†. Eur J Cardio-Thorac Surg 2018; 54: 203–208.

49.

Jacobs

Shahian

Badhwar

, et al. The Society of Thoracic Surgeons 2021 adult cardiac surgery risk models for multiple valve operations. Ann Thorac Surg 2022; 113: 511–518.

50.

Hemmann

Sirotina

De Rosa

, et al. The STS score is the strongest predictor of long-term survival following transcatheter aortic valve implantation, whereas access route (transapical versus transfemoral) has no predictive value beyond the periprocedural phase. Interact Cardiovasc Thorac Surg 2013; 17: 359–364.

51.

Holmes

Rongione

, et al. The long-term safety and efficacy of concomitant Cox maze procedures for atrial fibrillation in patients without mitral valve disease. J Thorac Cardiovasc Surg 2019; 157: 1505–1514.

52.

Wei

Liu

, et al. Age, creatinine and ejection fraction (ACEF) score: a simple risk-stratified method for infective endocarditis. QJM 2019; 112: 900–906.

53.

Barili

Pacini

Grossi

, et al. Reliability of new scores in predicting perioperative mortality after mitral valve surgery. J Thorac Cardiovasc Surg 2014; 147: 1008–1012.

54.

Wykrzykowska

Garg

Onuma

, et al. Value of age, creatinine, and ejection fraction (ACEF score) in assessing risk in patients undergoing percutaneous coronary interventions in the “All-Comers” LEADERS trial. Circ Cardiovasc Interv 2011; 4: 47–56.

55.

Stähli

Wischnewsky

Jakob

, et al. Predictive value of the age, creatinine, and ejection fraction (ACEF) score in patients with acute coronary syndromes. Int J Cardiol 2018; 270: 7–13.

56.

Capodanno

Marcantoni

Ministeri

, et al. Incorporating glomerular filtration rate or creatinine clearance by the modification of diet in renal disease equation or the Cockcroft–Gault equations to improve the global accuracy of the Age, Creatinine, Ejection Fraction [ACEF] score in patients undergoing percutaneous coronary intervention. Int J Cardiol 2013; 168: 396–402.

57.

Gatti

Perrotti

Obadia

, et al. Simple scoring system to predict in-hospital mortality after surgery for infective endocarditis. J Am Heart Assoc 2017; 6: e004806.

58.

Gatuz

Cuenza

. AEPEI scoring system to predict mortality among patients diagnosed with infective endocarditis after surgery at the Philippine Heart Center: a retrospective analytical cross-sectional study. Eur Heart J 2022; 43: ehab849.108.

59.

Gatti

Benussi

Gripshi

, et al. A risk factor analysis for in-hospital mortality after surgery for infective endocarditis and a proposal of a new predictive scoring system. Infection 2017; 45: 413–423.

60.

Fernández-Cisneros

Hernández-Meneses

Llopis

, et al. Risk scores’ performance and their impact on operative decision-making in left-sided endocarditis: a cohort study. Eur J Clin Microbiol Infect Dis 2023; 42: 33–42.

61.

da Costa

MAC

Wollmann

Campos

ACL

, et al. Risk index for death by infective endocarditis: a multivariate logistic model. Rev Bras Cir Cardiovasc 2007; 22: 192–200.

62.

Varela

López-Menéndez

Redondo

, et al. Mortality risk prediction in infective endocarditis surgery: reliability analysis of specific scores^†. Eur J Cardio-Thorac Surg 2018; 53: 1049–1054.

63.

De Feo

Cotrufo

Carozza

, et al. The need for a specific risk prediction system in native valve infective endocarditis surgery. Sci World J 2012; 2012: 1–8.

64.

Di Mauro

Dato

GMA

Barili

, et al. A predictive model for early mortality after surgical treatment of heart valve or prosthesis infective endocarditis. The EndoSCORE. Int J Cardiol 2017; 241: 97–102.

65.

Gatti

Sponga

Peghin

, et al. Risk scores and surgery for infective endocarditis: in search of a good predictive score. Scand Cardiovasc J 2019; 53: 117–124.

66.

Varela Barca

López-Menéndez

Redondo Palacios

, et al. External validation of the new predictive model for early mortality after surgical treatment of infective endocarditis: analysis of the reliability of EndoSCORE. Int J Cardiol 2018; 268: 99.

67.

Park

Chu

Peterson

, et al. Validated risk score for predicting 6-month mortality in infective endocarditis. J Am Heart Assoc 2016; 5: e003016.

68.

Fernandez-Felix

Barca

Garcia-Esquinas

, et al. Prognostic models for mortality after cardiac surgery in patients with infective endocarditis: a systematic review and aggregation of prediction models. Vol. 27, Clinical Microbiology and Infection. Madrid: Elsevier B.V, 2021. pp.1422–1430.

69.

Singal

Kamath

. Model for end-stage liver disease. J Clin Exp Hepatol 2013; 3: 50–60.

70.

Lin

Liu

Xiao

, et al. The MELD-XI score predicts 3-year mortality in patients with chronic heart failure. Front Cardiovasc Med 2022; 9: 985503.

71.

Wei

Luo

, et al. Risk prediction in infective endocarditis by modified MELD-XI score. Eur J Clin Microbiol Infect Dis 2018; 37: 1243–1250.

72.

Jaglal

Naylor

. Multicenter validation of a risk index for mortality, intensive care unit stay, and overall hospital length of stay after cardiac surgery. Steering Committee of the Provincial Adult Cardiac Care Network of Ontario. Circulation 1995; 91: 677–684.

73.

Gatti

Chocron

Obadia

, et al. Using surgical risk scores in nonsurgically treated infective endocarditis patients. Hellenic J Cardiol 2020; 61: 246–252.

74.

Martínez-Sellés

Muñoz

Arnáiz

, et al. Valve surgery in active infective endocarditis: a simple score to predict in-hospital prognosis. Int J Cardiol 2014; 175: 133–137.

75.

Olmos

Vilacosta

Habib

, et al. Risk score for cardiac surgery in active left-sided infective endocarditis. Heart 2017; 103: 1435–1442.

76.

Martins

ABB

da Cruz Lamas

. Prognostic scores for mortality in cardiac surgery for infective endocarditis. Arq Bras Cardiol 2020; 114: 525–529.

77.

Lopez Díaz

Fernandez-Hidalgo

Revilla

, et al. Internal and external validation of a model to predict adverse outcomes in patients with left-sided infective endocarditis. Heart 2011; 97: 1138–1142.

78.

Alves

Pivatto Júnior

Filippini

, et al. SHARPEN score accurately predicts in-hospital mortality in infective endocarditis. Eur J Intern Med 2021; 92: 124–127.

79.

Chee

Tan

Ngiam

, et al. The SHARPEN clinical risk score predicts mortality in patients with infective endocarditis: an 11-year study. Int J Cardiol 2015; 191: 273–276.

80.

Habertheuer

Geirsson

Gleason

, et al. STratification risk analysis in OPerative management (STOP score) for drug-induced endocarditis. J Card Surg 2021; 36: 2442–2451.

81.

Luo

Huang

Liu

, et al. Machine learning-based risk model for predicting early mortality after surgery for infective endocarditis. J Am Heart Assoc 2022; 11: e025433.

82.

Muñoz

Kestler

De Alarcon

, et al. Current epidemiology and outcome of infective endocarditis: a multicenter, prospective, cohort study. Medicine 2015; 94: e1816.

83.

Anantha Narayanan

Mahfood Haddad

Kalil

, et al. Early versus late surgical intervention or medical management for infective endocarditis: a systematic review and meta-analysis. Heart 2016; 102: 950–957.

84.

Carino

Fernández-Cisneros

Hernández-Meneses

, et al. The fate of active left-side infective endocarditis with operative indication in absence of valve surgery. J Card Surg 2020; 35: 3034–3040.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.16 MB

0.01 MB