A new mortality prediction calculator for patients with infective endocarditis could save lives

A new mortality risk calculator for patients with infective endocarditis undergoing cardiac surgery allows for a more accurate prediction of possible postoperative complications, thanks to a study led by researchers from the Epidemiology and Public Health Networking Biomedical Research Centre (CIBERESP) and the Clinical Biostatistics Unit of the Ramón y Cajal University Hospital (IRYCIS), which has been published in the journal Clinical Microbiology and Infection (CMI). Vall d’Hebron University Hospital also participated in the study.

The calculator is available to clinicians through a free-to-access online platform which allows them to identify patients with a high risk of mortality, supporting their decision-making and helping them to ensure a better care and surgical management of these patients.

Currently, there is a great interest in understanding the prognosis of patients who may need to undergo surgery for infective endocarditis: a rare disease associated with high morbidity and mortality rate whose management is often complex. The treatment of choice for a large number of patients with infective endocarditis (20-50%) is surgery; however, the mortality rate associated with cardiac surgery remains high.

Finding the balance between the risk and benefit of the procedure

Cardiac surgery for infective endocarditis consists of the excision and resection of the tissues affected by inflammation and infection, allowing for the cardiac reconstruction of the damaged areas. Alongside antibiotic treatment, this contributes to the eradication of the causative microorganisms and reduces the morbidity and mortality rates associated with the disease. However, the decision to carry out surgery for infective endocarditis remains a challenge in clinical practice, and must only be taken after carefully weighing up the risks and benefits of this procedure. While surgery may be the definitive treatment for this disease, it represents a challenge for the cardiac surgeon due to its high complexity.

In our study, the research team identified 11 different prediction models for postoperative mortality that have been developed over the course of the last decade. They analysed both the design and the methodological quality of the studies included in the systematic review, as well as the risk of unreliable results. It was found that eight of the eleven models presented shortcomings in their statistical methods, or were developed using a very limited sample size. This suggests that these models may be unreliable and that their predictions may not be accurate when applied to new patients. As such, they could lead to incorrect healthcare decisions in clinical practice.

On this issue, Núria Fernández Hidalgo, specialist in the Infectious Diseases Department of Vall d'Hebron University Hospital and researcher in the Infectious Diseases Research Group of the Vall d'Hebron Research Institute (VHIR), and Laura Varela, surgeon at the Fundació Jiménez Díaz University Hospital, stated: “The estimation of risk in patients suffering from infective endocarditis represents one of the most challenging issues in the clinical management of these complex patients. While risk models may not alter the clinical decision-making process in terms of whether or not to offer an operation to a patient, they are very useful for the decision-making in endocarditis teams, providing quality assurance and helping to inform patients and their families”.

Development of a meta-model using the national GAMES registry

Researchers on this study developed a meta-model by applying regression techniques that permit the aggregation of results from existing models using data from the GAMES (Support Group for the Management of Infective Endocarditis in Spain) national endocarditis registry. For the present study, 1,453 infective episodes of adult patients undergoing cardiac surgery with a preoperative diagnosis of active infective endocarditis were selected from 34 Spanish hospitals between 2008 and 2018.

The proposed meta-model uses information that can be collected routinely, such as age, sex, kidney failure, previous cardiac surgery, chronic lung disease, pulmonary hypertension, left ventricular ejection fraction, pre-operative critical status, New York Heart Association (NYHA) functional status, presence of paravalvular complications (abscess and/or fistula), the urgency of the procedure, number of valves/prostheses treated, valve location and aetiology of infection. This information is used to estimate the probability of an infective endocarditis patient dying in the 30 days following surgery or during their admission to hospital.

Borja M. Fernández Félix, the leading author and researcher of CIBERESP, said that “the online meta-model calculator is a tool that can be used to predict the risk of postoperative mortality in patients with infective endocarditis. It has been developed and validated in-house following a robust methodology that has allowed us to aggregate the best available evidence using data from the national endocarditis registry. The results have shown that the meta-model has a better predictive performance than pre-existing models. The calculator, which is freely available online, will assist clinicians in their decision-making and provide insight into the individual risk profile of each patient”.

This study involved researchers from the Autonomous University of Madrid, the Spanish Network for Research in Infectious Diseases (REIPI), the Support Group for the Management of Infective Endocarditis in Spain (GAMES), the Cardiovascular Surgery Department of the Fundación Jiménez Díaz University Hospital and the Infectious Diseases Departments of Ramón y Cajal University Hospital and Vall d'Hebron University Hospital, in collaboration with researchers from the Epidemiology and Cardiovascular Research Unit of the Cardiology Department and the Vall d'Hebron Research Institute (VHIR) and the CIBERESP Clinical Epidemiology Programme of Ramón y Cajal, Vall d’Hebron and Cruces university hospitals.

The risk calculator can be accessed online:

https://www.evidencio.com/models/show/2498