In-hospital Outcomes of Infective Endocarditis from 1978 to 2015: Analysis Through Machine-Learning Techniques

Abstract

Background: Early identification of patients with infective endocarditis (IE) at higher risk for in-hospital mortality is essential to guide management and improve prognosis.

Methods: A retrospective analysis was conducted of a cohort of patients followed up from 1978 to 2015, classified according to the modified Duke criteria. Clinical parameters, echocardiographic data, and blood cultures were assessed. Techniques of machine learning, such as the classification tree, were used to explain the association between clinical characteristics and in-hospital mortality. Additionally, the log-linear model and graphical random forests (GRaFo) representation were used to assess the degree of dependence among in-hospital outcomes of IE.

Results: This study analyzed 653 patients: 449 (69.0%) with definite IE; 204 (31.0%) with possible IE; mean age, 41.3 ± 19.2 years; 420 (64%) men. Mode of IE acquisition: community-acquired (67.6%), nosocomial (17.0%), undetermined (15.4%). Complications occurred in 547 patients (83.7%), the most frequent being heart failure (47.0%), neurologic complications (30.7%), and dialysis-dependent renal failure (6.5%). In-hospital mortality was 36.0%. The classification tree analysis identified subgroups with higher in-hospital mortality: patients with community-acquired IE and peripheral stigmata on admission; and patients with nosocomial IE. The log-linear model showed that surgical treatment was related to higher in-hospital mortality in patients with neurologic complications.

Conclusions: The use of a machine-learning model allowed identification of subgroups of patients at higher risk for in-hospital mortality. Peripheral stigmata, nosocomial IE, absence of vegetation, and surgery in the presence of neurologic complications are predictors of fatal outcomes in machine learning-based analysis.

Figure 1

Classification tree. Interaction between variables: fever, age, peripheral stigmata, vegetation on the first transthoracic echocardiography (TTE), mode of acquisition, Duke criteria. Hospital discharge = 0; death = 1 (terminal nodes). The variables were grouped in the classification tree, allowing the identification of subgroups of patients with similar characteristics, for whom different mortality rates were found: Node 15 = fever + community-acquired infective endocarditis (IE) + peripheral stigmata (n = 165), 36.4% mortality; Node 14 = fever + community-acquired IE + definite IE diagnosis + age > 53 years (n = 62), 43.5% mortality; Node 12 = fever + community-acquired IE + definitive IE diagnosis + age < 53 years + absence of vegetation on the first TTE (n = 36), 38.9% mortality; Node 6 = fever + nosocomial IE (n = 68), 51.5% mortality.

Figure 2

Graphical log-linear model and graphical random forests (GRaFo) representation. (A) 1978-2015, (B) 1978-1999, and (C) 2000-2015. The weights on the edges correspond to Cramer’s V coefficients (V). Outcome = death. This representation assesses the degree of dependence between the variables (complications and death). In the analysis of (A) the whole cohort, for the 2-by-2 interaction (blue lines), heart failure (HF) was associated with dialysis-dependent renal failure (DRF) (V = 0.11) and with surgical treatment (ST) (V = 0.30). Neurologic complications (NC), HF, and DRF were associated with mortality (V = 0.21, 0.20, 0.15, respectively). For the 3-by-3 interaction (red lines) of mortality, ST, and NC, a greater dependence between mortality and NC (V = 0.15) was observed. The dependence relationships found in (A) and (B) are the same. In the 2000-2015 group (C), there is a dependence relationship between surgical treatment and mortality, conditioned to the presence of HF and of the other variables.