Risk calculator of multimorbid risk of rehospitalisation and death from heart failure – including the contribution of the gut microbiome

Background The elucidation of the contributory role of multimorbidity to heart failure (HF) including the gut-heart axis has added a new dimension to our understanding of HF pathophysiology which is not reflected in currently available risk scores. The present investigation aimed to develop and validate a novel risk score model of multimorbidity for HF risk stratification. Methods A risk model was developed based on the contribution of markers associated with HF multimorbidities on outcomes of mortality and/or rehospitalization due to HF (death/HF) at one year. Two independent HF cohorts were combined and randomly split 70:30 using a split-sample validation approach for training and validation cohorts which were not significantly different for investigated variables. Backward logistic regression was used to develop the risk model with a further scoring system to create a simple risk calculator. Results A final 11-variable risk model (age, previous HF hospitalization, NYHA group III/IV, NT-proBNP, diastolic blood pressure, loop diuretic use, beta-blocker non-use, creatinine, COPD, diabetes, and combined gut metabolites) showed a diagnostic performance of 0.71 in the training cohort (C-statistic validation cohort, 0.70, p<0.001). A risk score/calculator was further developed based on this model with categorization into three (low-, mid- and high-) and two (low- and high-) risk groups, with both approaches demonstrating increased incidence of death/HF in patients at the highest risk (p<0.001). Conclusion A novel risk model and score were derived which showed the contribution of comorbidities including the added value of the gut-heart axis on risk stratification of HF patients on rehospitalization and death.