Table 1 PICOTS

Population

Human adult patients aged 18 or older, diagnosed with any type of HF.

Intervention

(Model)

Multivariable models (i.e. models that contain two or more variables) for predicting any of the HF clinical outcomes listed below, or a combination of them, which considers, and possibly contains, prognostic factors, particularly biomarker concentrations, measured at baseline, on admission, or at discharge, or percentage change during hospitalization. The purpose of the model must be to yield absolute risk probabilities for individual patients. The biomarkers do not need to be part of the final model but considered as candidate predictors.

Outcomes

a) Mortality (either all-cause mortality, sudden cardiac death, or death from progressive pump failure); b) HF-related hospitalisation; c) need for cardiac transplantation; d) mechanical assist device implantation, independent of other present co-morbidities; and e) major adverse cardiovascular events (MACE) such as non-fatal stroke, non-fatal myocardial infarction, and cardiovascular death.

Any composite of these outcomes will also be considered

Timing

No constraint will be imposed on the prediction horizon as this can vary according to the outcome predicted by each particular model. For instance, mortality could be predicted at 1, 2 or 3 years whereas re-hospitalisation could be predicted at 7 days, 1 months, or 6 months.

The timing of predictor measurements could be at diagnosis of HF, discharge after a HF-related hospitalisation, or start of study recruitment.

Setting

Any setting relevant for the care of people with HF (e.g. primary care, hospital care, including emergency departments, cardiological departments, general medicine departments, intensive care units, or coronary care units).