Retrospective assesment of prognosis scoring systems which used to evaluate patients who admitted emergency medicine service with heart failure

The study is designed to investigate efficiency and clinical applicability of clinical scoring systems which used to determine the prognosis of patients who admitted emergency medicine service with heart failure. 737 heart failure patients were included to study. When 4 clinical scoring algorithms (Acute Heart Failure Index, ADHERE Tree, EFFECT, BWH) which used in our study analysed; in hospital morbidity rates of low risk groups were 13,2%, 13,3%, 13,6%, 16,6^% and 30 day mortality rates of low risk groups were 8,5%, 10,2%, 6,4%, 11,3% respectively. When very low risk + low risk groups and medium risk + high risk + very high risk groups compared, it was found that medium risk + high risk + very high risk groups were associated with 4,360 times increased risk (OR: 4,360 95% CI) of 30 day mortality. Also it was found that the area under ROC curve were larger or 30 day mortality in EFFECT clinical scoring algorithm group (0,706). It was found that, the most applicable clinical scoring algorithm was EFFECT. In our study it was found that; clinical scoring algorithms which analyse initial vital signs and laboratory results are applicable for detecting low risk groups. Clinical scoring systems could be used to detect low risk groups for discharge or short follow-up decisions; depending on more enhanced algorithms. For standardizion and reliability of algorithms,there is need for bigger and multicentered studies.