RESUMO
BACKGROUND: Accurate identification of patients at risk of major adverse cardiac events (MACE) places a substantial burden on emergency physicians (EPs). Bayesian nomogram for risk stratification in low- to intermediate-risk cardiovascular patients has not been investigated previously. OBJECTIVE: The objective of this study was to develop a comparative diagnostic model using Bayesian statistics for exercise treadmill test (ETT) and stress echocardiogram (ECHO) to calculate post-test diagnostic risk of MACE using HEART (history, electrocardiogram, age, risk factors, and troponin) risk score as predictor of pretest probability. METHODS: Stratification was made by applying HEART scores for the prediction of MACE. Likelihood ratios (LR) were calculated using pooled sensitivity and specificity of ETT and ECHO from the American College of Cardiology Foundation/American Heart Association systematic review. Post-test probabilities were obtained after inserting HEART score and LR into Bayesian nomogram. Analysis of variance was used to assess statistical association. RESULTS: Positive LR (LR+) for ETT was 4.56 and negative LR (LR-) was 0.27; for ECHO, LR+ 5.65 and LR- 0.15. Bayesian statistical modeling post-test probabilities for LR+ and low HEART risk yielded a post-test probability for ETT of 7.75% and 9.09% for ECHO; intermediate risk gave 47.62% and 52.63%, respectively. For LR-, low HEART risk post-test probability for ETT was 0.46% and for ECHO 0.26%; intermediate risk probabilities were 4.48% and 2.49%, respectively. LR- was statistically significant in ruling out MACE with ECHO (p < 0.001), but no significant differences were seen for LR+ (p = 0.64). CONCLUSIONS: This Bayesian analysis demonstrated slight superiority of stress ECHO over ETT in low- and intermediate-risk patients in ruling out MACE.