Overestimation of the pretest probability of coronary heart disease with the Diamond and Forrester algorithm in low prevalent Mediterranean area

Abstract Funding Acknowledgements Type of funding sources: None. Background Initial diagnostic approach for chest pain includes: nature of pain (non coronary, typical or atypical), sex and age. This probabilistic Bayesian method is called the Diamond and Forrester algorithm. It uses as benchmarks coronariography and autopsy, but strongly depends of ischemic heart disease (IHD) prevalence. In low prevalent countries of IHD, as Mediterranean area, the very high negative predictive value of cardiac CT could overcome Diamond and Forrester algorithm. Objective To compare pre-test probability of IHD in patients with atypical chest pain using the algorithm of Diamond and Forrester. (expected) opposite to cardiac computed tomography (observed). Methods Patients examined with cardiac CT. All patients were pretreated with metoprolol 5–20 mg. e.v., unless contraindications, to obtain sinus rhythm <60 bpm and 0.4 mg of sublingual nitro-glycerine. Prospective studies or retrospective at fixed phase at 75% of RR interval when appropriated were used for standardized reconstruction. Postprocessing accomplished with double reading. Results A subset of 175 consecutive PP submitted for atypical chest pain was compared according to its expected prevalence (Diamond and Forrester) and that observed. Mean age 66 ± 12 years, 54% male, prevalence of hypertension was 50%, hypercholesterolemia 36%, diabetes 19% and smoking 11%. CHD was demonstrated in 37% vs 58% expected from the method of Diamond and Forrester (p <0.05). Age related dichotomic analysis of the expected probability disclosed the highest difference when age <60 yo. (expected 42% vs. observed 11%; p<0.05) and lesser when ≥60 yo (expected 64% vs. observed 47%; p<0.09). Conclusion In low prevalent Mediterranean area, clinical evaluation of patients with a history of atypical chest pain using Bayesian calculation overestimates the true prevalence of coronary heart disease, especially in those less than 60 year old. Our results would remark the crescent role of cardiac CT as an efficient, gatekeeper tool for overwhelming and costly health systems.