Background Decision making in chest pain of uncertain origin is challenging.

Objectives To evaluate the predictive value of simple characteristics of pain presentation in patients coming to the emergency department with chest pain and without electrocardiogram ischaemia or raised troponin.

Methods 789 patients were studied. The following categorical pain characteristics were collected: effort related pain, pressing character, radiation, associated symptoms, and ≥2 episodes in 24 h. Additionally, a predefined semi-quantitative pain score including seven items (Geleijnse score) was completed. Risk factors and co-morbidities were also recorded. The primary and secondary endpoints were cardiac events at 30 days and at 1 year.

Results After adjusting for risk factors and co-morbidites, the pain characteristics associated with the primary and secondary endpoints were effort related pain (HR=2.1, 95% CI 1.5 to 3.0, p=0.0001; HR=1.8, 95% CI 1.3 to 2.5, p=0.0003) and ≥2 episodes in 24 h (HR=2.4, 95% CI 1.7 to 3.5, p=0.0001; HR=2.3, 95% CI 1.7 to 3.2, p=0.0001). Both variables retained their predictive value in women, diabetics and elderly (>70 years) patients. The discriminatory capacity of the predictive models including these two pain characteristics for the primary and secondary endpoints (C-statistic 0.76 and 0.76) was better than using the complex semi-quantitative pain score (C-statistic 0.69 and 0.71).

Conclusion In patients presenting to the emergency department with chest pain and without electrocardiogram ischaemia or raised troponin, effort related pain and ≥2 episodes in 24 h are the main characteristics to be considered for decision making.