Background: Treatment after acute coronary syndrome (ACS) is influenced by the GRACE score which predicts the risk of death at 6 months. We investigated how well the GRACE score performed in COPD patients and how it was used to guide treatment compared to non-COPD patients.

Methods: Patients with an admission for ACS between 2003-2013 were identified in the UK MINAP database. COPD patients had a record of obstructive airway disease, smoking history and were aged >35. GRACE scores were calculated using a previously validated method1 and converted to predicted risk of death. Discrimination (c-statistic) and calibration (Hosmer-Lemeshow Chi2) measures were calculated for COPD and non-COPD patients. We then used logistic regression adjusted for GRACE score to compare management for COPD and non-COPD patients with the same GRACE score.

Results: 483798 patients with ACS were included. 58940 (12.2%) had COPD. Discrimination and calibration of GRACE scores were poorer for COPD patients (Table 1). Compared to non-COPD patients with the same GRACE score, COPD patients were under-treated and under-investigated (Table 2).