Predictors of elevated follow-up high sensitivity troponin after an initial normal value during acute exacerbations of chronic obstructive pulmonary disease

Abstract

Abstract Funding Acknowledgements Type of funding sources: None. BACKGROUND Elevation of cardiac troponin (cTn) may be seen in acute exacerbations of chronic obstructive pulmonary disease (AECOPD) and is both diagnostic for myocardial injury and confers a poorer prognosis. Serial cTn testing after an initial normal value is often performed even though the yield is low. PURPOSE We aim to investigate factors predicting an abnormally elevated cTn after an initial normal cTn value. METHODS Retrospective data analysis of patients with AECOPD admitted between 1st January 2017 to 30th June 2017 with at least 2 serial high sensitivity cTn I (HScTnI) performed within 24 hours of presentation. Binary logistic regression was used to explore predictors of abnormally elevated (>40 ng/L) serial HScTnI after an initial normal value. RESULTS There were 240 cases with an initial normal HScTnI (mean age 74 years, 92.1% male). Amongst these, 97 cases had another follow-up HScTnI performed within 24 hours of presentation. Patients in the groups with or without follow-up HScTnI were similar in terms of demographics, COPD stage, cardiovascular risk factors, history of coronary disease and vital parameters. Significantly more patients in the group with follow-up HScTnI had ischaemic electrocardiographic changes (10.3 vs 2.1%, p = 0.006). There were 22/97 (22.7%) cases with abnormally elevated follow-up HScTnI. The mean time interval between the initial and follow-up HScTnI was not significantly different between the groups with or without elevated HScTnI (12.8 vs 10.4 hours, p = 0.858). Cases with elevated follow-up HScTnI had significantly higher heart rate (109 vs 97bpm, p = 0.007), initial HScTnI (24 vs 17ug/L, p = 0.009), C reactive protein (43.1 vs 7.6mg/L, p = 0.004) and procalcitonin (0.16 vs 0.07ug/L, p = 0.014) levels. They also had a greater proportion of chest radiographs showing consolidation (15 vs 9%, p = 0.046). On univariate analysis, heart rate (OR 1.037 [1.009 - 1.066], P = 0.009), temperature (OR 1.897 [1.119 - 3.217], p = 0.017), initial HScTnI value (OR 1.063 [1.014 - 1.115], p = 0.012) and neutrophil counts (OR 1.039 [1.000 - 1.078], p = 0.049) were significant predictors of an abnormally elevated follow-up HScTnI. Only heart rate remained as a significant predictor (Adj OR 1.050 [CI 1.015 - 1.086] p = 0.005) after multivariate analysis. CONCLUSIONS Less than a quarter of AECOPD patients with an initial HScTnI had an elevated level on serial testing. A higher heart rate is a significant predictor of a subsequent elevated HScTnI after an initial normal value and may identify patients for selective serial cTn testing. Larger studies are needed to identify predictors that can reliably risk stratify patients for serial cTn testing.