TRADITIONAL CHINESE MEDICINE FOR HEART FAILURE ADMISSIONS IN WESTERN MEDICINE HOSPITALS IN CHINA: ANALYSIS FROM THE CHINA PEACE RETROSPECTIVE HEART FAILURE STUDY

Background: Traditional Chinese Medicine (TCM) is used in the treatment of many conditions, including heart failure (HF), though is not well characterized. Methods: We conducted a retrospective analysis of TCM use during hospitalizations for HF in China. We randomly sampled Western-medicine hospitals, stratified by economic-geographical regions, systematically sampling hospitalizations with a discharge diagnosis of HF. Patient-level data on clinical characteristics, management and in-hospital outcomes were obtained through central medical record abstraction. TCMs were categorized by the primary active ingredient. We assessed frequency of use among patients and hospitals, and the factors associated with TCM use, using hierarchical logistic regression models to account for patients' clustering within hospitals. We assessed variability in hospital-level TCM use with the median odds ratio (OR), TCM use in lieu of evidence-based therapies, and in-hospital bleeding and mortality outcomes by TCM using adjusted multivariable hierarchical models. Results: We included 10,201 patients (median age 73 years; 49% women) from 189 hospitals. In this population, 73.9% received TCM; 83.6% of whom was administered intravenously. The most commonly used TCM was Salvia miltiorrhiza (50.0%). Nearly all hospitals (99.4%) used TCM, though we observed substantial variation across hospitals (median OR=3.17; 95% CI: 2.72-3.62). Patients with coronary artery disease (OR: 1.80; 95% CI: 1.60-2.03) or stroke (OR: 1.34; 95% CI: 1.17-1.53) were more likely to receive TCM. Hospitals in Central (OR: 1.68; 95% CI: 1.05-2.69) or Western (OR: 1.72; 95% CI: 1.11-2.67) regions were more likely than hospitals in Eastern regions to use TCM. TCM was not used in lieu of evidence-based therapies. We observed a modestly significant increase in in-patient bleeding (OR: 1.35, 95% CI: 1.01-1.81) and mortality (OR: 1.36, 95% CI: 1.04-1.77) with Salvia miltiorrhiza, though not with other TCMs. Conclusion: TCM was widely used among patients admitted for HF, though the proportion varied substantially by hospital. While TCM was not used in lieu of evidence-based therapies for HF, we found a signal for harm with the most widely used TCM.