PT - JOURNAL ARTICLE AU - Chen, Yang AU - Sikkel, Markus AU - Shah, Jaymin TI - 20 Factors Which Impact on Mortality and Readmission in Patients with Heart Failure: Real World Longitudinal Data AID - 10.1136/heartjnl-2016-309890.20 DP - 2016 Jun 01 TA - Heart PG - A12--A13 VI - 102 IP - Suppl 6 4099 - http://heart.bmj.com/content/102/Suppl_6/A12.3.short 4100 - http://heart.bmj.com/content/102/Suppl_6/A12.3.full SO - Heart2016 Jun 01; 102 AB - Background Heart failure remains challenging to manage and treat globally. Much of the data on its prognosis stems from clinical trials, where cohorts are often younger and less sick than the groups of patients usually treated in hospitals and communities. Large-scale studies examining real world cohorts and factors that affect their outcomes are lacking.Design A retrospective cohort study of 3626 patients admitted with a diagnosis of heart failure over a 9-year period. We investigated the effects of baseline characteristics, co-morbidities and echocardiographic findings on in-hospital and overall mortality along with hospital readmission. Mean ages were 74.8 and 78.7 years for males and females respectively. The average follow-up time was 3 years for mortality and 2.43 years for readmission.Results Increasing age is associated with higher mortality rate both in-hospital OR 1.04 [1.03–1.05 95% CI] and throughout the follow-up period (overall mortality) HR 1.03 [1.02–1.03 95% CI]. Being female had a protective effect for overall mortality HR 0.82 [0.74–0.91 95% CI]. (Table 1). Ethnicity had a mixed effect. For in-hospital mortality, there was an adverse association for Mixed Asian OR 1.5 [1.00–2.27 95% CI] and Chinese backgrounds OR 1.53 [1.11–2.13 95% CI]. For readmission, there was significant variation amongst different ethnic groups – Black patients were at highest risk HR 1.25 [1.02–1.54 95% CI] in contrast to Chinese patients who had the lowest risk HR 0.68 [0.53–0.86 95% CI]. (Table 2).View this table:Abstract 20 Table 1 Univariable and ultivariable hazard ratios for follow-up mortalityView this table:Abstract 20 Table 2 Univariable and multivariable hazard ratios for follow-up readmissionSubgroup analysis of echocardiography data demonstrated that compared to patients with severe systolic dysfunction (LVEF <35%), patients at all other stages of systolic dysfunction, including Heart Failure with Preserved Ejection Fraction (LVEF>60%), did not have any difference in mortality. This remained true for both in-hospital (0.86 OR [0.24–3.05 95% CI] and overall mortality 1.19 HR [0.77–1.83 95% CI].Conclusion In a real world heart failure population, we have demonstrated novel associations between ethnicity and disease trajectory and confirmed that LVEF is a poor prognostic marker. Future work is planned to assess why ethnic groups may have different outcomes, including examination of patient understanding and engagement with healthcare.