Effect of baseline characteristics on mortality in the SURVIVE trial on the effect of levosimendan vs dobutamine in acute heart failure: Sub-analysis of the Finnish patients. (15th July 2016)

• Record Type: Journal Article
• Authors: Kivikko, Matti; Pollesello, Piero; Tarvasmäki, Tuukka; Sarapohja, Toni; Nieminen, Markku S.; Harjola, Veli-Pekka
• Is Part Of: International journal of cardiology. Volume 215(2016)
• Page Start: 26
• Page End: 31
• Publication Date: 2016-07-15
• Subjects: Acute heart failure -- Clinical trial subanalysis -- Mortality -- Levosimendan -- Dobutamine -- Inotropes; Cardiology -- Periodicals; Electronic journals; 616.12
• DOI: 10.1016/j.ijcard.2016.04.064
• Languages: English
• ISSNs: 0167-5273

Abstract

Abstract: Background: In the SURVIVE trial, including 1327 acute heart failure patients, no statistically significant difference between levosimendan and dobutamine in the 180-day all-cause mortality was seen. Country-specific differences in outcome were, however, present. In the Finnish sub-population in fact, mortality was significantly lower in levosimendan treated patients. We aim to understand the reasons for this disparity. Methods: The risk factors for all-cause mortality were identified in the whole study population using multivariate Cox proportional hazards regression analysis. Those factors were evaluated in the 95 patients of the Finnish sub-population. Results: The treatment by country interaction for mortality in Finland vs. other countries was significant, p = 0.029. Levosimendan treated patients had a lower 180-day mortality compared to dobutamine treated (17% vs. 40%, p = 0.023) in the Finnish sub-population. Baseline variables predicting survival in the whole SURVIVE trial population included age, systolic blood pressure, heart rate, myocardial infarction during admission, levels of NT-pro-BNP, glucose, creatinine, and alanine transferase, use of ACE inhibitors and β-blockers, oliguria, time from hospital admission to randomization, history of cardiac arrest, and left ventricular ejection fraction. Finnish patients were more frequently treated with β-blockers (88% vs. 52%, p < 0.0001), their study treatment was started earlier (mean ± SD 41 ± 40 h vs. 81 ± 154; p < 0.0001), and they had more often acute myocardial infarction at admission (39% vs. 16%, p < 0.0001). Conclusion: The lower mortality in the Finnish patients treated with levosimendan was associated with higher use of β-blockers, higher frequency of myocardial infarction at admission, and shorter delay between randomization and start of treatment.