Masahiro Ushio1, Moritoki Egi2, Junji Wakabayashi3, Taichi Nishimura3, Yuji Miyatake4, Norihiko Obata3, Satoshi Mizobuchi3. 1. Department of Anesthesia, Kakogawa West City Hospital, Kakogawa City, Hyogo, Japan. 2. Department of Anesthesiology, Kobe University Hospital, Kobe City, Hyogo, Japan. Electronic address: moriori@tg8.so-net.ne.jp. 3. Department of Anesthesiology, Kobe University Hospital, Kobe City, Hyogo, Japan. 4. Department of Surgery Related, Kobe University Graduate School of Medicine, Kobe City, Hyogo 650-0017, Japan.
Abstract
OBJECTIVE: To determine the effects of milrinone on short-term mortality in cardiac surgery patients with focus on the presence or absence of heterogeneity of the effect. DESIGN: A systematic review and meta-analysis. SETTING AND PARTICIPANTS: Five hundred thirty-seven adult cardiac surgery patients from 12 RCTs. INTERVENTIONS: Milrinone administration. MEASUREMENTS AND MAIN RESULTS: The authors conducted a systematic Medline and Pubmed search to assess the effect of milrinone on short-term mortality in adult cardiac surgery patients. Subanalysis was performed according to the timing for commencement of milrinone administration and the type of comparators. The primary outcome was any short-term mortality. Overall analysis showed no difference in mortality rates in patients who received milrinone and patients who received comparators (odds ratio = 1.25, 95% CI 0.45-3.51, p = 0.67). In subanalysis for the timing to commence milrinone administration and the type of comparators, odds ratio for mortality varied from 0.19 (placebo as control drug, start of administration after cardiopulmonary bypass) to 2.58 (levosimendan as control drug, start of administration after cardiopulmonary bypass). CONCLUSIONS: Among RCTs to assess the effect of milrinone administration in adult cardiac surgery patients, there are wide variations of the odds ratios of administration of milrinone for short-term mortality according to the comparators and the timing of administration. This fact may suggest that a simple pooling meta-analysis is not applicable for assessing the risk and benefit of milrinone administration in an adult cardiac surgery cohort. Copyright Â
OBJECTIVE: To determine the effects of milrinone on short-term mortality in cardiac surgery patients with focus on the presence or absence of heterogeneity of the effect. DESIGN: A systematic review and meta-analysis. SETTING AND PARTICIPANTS: Five hundred thirty-seven adult cardiac surgery patients from 12 RCTs. INTERVENTIONS:Milrinone administration. MEASUREMENTS AND MAIN RESULTS: The authors conducted a systematic Medline and Pubmed search to assess the effect of milrinone on short-term mortality in adult cardiac surgery patients. Subanalysis was performed according to the timing for commencement of milrinone administration and the type of comparators. The primary outcome was any short-term mortality. Overall analysis showed no difference in mortality rates in patients who received milrinone and patients who received comparators (odds ratio = 1.25, 95% CI 0.45-3.51, p = 0.67). In subanalysis for the timing to commence milrinone administration and the type of comparators, odds ratio for mortality varied from 0.19 (placebo as control drug, start of administration after cardiopulmonary bypass) to 2.58 (levosimendan as control drug, start of administration after cardiopulmonary bypass). CONCLUSIONS: Among RCTs to assess the effect of milrinone administration in adult cardiac surgery patients, there are wide variations of the odds ratios of administration of milrinone for short-term mortality according to the comparators and the timing of administration. This fact may suggest that a simple pooling meta-analysis is not applicable for assessing the risk and benefit of milrinone administration in an adult cardiac surgery cohort. Copyright Â