A retrospective evaluation was performed in a cohort of consecutive patients. Patients admitted to the pediatric Critical Care Unit (CCU) ofThe Hospital for Sick Children starting in January 2004 who received milrinone intravenously were eligible for inclusion. There were no exclusion criteria, which allowed patients with a variety of indications to be included in the analysis. Patients receiving milrinone were identified through the electronic patient database of the CCU.

To determine milrinone dosing practices, several charac­teristics of milrinone therapy were abstracted from the patients’ charts: loading dose, maintenance infusion rate, tapering regimen, interruptions, and overall duration of therapy. Load­ing doses given immediately postoperatively were included, to capture as many patients as possible. Patients who received milrinone before admission to the CCU were captured by means of medication lists from anesthesiology records, transfer letters, or pharmacy records. Children whose milrinone infusion was stopped and restarted were considered to have had an interruption in their therapy; the total duration of milrinone therapy was calculated as the sum of all infusions. Concomitant medications for circulatory support, as well as anti-arrhythmics, antihypertensives, potential nephrotoxins, anticoagulants, and “platelet-lowering” drugs, were also recorded. All doses were expressed on a per-kilogram basis. Death in the CCU and various patient characteristics (age, weight, sex, diagnosis, type of surgery, duration of use of aortic clamp, duration of cardiopulmonary bypass, length of stay, renal function, and other laboratory data) were recorded to capture factors potentially associated with development of low cardiac output syndrome, thrombocytopenia, or arrhythmia.

The clinical outcomes were effectiveness, assessed in terms of prevention of low cardiac output syndrome, and develop­ment of adverse events (specifically, thrombocytopenia and arrhythmia). Outcomes were defined on the basis of the literature review, as well as consensus among the clinical investigators and staff intensivists. Low cardiac output syndrome was defined as a difference in oxygen saturation between arterial and mixed venous blood (AVO2) of at least 30% or an increase in serum lactate of more than 2 mmol/L.5 Thrombocytopenia was defined as a platelet count of less than 50 x 109/L. The occurrence and type of arrhythmia were determined by an electrophysiologist, who reviewed the chart notes and electrocardiograms. Renal dysfunction was defined using the age-specific criteria of the pediatric logistic organ dysfunction score. buy antibiotics canada

A sample size of 200 patients was chosen to support multiple regression with up to 10 potentially predictive variables and to provide reasonably precise estimates of the rates of low cardiac output syndrome and potentially adverse events.