Shah 2013.

Study characteristics
Methods	Study design: randomised controlled trial, parallel arm Setting: Ahmedabad, India (type of hospital not stated) Study start/end dates or study duration: 2009 to 2013 (4 years) Funding sources: not stated Declarations of interest: not stated
Participants	Sample size: 50 Patient age: 2 to 12 years (levosimendan group 4.1 ± 1.3 years, milrinone group 4.0 ± 1.5 years) Patient sex: 16 male and 9 female in the levosimendan group, 17 male and 8 female in the milrinone group Diagnosis/severity: biventricular and univentricular physiology with the following diagnoses: septum primum atrial septal defect, ventricular septal defect, tetrology of fallot, atrial septal defect with pulmonic stenosis, ventricular septal defect with pulmonic stenosis, unobstructed total anomalous pulmonary venous connection and partially anomalous pulmonary venous connection, atrio‐ventricular septal defect, anomalous left coronary artery from pulmonary artery, type I and II Ebstein’s anomaly, double‐chambered right ventricle, Glenn’s surgery Inclusion criteria: patients posted for on‐pump paediatric cardiac surgery, in which probable maximum aortic cross clamp time 60 to 90 minutes
Interventions	Intervention 1: levosimendan (n = 25): intravenous bolus infusion of levosimendan 12 μg/kg over 10 minutes after cross clamp removal followed by 0.1 μg/kg/min Intervention 2: milrinone (n = 25): intravenous bolus infusion of milrinone 50 μg/kg over 10 minutes followed by 0.5 μg/kg/min Concomitant interventions: Both groups received dopamine infusion of 5 to 8 μg/kg/min after cross clamp removal. "Both study drugs along with dopamine were stopped average 24‐48 hours after shifting to SICU from operation theatre." Follow‐up period: "After the surgery, patients were monitored for the above said parameters in the post operative period at interval of 0,6,12,18 and 24 hours after shifting to SICU."
Outcomes	Primary outcome(s) Heart rate (ECG) Mean arterial pressure Pulmonary artery pressure Secondary outcome(s) Stroke volume, cardiac output, systemic vascular resistance, pulmonary vascular resistance, pulmonary capillary wedge pressure (estimated by transesophageal echocardiography) Respiratory rate and arterial partial pressure of oxygen, urine output, laboratory investigations such as serum creatinine level, platelet counts, troponin I level (0, 12, 24 hours) and serum lactate level (0, 24 hours), fluid requirements, adverse events After the surgery, patients were monitored for the above parameters in the postoperative period at intervals of 0, 6, 12, 18, and 24 hours after shifting to surgical ICU. Mortality: duration of follow‐up unclear Length of intensive care stay Duration of mechanical ventilation Number of patients requiring ECMO Number of patients experiencing adverse effects: arrhythmia, hypotension, hypokalaemia, thrombocytopaenia, anaemia, elevated serum levels of liver enzymes (aminotransferases), need for renal replacement therapy
Notes	For the outcome hypotension, there are two sets of results (table 2, without a definition and table 3, defined as the requirement for norepinephrine). We chose the data set with definition of this outcome. Intention‐to‐treat analysis: not stated Trialists contacted: no Unpublished information received: no
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	No random sequence generation described
Allocation concealment (selection bias)	High risk	No allocation concealment described, only "patients were randomly allocated into two groups."
Blinding of participants and personnel (performance bias) All outcomes	High risk	No blinding reported (other than the label "double blind study", even though one of the study drugs was levosimendan, which is a yellow solution and could therefore be distinguished from other infusions by the patient and bedside personnel.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	No blinding of outcome assessors described
Incomplete outcome data (attrition bias) All outcomes	Low risk	No dropouts described, results based on complete study population
Selective reporting (reporting bias)	Low risk	No bias suspected
Other bias	Unclear risk	Funding and possible conflicts of interest not reported