Schwartz 1992.

Study characteristics
Methods	Randomised clinical trial at 32 centres in Italy. Duration was not mentioned. However, the study was completed on December 1983
Participants	1013 participants younger than 65 years of age who had been admitted to the coronary care unit with diagnosed myocardial infarction 20 to 40 days after onset of symptoms were included in the trial. The diagnosis of acute MI was made when the following criteria were present: (1) typical electrocardiographic findings with development of pathological Q waves and/or ST‐segment evolutive changes in the anterior leads; and (2) typical changes in creatinine phosphokinase, lactic dehydrogenase, and glutamic oxaloacetic transaminase serum levels Male:female = not reported for the entire group of randomised patients, but the proportion of male:female in the high‐risk group was 132:12 Mean age = not reported for the entire group of randomised patients, but mean age in the high‐risk group was 51.4 years Exclusion criteria: age > 65 years; heart rate < 50 beats/min; clinically overt heart failure (NYHA III and IV); first‐degree heart block with PQ > 0.24 seconds or second‐ or third‐degree heart block; sick sinus syndrome; insulin‐dependent diabetes; history of asthma or chronic pulmonary obstructive disease needing treatment of IV theophylline, or systemic cortisone or beta‐2 stimulants; intermittent claudication with necrotic lesions; haemodynamically significant valvular disease; need for 'open' treatment with beta‐blockers; other non‐cardiac condition serious enough to worsen the short‐term prognosis, or to confuse endpoint evaluations, such as concurrent neoplastic disease, alcoholism, drug addiction, or psychiatric disease; inability to follow up with patients for administrative reasons; unlikely patient cooperation with regards to treatment compliance and adherence to scheduled follow‐up visits
Interventions	Study randomised patients into high‐ and low‐risk groups according to whether participants experienced at least 1 episode of ventricular fibrillation, ventricular flutter, or ventricular tachycardia. The high‐risk group was further divided into 3 groups: oxprenolol, placebo, or left cardiac sympathetic denervation (LCSD), and the low‐risk group was divided into oxprenolol or placebo. When data were reported in our review, oxprenolol and placebo groups in high‐ and low‐risk groups were pooled Experimental group: oxprenolol (160 mg 1 tablet given once daily in the morning) (n = 485) Control group: placebo (1 tablet given once in the morning) (n = 488) Co‐intervention: long‐acting nitrates and nifedipine Excluded medication: beta‐blockers, amiodarone, verapamil, prenylamine, antiplatelet drugs
Outcomes	Outcomes: cardiac death, total death, fatal and non‐fatal reinfarction Time point reported: 22 months
Notes	Email was not found Mean time from being qualified for inclusion to randomisation was 30 days MACE was calculated and was reported (total cardiac deaths + non‐fatal reinfarctions) Ciba‐Geigy was involved in the trial
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	"Patients in the low‐risk group were randomly assigned, under double‐blind conditions and in equal proportions..." "Patients who gave their informed written consent entered the study and were randomised according to which of the two risk groups they belonged. For each group, randomization was balanced within each center in preset blocks of six patients..."
Allocation concealment (selection bias)	Low risk	Treatment allocation was obtained by opening the treatment allocation envelope sent from the Study Coordination Committee
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Blindness was maintained at all stages for pharmacologically treated patients "...and for keeping the randomization code that had to be opened only if necessary; as a matter of fact, the Policy Committee never opened the randomization code throughout the study" "...the pharmacologic treatments were started immediately under double blind conditions..."
Blinding of outcome assessment (detection bias) All outcomes	Low risk	All data were processed by the Medical Department of Ciba‐Geigy Italy. Periodic random checks of the consistency of recorded and stored data were made by a statistician from the University of Milan, not otherwise involved in the study "...the End Points Evaluation Committee scrutinized the data of those patients who reached an end point and classified the events while maintaining blindness"
Incomplete outcome data (attrition bias) All outcomes	Low risk	No dropout was reported
Selective reporting (reporting bias)	Unclear risk	No protocol could be found; however mortality was reported
Other bias	Low risk	No other biases were found