TOPCAT.

Study characteristics
Methods	Study design: parallel RCT Centres: "233 sites in 6 countries (1151 participants in the United States, 326 in Canada, 167 in Brazil, 123 in Argentina, 1066 in Russia, and 612 in Georgia)" Start of enrolment: August 2006 End of enrolment: January 2012 Mean follow‐up: 3.3 years Run‐in period: no
Participants	Inclusion criteria: "≥ 50 years old, "had at least one sign and at least one symptom of heart failure on a prespecified list of clinically defined signs and symptoms, a left ventricular ejection fraction of 45% or more as measured at the local site by means of echocardiography or radionuclide ventriculography, controlled systolic blood pressure (defined as a target systolic blood pressure of < 140 mm Hg or ≤ 160 mm Hg if the patient was taking three or more medications to control blood pressure), and a serum potassium level of less than 5.0 mmol per liter. In addition, eligible patients had a history of hospitalization within the previous 12 months, with management of heart failure a major component of the care provided (not adjudicated by the clinical‐events adjudication committee), or an elevated natriuretic peptide level within 60 days before randomization (a brain natriuretic peptide [BNP] level ≥ 100 pg per milliliter or an N‐terminal pro‐BNP [NTproBNP] level ≥360 pg per milliliter)." Exclusion criteria: "severe systemic illness with a life expectancy of less than 3 years, severe renal dysfunction (an estimated glomerular filtration rate [GFR] of <30 ml per minute per 1.73 m2 of body‐surface area or a serum creatinine level that was ≥ 2.5 mg per deciliter [221 μmol per liter]), and specific coexisting conditions, medications, or acute events." Randomised (N): 3445 (1722 intervention, 1723 control) Withdrawn (N): 311 for reasons other than death (160 intervention, 151 control) Lost to follow‐up (N): 132 (67 intervention, 65 control) Analysed (N): 3445 (1722 intervention, 1723 control) Age (years, median, IQR): intervention: 68.7, 61.0 to 76.4; control: 68.7, 60.7 to 75.5 Sex (% men): intervention: 48.4; control: 48.5 Ethnicity (%): intervention: white 88.6, control: white 89.2 Systolic blood pressure (mmHg, median, IQR): intervention: 130,120‐139; control: 130, 120‐140 Heart rate (beats/min, median, IQR): intervention: 68, 62‐76; control: 68, 62‐76 BMI (median, IQR): intervention: 31, 27‐36; control: 31, 27‐36 Serum creatinine (mg/dL, median, IQR): intervention: 1.0, 0.9‐1.2; control: 1.1, 0.9‐1.2 B‐type natriuretic peptide (pg/mL): only in subgroup NT pro B‐type natriuretic peptide (pg/mL): only in subgroup LVEF (%, median, IQR): intervention: 56, 51‐61; control: 56, 51‐62 NYHA class I (%): intervention: 3.3; control: 3.1 NYHA class II (%): intervention: 63.3; control: 64.1 NYHA class III (%): intervention: 33.0; control: 32.1 NYHA class IV (%): intervention: 0.4; control: 0.5 Hypertension (%): intervention: 91; control: 92 Diabetes (%): intervention: 33; control: 32 Atrial fibrillation (%): intervention: 35; control: 35 Hospitalisation for heart failure: not reported Coronary heart disease (%): intervention: 26; control: 26 Stroke (%): intervention: 7; control: 8 Diuretic (%); intervention: 81; control: 82 Digoxin not reported Beta‐blocker (%): intervention: 78; control: 77 ACEI or ARB (%): intervention: 84; control: 84 MRA (%): 0
Interventions	Intervention: spironolactone "Study drugs were initially administered at a dose of 15 mg once daily, which was increased to a maximum of 45 mg daily during the first 4 months after randomization. Subsequent dose adjustments were made as required." Comparator: matching placebo Concomitant medication: "Study patients continued to receive other treatments for heart failure and coexisting illnesses throughout the trial."
Outcomes	Planned: From NCT record 21 April 2006: primary outcomes: cardiovascular mortality, aborted cardiac arrest, composite of hospitalisation for the management of heart failure (ie hospitalisation for non‐fatal myocardial infarction or non‐fatal stroke). Secondary outcomes: all‐cause mortality, composite of cardiovascular mortality or cardiovascular related hospitalization (i.e. hospitalization for non‐fatal myocardial infarction, non‐fatal stroke, or the management of heart failure), hospitalization for the management of heart failure incidence rate, sudden death or aborted cardiac arrest Reported: "composite of death from cardiovascular causes, aborted cardiac arrest, or hospitalization for the management of heart failure; myocardial infarction; stroke; hospitalisation from any cause; hyperkalemia (potassium level, ≥ 5.5 mmol per liter); hypokalemia (potassium level, < 3.5 mmol per liter); an elevated serum creatinine level (≥ 2 times the baseline value and above the upper limit of the normal range); serum creatinine level of 3.0 mg per deciliter (265 μmol per liter) or higher; serious adverse events"
Notes	NCT record reports on QoL but no usable data. Hamo 2015 reports baseline QoL data but not by intervention arm. Solomon 2016 reports data for four LVEF groups for HF hospitalisation, CV death, death (table 2) ‐ 40‐49%, 50‐54.99%, 55‐59.99%, 60% and over. Data for all‐cause mortality, lost to follow up, hyperkalemia differ between Pitt 2014 and NCT results. Emailed investigators to ask for end scores for QoL KCCQ, clarification on withdrawals due to adverse events and subgroup data for primary outcomes. No response.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"Eligible participants were randomly assigned to receive either spironolactone or placebo in a 1:1 ratio with the use of permuted blocks." "the randomization software will return a Treatment Allocation Code corresponding to either spironolactone or placebo"
Allocation concealment (selection bias)	Low risk	"The nurse coordinator will utilize a master list of Treatment Allocation Codes to determine which labelled study drug packet to provide to the subject."
Blinding of participants and personnel (performance bias) All outcomes	Low risk	"Subjects and treating physicians will be blinded to whether subjects are receiving spironolactone or placebo"
Blinding of outcome assessment (detection bias) All outcomes	Low risk	"Data were collected and managed electronically by the New England Research Institutes Clinical Trial Coordinating Center, which also coordinated site monitoring and analyzed the trial results (with independent verification at Brigham and Women’s Hospital)."
Incomplete outcome data (attrition bias) All outcomes	Low risk	"All randomly assigned participants were included in all analyses according to the intention‐to‐treat principle."
Selective reporting (reporting bias)	Low risk	reported as planned
Other bias	Low risk	"sponsored by National Heart, Lung and Blood Institute, National Institutes of Health"