PARAGON‐HF.

Study characteristics
Methods	Study design: RCT Centres: 848 centres in 43 countries Start of enrolment: June 2014 End of enrolment: December 2016 Follow‐up: median 35 months Run‐in period: "Single‐blind run‐in period", Valsartan run in phase median 15 days (interquartile range, 12‐22). Sacubitril‐valsartan run in phase ‐ 19 days (interquartile range, 15 to 23 days)
Participants	Inclusion criteria: "Eligibility requirements at screening included an age of 50 years or older, signs and symptoms of heart failure, NYHA class II to IV, an ejection faction of 45% or higher within the previous 6 months, elevated level of natriuretic peptides (with different cutoffs depending on the occurrence of recent hospitalization for heart failure and presence of atrial fibrillation or flutter), evidence of structural heart disease, and diuretic therapy." Exclusion criteria: Any prior measurement of LVEF <40%, ACS, cardiac surgery, other major CV surgery within 3 months, or urgent PCI within 3 months or and elective PCI within 30 days pror to entry, Any clinical event within 6 months prior to entry that could have reduced the LVEF (MI or CABG) unless echo confirms an EF>45%, Current acute decompensated HF requiring therapy, Patients who require treatment with 2 or more of the following: an angiotensin converting enzyme inhibitor (ACEI), an angiotensin receptor blocker (ARB) or a renin inhibitor, Alternative reason for shortness of breath such as: significant pulmonary disease or severe COPD, hemoglobin (Hgb) <10 g/dl, or body mass index (BMI) > 40 kg/m2, Systolic blood pressure (SBP) ≥ 180 mmHg at entry, or SBP >150 mmHg and <180 mmHg at entry unless the patient is receiving 3 or more antihypertensive drugs, or SBP < 110 mmHg at entry. Randomised (N): 4822 (intervention: 2419; control: 2403) Withdrawn (N): for reasons other than death: 26 (intervention: 12 patients ‐ good clinical practice violation; control: 14 patients ‐ due to good clinical practice vioation) Lost to follow‐up (N): 2 Analysed (N): 4796 (intervention: 2407; control: 2389) Age (years, mean, SD): intervention: 72.7, 8.3; control: 72.8, 8.5 Sex (% men): intervention: 48.4; control: 48.2 Ethnicity (%): intervention: White 81.6, Black 2.2, Asian 12.3, Other 4.0; control: White 81.4, Black 2.1, Asian 13.0, Other 3.6 Systolic blood pressure (mmHg, mean, SD): intervention: 130, 15.6; control: 130.6, 15.3 Heart rate (beats/minute, SD): intervention: 70.6, 12.3; control: 70.3, 12.2 BMI (mean, SD): intervention: 30.2, 4.9; control: 30.3, 5.1 Serum creatinine (mg/dL, SD): intervention: 1.1, 0.3; control: 1.1, 0.3 B‐type natriuretic peptide not reported NT pro B‐type natriuretic peptide (pg/mL, median (IQR)): intervention: 904 (475 ‐ 1596); control: 915 (453 ‐ 1625) LVEF (%, mean, SD): intervention: 57.6, 7.8; control: 57.5, 8.0 NYHA class I (%): intervention: 3; control: 2 NYHA class II (%): intervention: 77.5; control: 77 NYHA class III (%): intervention: 19; control: 19.8 NYHA class IV (%): intervention: 0.3; control: 0.5 Hypertension (%): intervention: 95.7; control: 95.4 Diabetes (%): intervention: 43.5; control: 42.5 Atrial fibrillation (%): intervention: 32.2; control: 32.5 Hospitalisation for heart failure (%): intervention: 47.2; control: 49 Coronary heart disease (%): intervention: 37.4; control: 34.5 Stroke (%): intervention: 11.1; control: 10.1 Diuretic (%): intervention: 95.3; control: 95.9 Digoxin (%): not reported Beta‐blocker (%): intervention: 79.9; control: 79.5 ACEI/ARB (%): intervention: 86.2; control: 86.4 MRA (%): intervention: 24.6; control: 27.1
Interventions	Intervention: Sacubitril‐valsartan (target dose, 97 mg of sacubitril with 103 mg of valsartan twice daily) Comparator: valsartan (target dose, 160 mg twice daily) Concomitant medication: From protocol: "patients will continue to take optimal background therapy to treat co‐morbid conditions, as considered appropriate by the investigator and in accordance with standard therapy guidelines, with the exception of an ACEI or ARB as this will be replaced by study drug."
Outcomes	Planned: (from protocol): Primary objective: To compare sacubitril/valsartan to valsartan in reducing the rate of the composite endpoint of cardiovascular death and total (first and recurrent) HF hospitalizations in HF patients (NYHA functional class II to IV) with preserved EF (LVEF ≥45%). Secondary objectives: changes in the clinical summary score for HF symptoms and physical limitations, as assessed by the KCCQ), improving NYHA functional classification, delaying the time to first occurrence of a composite renal endpoint, delaying the time to all‐cause mortality Reported: "The primary outcome was a composite of total (first and recurrent) hospitalizations for heart failure and death from cardiovascular causes. Secondary outcomes were the change from baseline to 8 months in the clinical summary score on the Kansas City Cardiomyopathy Questionnaire (KCCQ)12 (scores range from 0 to 100, with higher scores indicating fewer symptoms and physical limitations); the change from baseline to 8 months in NYHA class; the first occurrence of a decline in renal function (decrease in the estimated glomerular filtration rate of ≥50%, development of end‐stage renal disease, or death due to renal failure) in a time‐to‐event analysis; and death from any cause in a time‐to‐first‐event analysis."
Notes	Funding: Novartis "The steering committee designed and oversaw the conduct of the trial and data analysis, in collaboration with the sponsor, Novartis."
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	From protocol: "all eligible patients will be randomized via Interactive Response Technology (IRT) to one of the treatment arms."
Allocation concealment (selection bias)	Low risk	From protocol: "A patient randomization list will be produced by the IRT provider using a validated system that automates the random assignedment of patients numbers to randomization numbers. These randomization numbers are linked to the different treatment arms, which in turn are linked to medication numbers"
Blinding of participants and personnel (performance bias) All outcomes	Low risk	From protocol: "patients, investigator staff, persons performaing the assessments, and data analysts will remain blind to the identity of the treatment from the time of randomization until database lock"
Blinding of outcome assessment (detection bias) All outcomes	Low risk	All the outcomes except KCCQ score and NYHA class were blindly adjudicated according to pre‐specified criteria.
Incomplete outcome data (attrition bias) All outcomes	Low risk	ITT performed on all outcomes.
Selective reporting (reporting bias)	Low risk	Primary and secondary outcomes reported as planned.
Other bias	Low risk	No other concerns