ACCORD BP 2010

Methods	Multicentre, 2 x 2 factorial design, ITT strategy. Participants and investigators were not masked to blood pressure goal. All participants in the ACCORD BP trial were randomly assigned to either intensive or standard glycaemic control, and were also randomly assigned to either intensive or standard blood pressure control. The follow‐up was 4 to 8 years (mean 4.7 years)
Participants	Men and women with type 2 diabetes mellitus and a glycated haemoglobin level of 7.5% or more, aged 40 to 79 years with cardiovascular disease or 55 to 79 years with anatomical evidence of a substantial amount of atherosclerosis, albuminuria, left ventricular hypertrophy, or at least two additional risk factors for cardiovascular disease (dyslipidaemia, hypertension, smoking or obesity). Exclusion criteria included body mass index (BMI) > 45, serum creatinine (sCR) level > 1.5 mg/dL and other serious illness. Participants with SBP between 130 and 180 mmHg who were taking three or fewer antihypertensive medications and who had the equivalent of a 24‐hour protein excretion rate < 1.0 g were also eligible for the blood pressure trial. Baseline characteristics for participants included in the review (% or mean ± SD): men/women (63/37%); age (62 ± 8 years); age ≥ 75 years (7%); SBP (138 ± 16 mmHg); DBP (74 ± 11 mmHg); current smoker (13%); ethnic group: White (62%), non White (38%). Types of drugs at 1 year: thiazides (51%), ACEI/ARB (84%), calcium channel blockers (CC)B (26%), beta blocker (57%), other (28%). Previous cardiovascular condition: IHD (86%), stroke (20%). Country: USA, Canada
Interventions	Standard target: SBP < 140 mmHg Lower (intensive) target: SBP < 120 mmHg
Outcomes	The primary end point for ACCORD was the first occurrence of a major cardiovascular event, which was defined as the composite of non fatal MI, non fatal stroke, or cardiovascular death. Prespecified secondary outcomes included the combination of the primary outcome plus revascularization or hospitalization for CHF; the combination of a fatal coronary event, non fatal MI, or unstable angina; non fatal MI; fatal or non fatal stroke; non fatal stroke; death from any cause; death from cardiovascular causes; and hospitalization or death due to heart failure
Funding sources	Supported by contracts from the NHLBI. The National Institute of Diabetes and Digestive and Kidney Diseases, the National Institute on Aging, the National Eye Institute and the Centers for Disease Control and Prevention also contributed funding. General Clinical Research Centers provide support at many sites. Several companies provided study medications
Declarations of interest	Drs Bigger, Buse, Byington, Corson, Cushman, Cutler, Evans, Friedewald, Gerstein, Goff, Grimm, Ismail‐Beigi, Katz, Peterson and Probstfield declared different types of relationships with NIH institutions and pharmaceutical companies (consultancy, grants, honoraria…)
Notes	The glycaemia ACCORD trial was stopped on 6 February 2008. The blood pressures achieved at the end of the trial were: standard target: SBP 133.5 ± 0.4 mmHg; lower target: SBP 119.3 ± 0.4 mmHg. A public repository provided individual patient data from hypertensive participants with established cardiovascular disease to be used in this systematic review
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"Randomization was performed centrally on the study’s Web site with the use of permuted blocks to maintain concealment of future study‐group assignments" (pp. 1575‐85)
Allocation concealment (selection bias)	Low risk	"Randomization was performed centrally on the study’s Web site with the use of permuted blocks to maintain concealment of future study‐group assignments" (pp. 1575‐85)
Blinding of participants and personnel (performance bias) All outcomes	High risk	The study design is not compatible with blinding of participants and personnel
Blinding of outcome assessment (detection bias) All outcomes	Low risk	"ACCORD utilized a centralized adjudication process for all deaths, and hospitalizations for myocardial infarction and strokes. Upon identification of a potential outcome, clinical site staff obtained medical records or details regarding the case. Personal identifiers and information that may have alerted adjudicators to treatment assignment (e.g. A1C values) were masked by the clinical site and the medical records sent to the Coordinating Center" (pp. 1575‐85; Appendix 1)
Incomplete outcome data (attrition bias) All outcomes	Low risk	Consort diagram (section 2) (pp.1575‐85; Appendix 1)
Selective reporting (reporting bias)	Low risk	No reporting bias (protocol was checked)
Other bias	Unclear risk	Subgroup of participants with basal cardiovascular disease not predefined