Methods

PROBE study. Presumably multi‐centre (not stated)

4418 randomised and analysed
2212 (lower), 2206 (higher)

Participants

Elderly Japanese outpatients 65‐85 years with baseline BP > 160 mmHg
Mean 73.6 years, 60% women
Baseline BP 170/90, 12% had diabetes mellitus, 55% previous BP treatment, 13.5% smoked

Exclusions: current use of efonidipine, diastolic BP > 120 mmHg, secondary hypertension, recent stroke (< 6 months prior) or signs and symptoms of stroke, a recent MI or coronary angioplasty (< 6 months previously), angina pectoris requiring hospitalisation, CHF (NYHA) ≥ class II, persistent arrhythmia such as atrial fibrillation, dissecting aneurysm of the aorta or occlusive arterial disease, hypertensive retinopathy, serum aspartate aminotransferase or serum alanine aminotransferase levels > double the respective upper limits of normal, poorly controlled diabetes mellitus (fasting blood sugar ≥ 200 mg/dL or HbA1c ≥ 8%), renal disease (serum creatinine ≥ 1.5 mg/dl), malignant disease or collagen disease. People considered "unsuitable as subjects" were also excluded

Interventions

2 years lower BP target (systolic BP < 140 mmHg) versus higher BP target (systolic BP < 160 but ≥ 140 mmHg)

Run‐in period (4 weeks in untreated participants and 2–4 weeks in treated participants) to assess baseline BP during which participants were examined on at least two occasions

Untreated participants initially received efonidipine 20‐40 mg once daily (a long‐acting dihydropyridine calcium antagonist). In participants who were already receiving antihypertensive medications, a similar dose of efonidipine was added or substituted for one of the drugs being received before study entry without a washout period. Daily dose of efonidipine could be increased to 60 mg (once or twice daily) and antihypertensive drugs other than calcium antagonists were added, if needed

Study visits: with physicians every 2 or 4 weeks. During these visits BP drugs were titrated to the allocated target BP with the goal of reaching that achieved BP within 3 months of treatment allocation.

Achieved BP at study completion 135.9/74.8 (lower) versus 145.6/78.1 (higher)

Outcomes

The primary endpoint was the combined incidence of cerebrovascular disease (cerebral haemorrhage, cerebral infarction, transient ischaemic attack, and subarachnoid haemorrhage), cardiac and vascular disease (myocardial infarction, angina pectoris requiring hospitalisation, heart failure, sudden death, dissecting aneurysms of the aorta, and occlusive arterial disease), and renal failure (acute or chronic renal failure; doubling of the serum creatinine concentration to a value of 1.5 mg/dL or higher)

Cerebrovascular disease was diagnosed based on neurological and radiological examinations. Cardiac and vascular diseases were diagnosed using radiographic, echocardiographic, and biochemical methods in addition to signs and symptoms. Sudden death, defined as death from instantaneous, unanticipated circulatory collapse within 1 h of initial symptoms, was also included in cardiac and vascular disease. Arrhythmias such as atrial fibrillation were not included in the primary endpoint, but were considered adverse events.

Secondary endpoints were "deaths from any cause, morbidity other than cardiovascular disease, changes in BP and heart rate, and any problems in regard to safety."

All outcomes were assessed at 2 years. Participants who died within 28 days after the onset of any of the primary or secondary endpoints were considered to have died from these diseases.

Notes

Dates: the registration period was from 1 April 2001‐31 December 2002. The treatment period ended on 31 December 2004. All participants followed for two years

Funding Source: sponsored by Shionogi & Co Ltd (makers of efonidipine)
Declaration of potential conflicts of interest: not reported
Other: supported by the Japan Physicians Association and the Japanese Society of Hypertension collaborators, but did not mention how these organisations are funded

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Randomly assigned the subject to either treatment group using a computer‐generated list

Allocation concealment (selection bias)

Low risk

"The investigators sent a registration form describing the clinical characteristics of eligible patients to the registration office by facsimile. Immediately after registration, the registration office randomly assigned the subject to either treatment group using a computer‐generated list and informed the investigators of the treatment assignments."

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Participants and clinicians not blinded

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

Endpoint assessment committee was blinded and reasonably objective outcomes were used

Incomplete outcome data (attrition bias)
All outcomes

Low risk

An ITT analysis was performed on all randomised participants with the exception of the 1.6% and 1.7% of participants that were lost to follow‐up. Unclear how those lost to follow‐up were handled in the analysis

Selective reporting (reporting bias)

Low risk

Reporting appears complete

Other bias

Low risk

The study was funded by the makers of efonidipine but the study question and design were not product focused

Methods

PROBE study. Multicenter

1545 lower, 1534 higher in analysis (after removing loss to follow‐up and those that withdrew)

Participants

Japanese outpatients ≥ 70 and < 85 years with isolated systolic hypertension (systolic BP > 160 mm Hg and diastolic BP < 90 mm Hg) who were either previously untreated or who could be switched from their current medications to valsartan. It is unclear whether only participants that tolerated valsartan were randomised.

Exclusions: secondary or malignant hypertension, seated systolic BP ≥ 200 mmHg or diastolic BP ≥ 90 mmHg, cerebrovascular disorder or myocardial infarction in the 6 months prior to enrolment, coronary arterioplasty 6 months prior to enrolment or coronary arteriography planned in the 6 months following enrolment, severe heart failure (≥ NYHA functional classification III), severe aortic stenosis or valvular heart disease, atrial fibrillation/flutter or serious arrhythmia, renal dysfunction with a serum creatinine level of ≥ 2 mg/dL, serious liver disease, history of hypersensitivity to valsartan, and "other patients who are judged to be inappropriate for the study by the investigator or subinvestigator".

Mean age: 76.1 years. Baseline BP 170/81. 62.4% women, 13.0% had diabetes, 19.2% smoked

Interventions

Blood pressure targets of < 140 (lower) versus 140 to ≤ 150 mmHg (higher)

Staged dose adjustments: valsartan, 40‐80 mg once daily, was the first‐step therapy for all participants. If the target BP in each group was not achieved within 1‐2 months, the dose of valsartan was increased (if < 160 mg) and/or other antihypertensive agents (except angiotensin II type 1 receptor blockers) were added.

Participants visited the clinic a minimum of once every 3 months for 2 years. 56.1% of lower participants & 57.6% of higher participants received valsartan only. 43.9% of lower participants and 42.4% of higher participants received additional BP meds (most commonly a CCB). Mean medications n = 1.6 for both groups

Achieved BP at study completion: 136.6/74.8 (lower) versus 142/76.5 (higher)

Outcomes

The primary end point of this study was a composite of cardiovascular events: sudden death, fatal or nonfatal stroke, fatal or nonfatal myocardial infarction, death because of heart failure, other cardiovascular death, unplanned hospitalisation for cardiovascular disease, and renal dysfunction (doubling of serum creatinine to a level > 2.0 mg per 100 mL or introduction of dialysis).

Secondary end points were each component of the primary end point independently, total mortality, and new onset or exacerbation of angina pectoris. Cardiovascular death, fatal or nonfatal myocardial infarction, and fatal or nonfatal stroke excluding transient ischaemic attacks were evaluated as hard end points

Notes

Dates: participants were enrolled from February 2004‐August 2005 and followed up until March 2008 (median follow‐up 3.07 years)

Funding Source: this study was funded by a grant from the Japan Cardiovascular Research Foundation and supported by the Japanese Society of Hypertension

Declarations of potential conflicts of interest: all of the study authors report receiving lecture fees from various pharmaceutical companies in Japan, including Novartis Pharma Japan (maker of valsartan ‐ the first medication introduced)

Other: unclear how the funding agencies are themselves funded

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

"Randomization of target BP levels, i.e., SBP of <140mmHg (L group) or ≥140mmHg and <150 mmHg (M group), will be performed with a minimization method based on the following assignment factors using a computer program: Sex: male or female; Age: younger than 75 years or 75 years or older; Seated SBP: less than 175 mmHg or 175 mmHg or higher; Antihypertensive therapy: not being treated or being treated; and Institution."

Allocation concealment (selection bias)

Unclear risk

“the patients were randomly assigned by the VALISH Data Center according to the following factors...”

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Participants and clinicians were not blinded to treatment group

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

“End points and adverse events were blindly evaluated according to the prospective, randomised, open‐label, blinded end point design by the endpoint committee and the safety committee, respectively.”

Incomplete outcome data (attrition bias)
All outcomes

Unclear risk

More participants (181 = 5.9%) withdrew or were lost to follow‐up than experienced the primary outcome. This included 82/1627 assigned to the lower (< 140 mmHg) BP target and 99/1633 assigned to the higher target. No sensitivity analysis was performed

Selective reporting (reporting bias)

Low risk

Reporting appears complete

Other bias

Unclear risk

How participants were selected for the per‐protocol analysis was neither described in the main publication of findings (VALISH 2010), nor pre‐defined in the preceding Rationale and Design publication which stated that the per‐protocol analysis would exclude participants "...according to judging a criteria drawn up by the Statistical Committee of this study".

Methods

PROBE study. Unclear if single site or multicenter

363 lower and 361 higher participants randomised and analysed by ITT

Participants

Chinese general practice outpatients > 70 years with either SBP ≥ 150 mm Hg and/or diastolic BP ≥ 90 mm Hg or a diagnosis of hypertension and current antihypertensive medication

Exclusions: secondary hypertension, valvular heart disease, chronic kidney dysfunction (serum creatinine ≥ 3.0 mg/dL), previous myocardial infarction or stroke in the preceding 6 months, NYHA ≥ class III CHF, echocardiography determining left ventricular ejection fraction < 40%, hepatic dysfunction, autoimmune disorders, malignant tumour, Alzheimer's disease, and "other noncardiovascular diseases potentially causing death before the end of the study".

Mean age: 76.5 years. Baseline BP 160/84. 66% men, 23% had diabetes, 25% smoked

Interventions

Lower BP target of < 140/90 versus higher BP target of < 150/90. Participants were started with single‐drug treatment of an ACE inhibitor (benzene enalapril 10 mg/d), a beta ‐blocker (bisoprolol 2.5–5 mg or metoprolol 50–100 mg/d), a CCB (amlodipine 5–10 mg/d), or a diuretic (indapamide 1.5–2.5 mg/d). Presumably initial choice of therapeutic was up to the treating physician (not stated). It is unclear whether, or how, participants already treated at baseline were switched to study medications.

To achieve the target BP, 1, 2, or 3 additional antihypertensive drugs could be added stepwise. If quadruple antihypertensive therapy (CCB + beta ‐blocker + ACE inhibitor + diuretic) failed to achieve the BP goal increasing the dose of antihypertensive drugs was recommended.

BP was measured at 4 weeks, 3 months, 6 months, and every 6 months thereafter.

Achieved BP at study completion 135.7/76.2 (lower) versus 149.7/82.1 (higher)

Outcomes

The primary outcome was the combined incidence of fatal/nonfatal stroke, acute myocardial infarction, and other cardiovascular deaths (sudden death and heart failure death).

Secondary endpoints were deaths from any causes.

Notes

Dates: not reported. Mean follow‐up 4 years

Funding Source: not reported

Declarations of Interest: not reported

Other: an abstract was published in 2011 (first author Jin, Wei 2013) with a completed analysis and no mention of Wei (the lead author of the final publication) as a co‐investigator. E‐mail queries to co‐authors confirm that first author Wei was a late addition to the project.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

“...randomly assigned to either intensive antihypertensive treatment or standard treatment by using a computer‐generated table of random numbers.”

Allocation concealment (selection bias)

Unclear risk

Method of allocation not reported

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Participants and clinicians were not blinded

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

"In order to reduce investigation bias, endpoints were evaluated by the members of the Endpoint Evaluation Committee, who were blinded to the treatment assignments and the time course of BP."

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Lower BP target: 2 discontinued treatment, 1 withdrew consent, 1 lost to follow‐up (1.1%)
Higher BP target: 7 discontinued treatment, 5 withdrew consent, 2 lost to follow‐up (3.9%)

Although an ITT analysis was stated it is unclear how missing data was handled.

Selective reporting (reporting bias)

Low risk

Reported adverse events selectively only. Total serious adverse events not provided.

Other bias

High risk

An initial analysis of this study was published in abstract form with no mention of Wei (the lead author of the final publication) as a co‐investigator. E‐mail queries to co‐authors confirm that first author Wei was a late addition to the project.