Blood pressure lowering efficacy of dual alpha and beta blockers for primary hypertension
Abstract
Background
Drugs with combined alpha and beta blocking activity are commonly prescribed to treat hypertension. However, the blood pressure (BP) lowering efficacy of this class of beta blockers has not been systematically reviewed and quantified.
Objectives
To quantify the dose‐related effects of various types of dual alpha and beta adrenergic receptor blockers (dual receptor blockers) on systolic and diastolic blood pressure versus placebo in patients with primary hypertension.
Search methods
We searched the Cochrane Hypertension Group Specialised Register, the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, EMBASE and ClinicalTrials.gov for randomized controlled trials up to October 2014. The WHO International Clinical Trials Registry Platform (ICTRP) is searched for inclusion in the Group's Specialised Register.
Selection criteria
Randomized double blind placebo controlled parallel or cross‐over trials. Studies contained a beta blocker monotherapy arm with a fixed dose. Patients enrolled in the studies had primary hypertension at baseline. Duration of the studies was from three to 12 weeks. Drugs in this class of beta blockers are carvedilol, dilevalol and labetalol.
Data collection and analysis
Two review authors (GW and AL) confirmed the inclusion of studies and extracted the data independently. RevMan 5.3 was used to synthesize data.
Main results
We included eight studies examining the blood pressure lowering efficacy of carvedilol and labetalol in 1493 hypertensive patients. Five of the included studies were parallel design; three were cross‐over design. The two largest included studies were unpublished carvedilol studies. The estimates of BP lowering effect (systolic BP/diastolic BP millimeters of mercury; SPB/DBP mm Hg) were ‐4 mm Hg (95% confidence intervals (CI) ‐6 to ‐2)/‐3 mm Hg (95% CI ‐4 to ‐2) for carvedilol (>1000 subjects) and ‐10 mm Hg (95% CI ‐14 to ‐7)/‐7 mm Hg (95% CI ‐9 to ‐5) for labetalol (110 subjects). The effect of labetalol is likely to be exaggerated due to high risk of bias. Carvedilol, within the recommended dose range, did not show a significant dose response effect for SBP or DBP. Carvedilol had little or no effect on pulse pressure (‐1 mm Hg) and did not change BP variability. Overall, once and twice the starting dose of carvedilol and labetalol lowered BP by ‐6 mm Hg (95% CI ‐7 to ‐4) /‐4 mm Hg (95% CI ‐4 to ‐3) (low quality evidence) and lowered heart rate by five beats per minute (95% CI ‐6 to ‐4) (low quality evidence). Five studies (N = 1412) reported withdrawal due to adverse effects; the risk ratio was 0.88 (95% CI 0.54 to 1.42) (moderate quality evidence).
Authors' conclusions
This review provides low quality evidence that in patients with mild to moderate hypertension, dual receptor blockers lowered trough BP by an average of ‐6/‐4 mm Hg and reduced heart rate by five beats per minute. Due to the larger sample size from the two unpublished studies, carvedilol provided a better estimate of BP lowering effect than labetalol. The BP lowering estimate from combining carvedilol once and twice the starting doses is ‐4/‐3 mm Hg. Doses higher than the recommended starting dose did not provide additional BP reduction. Higher doses of dual receptor blockers caused more bradycardia than lower doses. Based on indirect comparison with other classes of drugs, the blood pressure lowering effect of dual alpha‐ and beta‐receptor blockers is less than non‐selective, beta1 selective and partial agonist beta blockers, as well as thiazides and drugs inhibiting the renin angiotensin system. Dual blockers also had little or no effect on reducing pulse pressure, which is similar to the other beta‐blocker classes, but less than the average reduction of pulse pressure seen with thiazides and drugs inhibiting the renin angiotensin system. Patients taking dual receptor blockers were not more likely to withdraw from the study compared to patients taking placebo.
PICOs
Plain language summary
Alpha and beta dual receptor blockers for treatment of high blood pressure
Background
Alpha and beta dual receptor blockers are a subclass of beta blockers which are commonly used to treat high blood pressure (BP). Drugs in this class include carvedilol (Coreg), labetalol (Trandate) and dilevalol (Unicard). We searched for and found all the relevant studies to examine how well this class of drugs lowered blood pressure.
Study characteristics
We found eight clinical studies in October 2014, that examined the blood pressure lowering effect of carvedilol and labetalol in 1493 participants with high blood pressure. These people were randomly assigned to receive either a fixed dose of dual receptor blockers or a placebo for 3 to 12 weeks.
Key results
On average, dual receptor blockers lowered systolic BP by six points, diastolic BP by four points and heart rate by five beats per minute in patients with mild to moderate high blood pressure. There were more data on the effects of carvedilol. On average, carvedilol lowered systolic BP by four points and diastolic BP by three points. Higher doses of dual receptor blockers caused more slowing of heart rate but not more lowering of BP. The BP lowering effect of dual receptor blockers was less than other classes of BP lowering drugs. Patients taking dual receptor blockers were not more likely to withdraw from the study compared to patients taking placebo.
Quality of the evidence
The quality of the evidence was judged to be low due to various types of bias that could exaggerate the effect. A low quality of evidence means future research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Authors' conclusions
Summary of findings
| Dual receptor blockers compared with placebo for primary hypertension | |||
| Patient or population: Adults with primary hypertension Intervention: Alpha and beta dual receptor blockers Comparison: Placebo | |||
| Outcomes | Mean estimates of combining once and twice starting dose (95% CI) | No of Participants | Quality of the evidence |
| Systolic blood pressure | ‐5.59 (95% CI ‐7.47 to ‐3.70)1,2,3 | 1007 (8) | Low4,5 |
| Diastolic blood pressure | ‐3.88 (95% CI ‐4.95 to ‐2.82)1,2,3 | 1007 (8) | Low4,5 |
| Heart rate | ‐4.62 (95% CI ‐5.71 to ‐3.54)1,2 | 977 (7) | Low4,5 |
| Pulse pressure | ‐1.89 (95% CI ‐3.58 to ‐0.20)1,2 | 1007 (8) | Very low4,5,6 |
| WDAE | 0.88 (95% CI 0.54 to 1.42) | 1412 (5) | Moderate7 |
| SBP: systolic blood pressure; DBP: diastolic blood pressure; WDAE: withdrawal due to adverse effects; 95%CI: 95% confident interval | |||
| GRADE Working Group grades of evidence | |||
| Footnotes | |||
|
| |||
Background
This review is part of a series of four Cochrane reviews, which also includes reviews studying the blood pressure lowering efficacy of non‐selective beta blockers (Wong 2014a), beta‐1 selective beta blockers (Wong 2008) and partial agonist beta blockers (Wong 2014b).
Description of the condition
Elevated blood pressure is a prevalent condition that is associated with an increased risk of adverse cardiovascular events including stroke, myocardial infarction, congestive heart failure and renal failure. Antihypertensive drug treatment has been shown to reduce the incidence of these adverse events in patients with moderate to severe elevations of blood pressure (Musini 2009; Wright 2009). There are a number of classes of antihypertensive drugs used to treat elevated blood pressure. Beta adrenergic receptor blockers (beta blockers) are one class of these drugs.
Description of the intervention
Beta blockers were originally marketed and used to treat angina. During their use in patients with angina, it was discovered that they also lowered blood pressure. Since then, they have received clinical attention because of their proven effectiveness for certain arrhythmias and to prevent recurrence in patients who have had a myocardial infarction.
Six systematic reviews are relevant to this proposed review. Wright 2000 assessed the mortality and morbidity associated with different types of beta blockers. The authors found that patients treated with non‐selective beta blockers post‐myocardial infarction had statistically significant reduction in total mortality compared to placebo, whereas those treated with beta1‐selective beta blockers or partial agonist beta blockers did not. A recent review assessed the effects of beta adrenergic blocking agents on morbidity and mortality in adults with hypertension (Wisonge 2007). This review concluded that beta blockers are not the best class of drugs to use as first‐line antihypertensive therapy. However, it is possible that this only relates to beta1‐selective beta blockers, as atenolol, a beta1‐selective beta blocker, was the beta blocker used in 75% of the studies.
Wright 2009 examined the mortality and morbidity outcomes of different classes of first‐line antihypertensive drugs in people with hypertension. This review found that beta blockers significantly reduced stroke and cardiovascular events but not all cause mortality and congestive heart disease (CHD). The clinical outcome results of beta blockers were inferior to low‐dose thiazides, ACE Inhibitors and calcium channel blockers.
Three systematic reviews have assessed the effects of beta blockers on blood pressure. A Cochrane review of beta blockers for hypertension during pregnancy showed that oral beta blockers decreased the incidence of severe hypertension and the need for additional antihypertensive therapy (Magee 2003). A systematic review of the dose‐response blood pressure lowering effect of beta blockers and other antihypertensive drugs was limited by the fact that it did not differentiate between the different classes of beta blockers (Law 2005). Finally, a Cochrane review of the blood pressure lowering efficacy of beta blockers as a second‐line treatment did not have enough studies to be able to differentiate between the different classes of beta blockers (Chen 2010a).
How the intervention might work
Beta adrenergic receptors are present in many body systems including the heart, blood vessels, kidneys and nervous system. At the present time, the mechanism whereby beta blockers lower blood pressure in humans is not known. Many hypothetical mechanisms have been proposed, which include: decreasing cardiac output, reducing renin production or modulating the sympathetic nervous system. It is likely that a combination of these mechanisms causes the blood pressure lowering effect.
There are two major types of adrenergic receptors: alpha and beta. Alpha adrenergic receptors are divided into two subtypes, alpha1 and alpha2. Alpha1 receptors participate in the regulation of cardiac contractility, heart rate, hepatic glucose release and vasoconstriction of the aorta, coronary artery and large resistance blood vessels. Alpha2 receptors are the dominant receptors that participate in vasoconstriction of pro‐capillary arterioles and central nervous system (CNS) negative feedback to inhibit the production of norepinephrine. Activation of alpha adrenergic receptors through reflex sympathetic activation would cause vasoconstriction and therefore raise blood pressure. This effect could counter the blood pressure lowering effect of beta adrenergic antagonism. Beta receptors have three subtypes. Beta1 and beta2 are the two subtypes that participate in cardiovascular regulation. The beta1 receptor is the dominant beta receptor that mediates the catecholamine effect in the heart. Activation of beta1 receptors increases heart rate and therefore cardiac output. Beta2 receptors primarily regulate peripheral beta‐adrenergic responses such as vasodilation, bronchodilation and the release of renin and lipolysis in various tissues (Goodman 2011).
Dual alpha and beta‐blocking drugs were designed to competitively inhibit both alpha and beta receptors. Currently, there are no known selective antagonists for beta3 receptors. Dual alpha and beta blockers have both alpha and beta‐blocking properties but no partial agonist activity. These drugs may be non‐selective or beta1 selective on beta receptors as well as selectively blocking alpha1 or alpha2 receptors. Carvedilol (Coreg), dilevalol and labetalol (Trandate) are the drugs that fit into this pharmacological class. Both carvedilol and labetalol are non‐selective beta blockers and have higher affinity to alpha1 receptors than alpha2 receptors (Goodman 2011).
Why it is important to do this review
Since it is possible that beta blockers with different mechanisms of action have different effects on the reduction of morbidity and mortality, it is crucial to determine whether they have different abilities to lower blood pressure. No published review has quantified the blood pressure lowering effect of beta blockers according to their subclass of beta‐blocking activity. This review could provide important information towards understanding the mechanism by which beta blockers lower blood pressure.
Furthermore, since physicians use blood pressure lowering drugs on a daily basis to manage patients with high blood pressure, it is important to know accurately the magnitude of blood pressure lowering effect of individual beta blockers and separate subclasses.
The information found in this review would be useful for clinicians, scientists designing future drug trials and authors of other systematic reviews.
Objectives
Primary objective
To quantify the dose‐related effects of various doses and types of dual alpha and beta adrenergic receptor blockers on systolic and diastolic blood pressure versus placebo in patients with primary hypertension.
Secondary objectives
-
To determine the effects of dual alpha and beta adrenergic receptor blockers on variability of blood pressure.
-
To determine the effects of dual alpha and beta adrenergic receptor blockers on pulse pressure.
-
To quantify the dose‐related effects of dual alpha and beta adrenergic receptor blockers on heart rate.
-
To quantify the effects of dual alpha and beta adrenergic receptor blockers in different doses on withdrawals due to adverse effects.
Methods
Criteria for considering studies for this review
Types of studies
Study design met the following criteria:
-
placebo‐controlled;
-
random allocation to beta adrenergic receptor blocker group and placebo group;
-
parallel or cross‐over design;
-
double blinded;
-
duration of follow‐up at least three weeks;
-
blood pressure measurements at baseline (following washout) and at one or more time points between 3 and 12 weeks after starting treatment.
Types of participants
Participants had a baseline blood pressure of at least 140 mm Hg systolic, a diastolic blood pressure of at least 90 mm Hg, or both, measured in a standard way. Patients did not have creatinine levels greater than 1.5 times the normal level. Participants were not restricted by age, gender, baseline risk or other co‐morbid conditions.
Types of interventions
Monotherapy with any dual alpha and beta adrenergic receptor blocker, including carvedilol, dilevalol or labetalol.
Studies in which titration to a higher dose was based on blood pressure response were not eligible.
Types of outcome measures
Primary outcomes
Change in trough (13 to 26 hours after the dose), peak (1 to 12 hours after the dose) or both systolic and diastolic blood pressure compared to placebo. If blood pressure measurements were available at more than one time within the acceptable window, the weighted means of blood pressures taken in the 3 to 12 week range were used.
Secondary outcomes
-
Change in standard deviation compared to placebo.
-
Change in pulse pressure compared to placebo.
-
Change in heart rate compared to placebo.
-
Number of patients who withdraw due to adverse effects compared to placebo.
Search methods for identification of studies
We searched the Database of Abstracts of Reviews of Effects (DARE; October 2014) for related reviews. We used previously published meta‐analyses on dose‐response of beta adrenergic receptor blockers to help identify references to studies.
We searched the Cochrane Hypertension Group Specialised Register (October 2014), the Cochrane Central Register of Controlled Trials (CENTRAL; 2014, Issue 9), MEDLINE (1946 to October 2014), EMBASE (1974 to October 2014) and ClinicalTrials.gov (all years to October 2014) for randomized controlled trials. We applied no language restrictions. The WHO International Clinical Trials Registry Platform (ICTRP) is searched for inclusion in the Group's Specialised Register.
A modified, expanded version of the standard search strategy of the Hypertension Group with additional terms related to beta adrenergic receptor blockers in general and all the specific drugs listed above was used to identify the relevant articles. The MEDLINE strategy (Appendix 1) was translated into CENTRAL (Appendix 2), EMBASE (Appendix 3), ClinicalTrials.gov (Appendix 4), and the Hypertension Group Specialised Register (Appendix 5).
We searched all the databases to identify potentially relevant citations. The initial screening of these abstracts excluded articles whose titles and abstracts indicated they were clearly irrelevant. The full text of remaining articles were retrieved (and translated into English where required). We searched the bibliographies of pertinent articles, reviews and texts for additional citations.
Data collection and analysis
Selection of studies
We imported the references and abstracts identified by our search into Reference Manager 11 software. We selected studies based on the criteria listed above. Two independent review authors assessed the eligibility of the studies using a study selection form. A third reviewer resolved discrepancies.
Data extraction and management
Two review authors independently extracted data using a standard form and then cross‐checked them. A second person confirmed all numeric calculations and graphic interpolations.
Assessment of risk of bias in included studies
We assessed the risk of bias with the standard Cochrane 'Risk of bias' tool. The domains assessed included allocation sequence generation, allocation concealment, blinding of participants, personnel and outcome assessors, completeness of participant follow‐up, handling of incomplete outcome data and protection against selective outcome reporting.
Measures of treatment effect
The position of the patient during blood pressure measurement might affect the blood pressure reading or true lowering effect. However, in order not to lose valuable data, data reported from any single positions were used, regardless of the position. When blood pressure measurement data were available from more than one position, sitting blood pressure was the first preference. If both standing and supine were available, standing blood pressure was used. Effect measures were reported as the mean difference (MD) in BP, heart rate and pulse pressure between the treatment and placebo groups with 95% confidence interval (CI). Risk ratio (RR) was reported for withdrawal due to adverse effects.
Dealing with missing data
In the case of missing information in the included studies, we contacted the investigators (using email, letter, fax or a combination) to obtain the missing information.
In the case of missing standard deviations (SD) of the change in blood pressure, we imputed the standard deviation, based on the information in the same study or from other studies using the same drug. We used the following hierarchy (listed from high to low preference) to impute standard deviation values:
-
standard deviation of change in blood pressure taken in a different position than that of the blood pressure data used;
-
standard deviation of blood pressure at the end of treatment;
-
standard deviation of blood pressure at the end of treatment measured in a different position than that of the blood pressure data used;
-
standard deviation of blood pressure at baseline (unless this measure was used for entry criteria);
-
mean standard deviation of change in blood pressure from other studies using the same drug.
Assessment of heterogeneity
We test for heterogeneity of treatment effect between the studies by using a standard Chi2 statistic for heterogeneity. We used the fixed‐effect model to obtain summary statistics of pooled studies, unless significant between‐study heterogeneity was present, in which case we used the random‐effects model.
Data synthesis
We used the Cochrane Review Manager software, RevMan 5.3 for data syntheses and analyses.
We combined data for changes in blood pressure and heart rate using a weighted mean difference method (WMD). We analyzed drop‐outs due to side effects by using relative risk. When we found statistically significant relative risk, we calculated risk difference, and number needed to harm.
When we used the generic inverse variance method to incorporate cross‐over studies into meta‐analysis, we used the formula listed in the Cochrane Handbook for Systematic Reviews of Interventions, section 16.4.6.1 (Higgins 2011) to calculate the standard deviation of the difference between treatment and placebo. The standard error and sample sizes shown in the data analyses tables were unadjusted for subgroup comparisons in order to minimize the loss of statistical power for the estimates. To avoid double counting of patients when comparing multiple subgroups or combining two subgroups for overall estimates, we adjusted the standard error or the sample size in analyses for studies containing multiple dosage subgroups.
Subgroup analysis and investigation of heterogeneity
If possible, subgroup analyses included:
-
Different regimens of the same active chemical entity.
-
Sex, Age and Race.
-
Co‐morbid conditions: ischemic heart disease, peripheral vascular disease, diabetes.
-
Baseline severity of hypertension: mild, moderate, severe.
Sensitivity analysis
The robustness of the results were tested using several sensitivity analyses, including:
-
Studies that were industry‐sponsored versus non‐industry sponsored
-
Studies with blood pressure data measured in the sitting position versus other positions
-
Studies with reported standard deviations of blood pressure change versus imputed standard deviations
Results
Description of studies
Results of the search
In order to save time and effort, our trial search coordinator developed a comprehensive search strategy so that all four subclasses of beta blockers were searched simultaneously (Appendix 1; Appendix 2; Appendix 3; Appendix 4; Appendix 5) and three additional beta blocker reviews (Wong 2008; Wong 2014a; Wong 2014b) used the same study inclusion criteria. Citations were sorted according to their respective subclasses afterward. Please refer to Figure 1 for the flow of study selection. The search was first run in May 2010; subsequent searches were performed up to October 2014. We identified a total of 22,195 citations, 8353 of which were confirmed to be duplicates. The review authors then screened 13,842 titles and abstracts, 13,229 citations of which were excluded. Based on titles and abstracts, they judged that 613potentially met the inclusion criteria; these were retrieved for detailed review. Four hundred and ninty‐seven full‐text articles did not meet our inclusion criteria and were excluded. One hundred and fifteen studies met our inclusion criteria. Twelve studies, among the 115 studies, studied dual receptor blockers. Four studies were excluded with reasons listed in the Characteristics of excluded studies table. Eight studies were included in this review.
Study flow diagram
Included studies
Please refer to the Characteristics of included studies table for detailed descriptions of each included study. We included eight RCTs, that examined the blood pressure lowering efficacy of dual receptor blockers in 1493 hypertensive patients for 3 to 12 weeks. Five of the eight included studies were parallel studies; the other three were cross‐over studies. Four RCTs studied 12.5 mg/day to 50 mg/day doses of carvedilol in 1381 patients; the other four RCTs studied 300 mg/day to 800 mg/day doses of labetalol in 112 patients. Weighted mean baseline DBP was 101 millimetres of mercury (mm Hg) in carvedilol studies and 105 mm Hg in labetalol studies. Baseline SBP was not given in most included studies.
Excluded studies
Four studies that met the inclusion criteria were excluded after detailed reading. The reasons for exclusion were lack of useful data for outcomes listed in the review and misleading information on methods. Please refer to the Characteristics of excluded studies table for the reasons of exclusion.
Risk of bias in included studies
Figure 2 shows the 'Risk of bias' summary of each included study.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Allocation
Kane 1976 and Lechi 1982 described the details of methods of randomization and allocation concealment. The other six included studies did not report any information on this matter, therefore, it was difficult to assess the risk of bias in this category. The baseline characteristics of parallel groups were similar, thus we did not find any evidence that suggested randomization was not properly done. It was impossible to assess the integrity of randomization in cross‐over studies without additional information. Five out of 8 of the included studies in this review were parallel studies therefore the risk of selection bias was low in this review. Even though we did not find any evidence to suspect any bias in randomization and allocation concealment, more precise reporting would allow us to better assess the risk of bias and quality of the evidence.
Blinding
Beta blockers generally present a high risk of detection bias as they lower heart rate. It was likely possible for assessors to detect the treatment group based on the change in heart rate. This risk could be mitigated by using an automated BP machine. However, none of the studies in this review reported using automated BP machines. Therefore, the risk for detection and performance bias was high in this review (Jadad 1996).
Incomplete outcome data
All but one included study, Frick 1976, reported the information on dropouts. The dropout rate was low in the included studies. Therefore the risk of attrition bias was low in this review.
Selective reporting
McPhillips 1988 did not report the data for heart rate according to intervention groups and Frick 1976 did not report withdrawal due to adverse effects (WDAE). Other than that, blood pressure and heart rate were reported in all studies. In addition, five of the eight studies reported useful WDAE. We judged that the risk of selective reporting bias was low in this review.
Publication bias
It is possible that large pivotal studies that would have been required to support the application for a FDA approval for labetalol were not published. As a result, the magnitude of the blood pressure lowering effect was likely over‐estimated for labetalol. It is important that both positive and negative data are made available to provide the best estimate of the true effect.
Effects of interventions
See: Summary of findings for the main comparison
Carvedilol
Carvedilol is indicated for the treatment of hypertension, left ventricular dysfunction following MI and heart failure in the U.S. (FDA). It is indicated only for heart failure in Canada (eCPS). However, it can be used off label for hypertension. The recommended doses for hypertension in the U.S. are 6.25 mg twice daily to 25 mg twice daily for the immediate release form and 20 to 80 mg daily for the controlled release form.
We included four studies that examined the blood pressure lowering efficacy of 6.25 to 50 mg/day carvedilol in 1381 mild to moderate hypertensive patients (Carvedilol GSK B100; Carvedilol GSK B101; McPhillips 1988; Weber 2006). Carvedilol GSK B100 and Carvedilol GSK B101 were two unpublished RCTs obtained from the manufacturer's web‐site. The unpublished data contributed 72% of the data (1000 of 1381 subjects). Patients enrolled in these studies were on average 52 to 56 years old, with no co‐morbidities that could affect their blood pressure. The weighted mean baseline DBP of the four included studies was 100.9 mm Hg. Three of the four studies measured blood pressure at trough hours (Carvedilol GSK B100; Carvedilol GSK B101; Weber 2006).
Carvedilol 6.25 mg/day did not significantly lower SBP or DBP compared to placebo (SBP = ‐1 mmHg [‐5, 3]; DBP= ‐1 mmHg [‐3, 2];Analysis 1.1 & Analysis 1.2). Starting from 12.5 mg/day (the recommended starting dose), carvedilol significantly lowered SBP and DBP compared to placebo (Analysis 1.1 & Analysis 1.2). Heterogeneity was not significant in any of the subgroups (I2=0%). All 4 studies had multiple dose subgroups, allowing direct comparison between doses. Tests for subgroup differences by direct comparison were not significant for SBP and DBP within the product monograph range of recommended doses. (12.5 mg/day, 25 mg/day and 50 mg/day). The 1x and 2x starting dose (12.5 mg/day and 25 mg/day) subgroups contained the largest patient sample in carvedilol. The estimate for combining the average trough BP lowering effect combining the 1x and 2x starting dose (12.5 mg/day and 25 mg/day) carvedilol was ‐4/‐3 mmHg.
Carvedilol 6.25 mg/day did not significantly lower heart rate compared to placebo (‐1 BPM [‐4, 1]; Analysis 1.3). Tests for subgroup differences between carvedilol 12.5 mg/day and higher doses were significant for heart rate (P < 0.0005; Analysis 1.3). Carvedilol did not significantly change pulse pressure compared to placebo in any of the dose subgroups (Analysis 1.4).
Blood pressure variability was assessed by comparing the end treatment SD of carvedilol and placebo groups with a t‐test. Twelve SDs from carvedilol subgroups and four SDs from placebo subgroups were included. Carvedilol did not significantly change blood pressure variability in SBP (P = 0.37) or DBP (P = 0.65). The weighted mean SD for carvedilol group (SBP/DBP) was 15.9/9.0 and for placebo group was 16.4/9.2.
Labetalol
Labetalol is indicated for the treatment of hypertension in both Canada and the U.S. (eCPS; FDA). The recommended doses for hypertension in U.S. and Canada are 100 mg twice a day to 400 mg twice a day.
Four studies examined the BP lowering efficacy of 400 to 800 mg/day labetalol in 112 hypertensive patients were included in the analysis (Frick 1976; Kane 1976; Lechi 1982; Rouffy 1986). The baseline characteristics of patients were not well described in the articles. The mean baseline BP of patients enrolled in the studies was 165/105 mm Hg. Sixty per cent of the patients were middle age men with no co‐morbidities.
We did not find any data for the recommended starting dose (200 mg/day) of labetalol. Both labetalol 400 and 800 mg/day subgroups significantly lowered SBP and DBP (Analysis 2.1; Analysis 2.2). Tests for subgroup differences were significant by direct comparison for SBP (p= 0.09) and DBP (p= 0.0006). Heterogeneity was not significant in any of the subgroups (both I2=0%). Both 400 and 800 mg/day labetalol significantly lowered heart rate (400 mg/day :‐7 BPM [‐9, ‐5]; 800 mg/day ‐8 [‐10, ‐5]; Analysis 2.3). The effect on heart rate was not significantly different between 400 and 800 mg/day labetalol (p=0.6).
Labetalol 400 mg/day significantly reduced pulse pressure (‐4 mmHg [‐7, ‐0], p=0.02; Analysis 2.4), but 800 mg/day subgroup did not significantly change pulse pressure (‐5 [‐10, 0], p=0.07; Analysis 2.4).
Pooled hemodynamic effects of dual blockers
Our analyses showed that the BP lowering effect of carvedilol was much smaller than labetalol. If the difference in effect size was true, it would not be appropriate to pool the data. However, the difference of effect size might be the result of small sample size or bias. The possible reasons for the differences between the drugs are explained in the discussion. Since only carvedilol provided data for the recommended starting dose (12.5 mg/day), the pooled effect size at the recommended starting does was the same as carvedilol. Dual blocker lowered BP (SBP/DBP) by ‐6 mmHg [‐8, ‐4]/‐4 mmHg [‐5, ‐3] at twice the recommended starting dose (25 mg/day carvedilol and 400 mg/day labetalol). The BP lowering effect was ‐9 mmHg [‐12, ‐7]/‐7 mmHg [‐8, ‐5] at four times the recommended starting dose (50 mg/day carvedilol and 800 mg/day labetalol). Please refer to Analysis 3.1; Analysis 3.2; Analysis 3.3; Analysis 3.4; Analysis 3.5 for the pooled effect of dual alpha and beta blockers.
summary of findings Table for the main comparison also summarizes the results for combining the starting dose and twice the starting dose (Analysis 4.1; Analysis 4.2; Analysis 4.3; Analysis 4.4).
Withdrawal due to adverse effects (WDAE)
We pooled the WDAE data of carvedilol and labetalol together in order to include all valuable data. Five studies provided data for WDAE in 1412 patients (Carvedilol GSK B100; Carvedilol GSK B101; McPhillips 1988; Rouffy 1986; Weber 2006). Four of the studies used carvedilol as the active treatment (Carvedilol GSK B100; Carvedilol GSK B101; McPhillips 1988; Weber 2006).There was no significant difference in relative risk (RR) of WDAE between dual blocker treatment and placebo (RR 0.88;95% CI 0.54 to 1.42;P = 0.6).
Subgroup and sensitivity analyses
With the exception of baseline BP, the baseline characteristics of patients were similar in all studies. None of the studies reported data separated by subgroups. Therefore, the subgroup analyses for sex, co‐morbid conditions, race, age or severity of disease was not possible.
Difference between published and unpublished data
We performed a sensitivity analysis in order to examine the differences between estimated mean effect size of published and unpublished studies. The estimated means (SBP/DBP) for the 2x starting dose subgroup (25 mg/day carvedilol and 400 mg/day labetalol) from the published data were ‐3/‐4 mm Hg as compared to ‐5/‐2 mm Hg from the unpublished data. The estimate of SBP/DBP in 4x starting dose (50 mg/day carvedilol and 800 mg/day labetalol) subgroup from published data was ‐8/‐7 mm Hg as compared to ‐5/‐4 mm Hg from the unpublished data. For the 50 mg/day carvedilol and 800 mg/day labetalol, the published estimate exceeded the unpublished estimate by ‐3/‐3 mm Hg.
Discussion
Summary of main results
Carvedilol
Carvedilol 12.5 mg/day and higher doses significantly lowered blood pressure (BP). This finding coincided with the product monograph recommendation. However, the current evidence was inconclusive whether there was a dose response effect. The BP lowering point estimates were higher for higher doses, but the tests for subgroup differences were not significant. The large number of subjects in carvedilol studies, particularly the data coming from the unpublished studies, provided good estimates of the blood pressure lowering effect of carvedilol. The average trough BP lowering effect of once (12.5 mg/day carvedilol) and twice (25 mg/day carvedilol) the recommended starting dose was only ‐4/‐3 mm Hg. This suggested that carvedilol was relatively ineffective at lowering BP compared to other classes of antihypertensive drugs or other beta blockers.
Differences in BP lowering effect between carvedilol and labetalol
The number of patients included in studies for labetalol was much smaller than for carvedilol. In addition, we did not find any data for 200 mg/day labetalol, which was the recommended starting dose. This suggested that some studies of the BP lowering effect of labetalol have been conducted but have never been published. From the studies available, labetalol appeared to significantly lower BP to a greater extent than carvedilol. The average peak BP lowering effect of labetalol at twice (400 mg/day labetalol) the recommended starting dose was ‐10 [‐14, ‐7]/‐7 [‐9, ‐5].
The pharmacology of both dual blockers was examined in order to explore the possible explanation for the differences. In basic pharmacology studies, carvedilol and labetalol showed similar properties in pharmacokinetic parameters such as time to maximum serum concentration, elimination half life and receptor affinity (eCPS; FDA; Goodman 2011; McTavish 1993). This suggests that the difference in BP lowering effect that we observed was unlikely to be the result of differences in pharmacological properties.
Next, we examined the differences in baseline BP. The mean baseline blood pressure of labetalol studies (105 mm Hg) was higher than in the carvedilol studies (101 mm Hg). Because of this, the absolute blood pressure lowering effect could be bigger for labetalol. However, if the BP lowering was expressed as the DBP percent change from baseline, it was still larger for labetalol (‐6%) as compared to carvedilol (‐4%). This suggested that differences in baseline BP could only partly explain the differences in effect size.
Three of the four carvedilol studies stated that the blood pressure was measured at trough hours (from 13 to 24 hours after last dose). None of the labetalol studies specified whether blood pressure was measured at peak or trough hours. However, since labetalol was taken two to three times a day, the blood pressure must have been measured at peak hours (from 1 to 12 hours after the dose) in these studies. Measurements during the peak effect time could result in greater blood pressure lowering effect. Therefore, the difference in time of measurements could also contribute to the difference in effect size.
The paucity of large studies assessing the BP lowering effect of labetalol indicated that the pivotal studies that would have been required to achieve a license in various developed countries have not been published and the evidence supporting the use of labetalol for hypertension was weak. Our search for unpublished data for labetalol did not yield any additional studies. The pharmacological properties, basic study design and study population were similar in carvedilol and labetalol studies. Compared to the smaller effect size for carvedilol, the labetalol estimates could be exaggerated due to lack of large pivotal studies. If the difference in baseline BP and time of measurements were taken into account, it is possible that the effect sizes of the two drugs are actually not different. In addition, labetalol needs to be taken two to three times a day compared to the carvedilol controlled‐release tablet, which is taken once daily for hypertension. Physicians and patients making decisions to use dual blockers should be made aware of the small BP lowering effect of this class of drugs.
Differences in published and unpublished data
From the sensitivity analysis with carvedilol, we found that published studies could overestimate the BP lowering effect by as much as ‐3 mm Hg systolic and ‐3 mm Hg diastolic compared to unpublished studies. This was one of the first examples in which the effect of publication bias could be quantified. It showed the amount of over‐estimation caused by a deliberate decision to publish studies that showed greater effect, while not publishing studies that showed less effect. This provides good support for the extra effort it takes to look for unpublished data. However, despite our efforts to search, many studies remained completely hidden. Cochrane has been advocating for the publication of all data, positive and negative, for many years. In recent years, pressure from Cochrane and the European regulatory agency has resulted in the release of some unpublished data from the industry (NY Times 2013).
In addition, the public are becoming more aware of the impact that publication bias may have on medical decision‐making. This is a positive change. It is likely that the unpublished data we found on the GlaxoSmithKline (GSK) web site are the result of the efforts by many Cochrane researchers pushing the company to release all RCT data.
Overall completeness and applicability of evidence
This review provides the most current evidence for the BP lowering efficacy of dual blockers. The studies included in this review fit our objectives in terms of population, goal of the studies and outcomes measured. Our search strategy was comprehensive and thorough. In addition to published data, our extensive search also led us to obtain unpublished data. Therefore, we are confident that this review represents the best evidence available for BP lowering efficacy of these drugs.
Quality of the evidence
summary of findings Table for the main comparison summarizes the effect size of the starting dose and twice the starting dose and provides a judgment of the quality of evidence in this review. This review included eight studies examining two dual blockers in 1493 hypertensive patients. The sample size was sufficient to allow a robust conclusion with regard to carvedilol, primarily because we were able to obtain data on two large unpublished RCTs. However, like all systematic reviews, we were limited by the data that were available to us.
The data were insufficient for labetalol and we found no RCTs for dilevalol. As discussed above, data from large pivotal studies were missing in the labetalol analyses. As a result, the magnitude of the blood pressure lowering effect was likely over‐estimated for labetalol. For this reason, the risk of publication bias for labetalol is high. The risk of publication bias for carvedilol is much lower because of the two unpublished studies. The quality of evidence would be upgraded to moderate if we only considered estimates for carvedilol.
Similar to the other three reviews on beta blockers (Wong 2008; Wong 2014a; Wong 2014b), most of the studies of dual blockers did not use automated BP machines, which could have mitigated the risk of detection bias caused by loss of blinding. For this reason, the risk of detection bias remains high in this review and the quality of evidence was downgraded by one level.
Similar to the included studies of the same three reviews, none of the studies included in this review reported pulse pressure as one of their outcomes. Data for pulse pressure were not direct measurements from the studies. We calculated pulse pressure by subtracting DBP from SBP. Indirectness of this outcome was the reason that quality of evidence was further downgraded for pulse pressure to very low.
Potential biases in the review process
The rigidity of the inclusion criteria minimized the potential of bias during the selection process. Study selection was based solely on methodology. The responsibility of the review authors was to determine if the methodology fit the inclusion criteria. No other part of the studies played a role in the decision. The strict inclusion criteria ensured that included studies were well conducted in order to minimize the risk of bias during the process.
Agreements and disagreements with other studies or reviews
Two published reviews, Chen 2010a and Law 2005, combined results from all subclasses of beta blockers and estimated the overall BP lowering effect of beta blockers. They did not report separate results for different subclasses of beta blockers or individual drugs. Chen 2010a also examined the additional effects of beta blockers when used as second‐line treatment of hypertension. Our review examined beta blockers as first‐line monotherapy for hypertension. Because of these differences, this review is not directly comparable to these reviews.
Other Cochrane reviews have compared antihypertensive drug classes with placebo and used similar inclusion and exclusion criteria to this review. Based on an indirect comparison with other classes, dual receptor blockers lowered trough BP by ‐6/‐4 mm Hg, a smaller magnitude on average than thiazide diuretics (‐9/‐4 mm Hg; Musini 2014), ACE inhibitors (‐8/‐5 mm Hg; Heran 2008a), ARBs (‐9/‐6 mm Hg; Heran 2008b), and renin inhibitors (‐8/‐5 mm Hg; Musini 2008). Dual receptor blockers also lowered BP by a smaller magnitude than nonselective beta blockers (‐10/‐7 mm Hg; Wong 2014a), partial agonists (‐8/‐4 mm Hg; Wong 2014b) and beta1 selective beta blockers (‐10/‐8 mm Hg; Wong 2008). However, the difference in mean baseline BP (SBP/DBP) between dual receptor blockers and partial agonists (151/101 mm Hg versus 175/107 mm Hg) might partly explain the difference in BP lowering effect between these two subclasses. In addition, most of the included studies in the nonselective beta blocker review measured BP during peak hours while the measurements of dual receptor blocker studies were made at trough (Wong 2014a).
Like other beta blockers, the fact that dual blockers reduce DBP to an amount similar to SBP means that they have little or no effect on pulse pressure. This is similar to the other classes of beta blockers, but less than the average reduction of pulse pressure of 5 mm Hg seen with thiazides (Musini 2014) and 3 mm Hg seen with drugs inhibiting the renin angiotensin system (Heran 2008a; Heran 2008b; Musini 2008). These two differences in effect on blood pressure, little or no effect on pulse pressure and a lesser blood pressure lowering effect, might provide some insight into why first‐line beta blockers do not reduce mortality and morbidity as much as first‐line thiazide diuretics (Wisonge 2007; Wright 2000; Wright 2009), first‐line calcium channel blockers (Chen 2010b) and first‐line drugs inhibiting the renin angiotensin system (Xue 2015).
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Comparison 1 Carvedilol vs placebo, Outcome 1 systolic blood pressure.
Comparison 1 Carvedilol vs placebo, Outcome 2 diastolic blood pressure.
Comparison 1 Carvedilol vs placebo, Outcome 3 Heart rate.
Comparison 1 Carvedilol vs placebo, Outcome 4 Pulse pressure.
Comparison 2 Labetalol vs placebo, Outcome 1 systolic blood pressure.
Comparison 2 Labetalol vs placebo, Outcome 2 diastolic blood pressure.
Comparison 2 Labetalol vs placebo, Outcome 3 Heart rate.
Comparison 2 Labetalol vs placebo, Outcome 4 Pulse pressure.
Comparison 3 Pooled overall effects, Outcome 1 systolic blood pressure.
Comparison 3 Pooled overall effects, Outcome 2 diastolic blood pressure.
Comparison 3 Pooled overall effects, Outcome 3 Heart rate.
Comparison 3 Pooled overall effects, Outcome 4 Pulse Pressure.
Comparison 3 Pooled overall effects, Outcome 5 withdrawal due to adverse effects.
Comparison 4 Adjusted combined once and twice starting dose effects, Outcome 1 systolic blood pressure.
Comparison 4 Adjusted combined once and twice starting dose effects, Outcome 2 diastolic blood pressure.
Comparison 4 Adjusted combined once and twice starting dose effects, Outcome 3 Heart rate.
Comparison 4 Adjusted combined once and twice starting dose effects, Outcome 4 Pulse Pressure.
| Dual receptor blockers compared with placebo for primary hypertension | |||
| Patient or population: Adults with primary hypertension Intervention: Alpha and beta dual receptor blockers Comparison: Placebo | |||
| Outcomes | Mean estimates of combining once and twice starting dose (95% CI) | No of Participants | Quality of the evidence |
| Systolic blood pressure | ‐5.59 (95% CI ‐7.47 to ‐3.70)1,2,3 | 1007 (8) | Low4,5 |
| Diastolic blood pressure | ‐3.88 (95% CI ‐4.95 to ‐2.82)1,2,3 | 1007 (8) | Low4,5 |
| Heart rate | ‐4.62 (95% CI ‐5.71 to ‐3.54)1,2 | 977 (7) | Low4,5 |
| Pulse pressure | ‐1.89 (95% CI ‐3.58 to ‐0.20)1,2 | 1007 (8) | Very low4,5,6 |
| WDAE | 0.88 (95% CI 0.54 to 1.42) | 1412 (5) | Moderate7 |
| SBP: systolic blood pressure; DBP: diastolic blood pressure; WDAE: withdrawal due to adverse effects; 95%CI: 95% confident interval | |||
| GRADE Working Group grades of evidence | |||
| Footnotes | |||
|
| |||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 systolic blood pressure Show forest plot | 4 | Mean Difference (IV, Fixed, 95% CI) | Subtotals only | |
| 1.1 (0.5 initial) 6.25 mg/day | 1 | 236 | Mean Difference (IV, Fixed, 95% CI) | ‐0.90 [‐5.20, 3.40] |
| 1.2 (initial dose) 12.5 mg/day | 3 | 494 | Mean Difference (IV, Fixed, 95% CI) | ‐3.53 [‐6.38, ‐0.68] |
| 1.3 (2x initial) 20 & 25 mg/day | 4 | 644 | Mean Difference (IV, Fixed, 95% CI) | ‐4.09 [‐6.52, ‐1.65] |
| 1.4 (4x initial) 40 & 50 mg/day | 4 | 527 | Mean Difference (IV, Fixed, 95% CI) | ‐6.51 [‐9.35, ‐3.66] |
| 2 diastolic blood pressure Show forest plot | 4 | Mean Difference (IV, Fixed, 95% CI) | Subtotals only | |
| 2.1 (0.5 initial) 6.25 mg/day | 1 | 236 | Mean Difference (IV, Fixed, 95% CI) | ‐0.70 [‐3.06, 1.66] |
| 2.2 (initial dose) 12.5 mg/day | 3 | 494 | Mean Difference (IV, Fixed, 95% CI) | ‐2.26 [‐3.88, ‐0.63] |
| 2.3 (2x initial) 20 & 25 mg/day | 4 | 644 | Mean Difference (IV, Fixed, 95% CI) | ‐2.99 [‐4.38, ‐1.60] |
| 2.4 (4x initial) 40 & 50 mg/day | 4 | 527 | Mean Difference (IV, Fixed, 95% CI) | ‐5.04 [‐6.64, ‐3.44] |
| 3 Heart rate Show forest plot | 3 | Mean Difference (IV, Fixed, 95% CI) | Subtotals only | |
| 3.1 (0.5 initial) 6.25 mg/day | 1 | 236 | Mean Difference (IV, Fixed, 95% CI) | ‐1.4 [‐3.89, 1.09] |
| 3.2 (initial dose) 12.5 mg/day | 2 | 474 | Mean Difference (IV, Fixed, 95% CI) | ‐2.15 [‐3.86, ‐0.43] |
| 3.3 (2x initial) 20 & 25 mg/day | 3 | 624 | Mean Difference (IV, Fixed, 95% CI) | ‐4.68 [‐6.05, ‐3.31] |
| 3.4 (4x initial) 40 & 50 mg/day | 3 | 506 | Mean Difference (IV, Fixed, 95% CI) | ‐6.33 [‐7.93, ‐4.74] |
| 4 Pulse pressure Show forest plot | 4 | Mean Difference (IV, Fixed, 95% CI) | Subtotals only | |
| 4.1 (0.5 initial) 6.25 mg/day | 1 | 236 | Mean Difference (IV, Fixed, 95% CI) | ‐0.20 [‐3.93, 3.53] |
| 4.2 (initial dose) 12.5 mg/day | 3 | 494 | Mean Difference (IV, Fixed, 95% CI) | ‐1.24 [‐3.72, 1.23] |
| 4.3 (2x initial) 20 & 25 mg/day | 4 | 644 | Mean Difference (IV, Fixed, 95% CI) | ‐1.19 [‐3.32, 0.93] |
| 4.4 (4x initial) 40 & 50 mg/day | 4 | 527 | Mean Difference (IV, Fixed, 95% CI) | ‐1.58 [‐4.06, 0.91] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 systolic blood pressure Show forest plot | 4 | Mean Difference (Fixed, 95% CI) | Subtotals only | |
| 1.1 (2x initial) 300 & 400 mg/day | 4 | 110 | Mean Difference (Fixed, 95% CI) | ‐10.36 [‐13.95, ‐6.77] |
| 1.2 (4x initial) 600 & 800 mg/day | 2 | 44 | Mean Difference (Fixed, 95% CI) | ‐19.69 [‐25.38, ‐13.99] |
| 2 diastolic blood pressure Show forest plot | 4 | Mean Difference (Fixed, 95% CI) | Subtotals only | |
| 2.1 (2x initial) 300 & 400 mg/day | 4 | 110 | Mean Difference (Fixed, 95% CI) | ‐6.59 [‐8.60, ‐4.58] |
| 2.2 (4x initial) 600 & 800 mg/day | 2 | 44 | Mean Difference (Fixed, 95% CI) | ‐14.58 [‐18.02, ‐11.15] |
| 3 Heart rate Show forest plot | 4 | Mean Difference (Fixed, 95% CI) | Subtotals only | |
| 3.1 (2x initial) 300 & 400 mg/day | 4 | 110 | Mean Difference (Fixed, 95% CI) | ‐7.07 [‐9.13, ‐5.01] |
| 3.2 (4x initial) 600 & 800 mg/day | 2 | 44 | Mean Difference (Fixed, 95% CI) | ‐7.67 [‐10.43, ‐4.91] |
| 4 Pulse pressure Show forest plot | 4 | Mean Difference (Fixed, 95% CI) | Subtotals only | |
| 4.1 (2x initial) 300 & 400mg/day | 4 | 110 | Mean Difference (Fixed, 95% CI) | ‐3.59 [‐6.73, ‐0.46] |
| 4.2 (4x initial) 600 & 800 mg/day | 2 | 44 | Mean Difference (Fixed, 95% CI) | ‐4.64 [‐9.65, 0.37] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 systolic blood pressure Show forest plot | 8 | Mean Difference (Fixed, 95% CI) | Subtotals only | |
| 1.1 0.5x initial doses | 1 | 236 | Mean Difference (Fixed, 95% CI) | ‐0.90 [‐5.19, 3.39] |
| 1.2 1x initial doses | 3 | 494 | Mean Difference (Fixed, 95% CI) | ‐3.53 [‐6.38, ‐0.69] |
| 1.3 2x initial doses | 8 | 754 | Mean Difference (Fixed, 95% CI) | ‐6.06 [‐8.08, ‐4.05] |
| 1.4 4x initial doses | 6 | 571 | Mean Difference (Fixed, 95% CI) | ‐9.16 [‐11.73, ‐6.60] |
| 2 diastolic blood pressure Show forest plot | 8 | Mean Difference (Fixed, 95% CI) | Subtotals only | |
| 2.1 0.5x initial doses | 1 | 236 | Mean Difference (Fixed, 95% CI) | ‐0.7 [‐3.05, 1.65] |
| 2.2 1x initial doses | 3 | 494 | Mean Difference (Fixed, 95% CI) | ‐2.26 [‐3.88, ‐0.64] |
| 2.3 2x initial doses | 8 | 754 | Mean Difference (Fixed, 95% CI) | ‐4.16 [‐5.30, ‐3.01] |
| 2.4 4x initial doses | 6 | 571 | Mean Difference (Fixed, 95% CI) | ‐6.74 [‐8.19, ‐5.29] |
| 3 Heart rate Show forest plot | 7 | Mean Difference (Fixed, 95% CI) | Subtotals only | |
| 3.1 0.5x initial doses | 1 | 237 | Mean Difference (Fixed, 95% CI) | ‐1.4 [‐3.89, 1.09] |
| 3.2 1x initial doses | 2 | 474 | Mean Difference (Fixed, 95% CI) | ‐2.14 [‐3.85, ‐0.43] |
| 3.3 2x initial doses | 7 | 734 | Mean Difference (Fixed, 95% CI) | ‐5.41 [‐6.56, ‐4.27] |
| 3.4 4x initial doses | 5 | 550 | Mean Difference (Fixed, 95% CI) | ‐6.67 [‐8.05, ‐5.29] |
| 4 Pulse Pressure Show forest plot | 8 | Mean Difference (Fixed, 95% CI) | Subtotals only | |
| 4.1 0.5x initial doses | 1 | 236 | Mean Difference (Fixed, 95% CI) | ‐0.2 [‐3.92, 3.52] |
| 4.2 1x initial doses | 3 | 494 | Mean Difference (Fixed, 95% CI) | ‐1.25 [‐3.73, 1.23] |
| 4.3 2x initial doses | 8 | 754 | Mean Difference (Fixed, 95% CI) | ‐2.19 [‐4.01, ‐0.37] |
| 4.4 4x initial doses | 6 | 571 | Mean Difference (Fixed, 95% CI) | ‐2.43 [‐5.18, 0.31] |
| 5 withdrawal due to adverse effects Show forest plot | 5 | 1412 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.88 [0.54, 1.42] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 systolic blood pressure Show forest plot | 8 | 1007 | Mean Difference (Fixed, 95% CI) | ‐5.59 [‐7.47, ‐3.70] |
| 1.1 1x initial doses | 3 | 374 | Mean Difference (Fixed, 95% CI) | ‐3.51 [‐7.09, 0.07] |
| 1.2 2x initial doses | 8 | 633 | Mean Difference (Fixed, 95% CI) | ‐6.39 [‐8.61, ‐4.16] |
| 2 diastolic blood pressure Show forest plot | 8 | 1007 | Mean Difference (Fixed, 95% CI) | ‐3.88 [‐4.95, ‐2.82] |
| 2.1 1x initial doses | 3 | 374 | Mean Difference (Fixed, 95% CI) | ‐2.26 [‐4.29, ‐0.24] |
| 2.2 2x initial doses | 8 | 633 | Mean Difference (Fixed, 95% CI) | ‐4.50 [‐5.75, ‐3.25] |
| 3 Heart rate Show forest plot | 7 | 977 | Mean Difference (Fixed, 95% CI) | ‐4.62 [‐5.71, ‐3.54] |
| 3.1 1x initial doses | 2 | 359 | Mean Difference (Fixed, 95% CI) | ‐2.16 [‐4.28, ‐0.04] |
| 3.2 2x initial doses | 7 | 618 | Mean Difference (Fixed, 95% CI) | ‐5.50 [‐6.77, ‐4.24] |
| 4 Pulse Pressure Show forest plot | 8 | 1007 | Mean Difference (Fixed, 95% CI) | ‐1.89 [‐3.58, ‐0.20] |
| 4.1 1x initial doses | 3 | 374 | Mean Difference (Fixed, 95% CI) | ‐1.22 [‐4.32, 1.88] |
| 4.2 2x initial doses | 8 | 633 | Mean Difference (Fixed, 95% CI) | ‐2.17 [‐4.19, ‐0.15] |
