Article Text
PDF
Abstract
Introduction Hypertension, a prevalent cardiovascular disease globally, poses significant health risks and economic burden. Evolving treatment targets necessitate more intensive strategies, such as low-dose triple or quadruple drug combinations. However, a systematic comparison of different low-dose antihypertensive combinations is still lacking. The aim of the present study is to systematically and comprehensively evaluate the blood pressure-lowering effect and the associated safety of diverse low-dose polypharmacy combinations in patients with hypertension.
Methods and analysis In this systematic review and network meta-analysis, randomised controlled trials comparing diverse low-dose polypharmacy combinations with placebo or active treatments in patients with hypertension will be eligible for inclusion. The primary outcomes are a reduction in systolic/diastolic blood pressure, the rate of target blood pressure, adverse effects, serious adverse effects and all-cause dropout after treatment. PubMed, Web of Science, Embase, Cochrane Library, Chinese Science Citation Database, Wanfang Medical Network, VIP Database and clinical trial registries will be systematically searched for relevant studies published from inception date to 18 January 2024. No language restrictions will be applied during the search process. Two independent reviewers will identify eligible trials and extract the data. Traditional pairwise meta-analysis will be conducted to analyse direct comparisons. A frequentist approach will be used to analyse the primary outcome for network comparisons, and cumulative rank probabilities will present the treatment hierarchy of all endpoints. Sensitivity analysis will be conducted using a Bayesian framework under a random-effects model. Subgroup analyses will be conducted according to sample size, quality of study and sponsorship, if the data allow. The Cochrane Risk of Bias Tool 2.0 will be used to assess the quality of the included studies. The Grading of Recommendations, Assessment, Development, and Evaluation system will be used to assess the strength of evidence.
Ethics and dissemination Since this study relies solely on published literature, no ethics approval is necessary. The results will be submitted to a peer-reviewed journal.
PROSPERO registration number CRD42024503239.
- hypertension
- systematic review
- network meta-analysis
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
STRENGTHS AND LIMITATIONS OF THIS STUDY
This network meta-analysis will establish comprehensive evidence to compare low-dose polypharmacy across various antihypertensive combinations.
The planned study has well-defined objective and rigorous inclusion criteria and will use advanced methods for data collection and synthesis.
The overall quality of evidence will be assessed using the Grading of Recommendations, Assessment, Development, and Evaluation, allowing for assessment of certainty of evidence for the network meta-analysis.
Given the diversity in low-dose polypharmacy combinations and potential differences in methodological characteristics, significant heterogeneity is anticipated.
Introduction
Hypertension is the most common cardiovascular disease globally, with significant health implications and economic burden.1 2 Despite this, the awareness, treatment and control rates for hypertension are suboptimal. It is estimated that 1.28 billion people worldwide suffer from hypertension, with 45.3% of patients undiagnosed. The treatment rate among women stands at 47%, while that of men stands at 38%. On the other hand, the control rate in women is at 23%, and for men it is at 18%.3 4 In 2019, hypertension was responsible for 10.8 million deaths globally.5 According to the ‘China Cardiovascular Health and Disease Burden Report’,6 the direct economic burden of hypertension in China has reached ¥210.3 billion.
Most hypertension guidelines recommend dual therapy as the initial treatment.1 2 7–11 However, the pathogenesis of hypertension involves multiple mechanisms,12 13 and with societal progress and improvements in quality of life, the targets for blood pressure control have become increasingly stringent, moving from the initial 140 mm Hg to 130 mm Hg, with some specific populations being advised to aim for an optimal pressure of 120 mm Hg.14 Thus, dual therapy may not meet all patient needs, prompting a minority of studies to explore the initiation of treatment with low-dose triple or quadruple drug combinations.15–19
With respect to the efficacy and safety of initiating hypertension treatment with low-dose polypharmacy, one pairwise meta-analysis20 directly compared low-dose quadruple or triple drug regimens with other treatment strategies. Another meta-analysis synthesised evidence from randomised controlled trials comparing the quadpill with monotherapy or placebo.21 However, mixed controls and multiple comparisons in direct pairwise comparisons may underestimate the true effect of low-dose polypharmacy, potentially reducing the validity of the findings. Moreover, an evidence network directly comparing low-dose polypharmacy among various antihypertensive combinations is still lacking, including direct comparisons of low-dose triple versus dual combinations, or low-dose triple versus quadruple combinations. These limitations are inherent to traditional pairwise meta-analyses.
Therefore, this systematic review and network meta-analysis aims to systematically and comprehensively assess the blood pressure-lowering effect and related safety of low-dose polypharmacy combinations in populations with primary hypertension and to compare them with placebo or various active treatments.
Methods and analysis
This systematic review and network meta-analysis was registered on the International Prospective Register of Systematic Reviews (PROSPERO) with registration number CRD42024503239.22 The study protocol adheres to the guidelines of the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols.23
Data sources and search strategy
The search will be performed independently by two researchers, and differences will be resolved by a discussion with a third reviewer. Seven databases, namely PubMed, Web of Science, Embase, Cochrane Library, Chinese Science Citation Database, Wanfang Medical Network and VIP Database, will be searched up to 18 January 2024. Additionally, unpublished studies will be searched on ClinicalTrials.gov, the EU Clinical Trials Register and the Chinese Clinical Trial Registry. Manual searches of references from identified articles and previous systematic reviews will also be conducted. No language restrictions will be applied during the search process. The included studies will be reverified for eligibility before the final data analysis. We will search for original research articles using the following search terms (detailed search strings for each database are presented in online supplemental appendix S1):
Supplemental material
“hypertension” OR “hypertensive” OR “high blood pressure” OR “raised blood pressure” OR “elevated blood pressure” OR “HBP” OR “HTN”.
“low dose” OR “small dose” OR “half dose” OR “quarter dose” OR “minimal dose” OR “low dosage” OR “small dosage” OR “half dosage” OR “quarter dosage” OR “minimal dosage”.
“randomized”.
1 and 2 and 3.
Inclusion and exclusion criteria
Study selection will be conducted independently by two researchers, with any disagreements resolved by a third researcher. We will use professional translation services or engage with bilingual colleagues to translate and review any articles identified in non-native languages.
Eligibility criteria
Population: individuals aged 18 and over, regardless of sex or race, diagnosed with primary hypertension.
Intervention: combinations of dual, triple or quadruple low-dose antihypertensive drugs, regardless of fixed-dose combinations or free-drug combinations. All components must be in low doses and used in initiating treatment.
Comparison: active antihypertensive regimens or placebo. For active comparisons, the defined daily dose (DDD) must match that of the intervention group. If the control is also a combination therapy, it must contain a different number of antihypertensive drug classes from the intervention group. The DDD of placebo is set to 1.
Outcomes: must include at least one of the following outcomes: reduction in systolic/diastolic blood pressure, rate of achieving target blood pressure, and proportion of adverse events (AEs), serious adverse events (SAEs) and all-cause dropout after at least 4 weeks of treatment.
Study design: randomised controlled trials.
Exclusion criteria
Ongoing studies.
Studies for which relevant data cannot be obtained despite efforts.
Data management and extraction
Data will be recorded using EpiData (V.3.1). Two researchers will independently review the literature for eligibility, extracting relevant information including title, first author, year of publication, inclusion range of clinic blood pressure, inclusion of previously treated patients, number of study arms, detailed information on medications (name, dosage, administration), observation indicators, observation duration, study design (type, blinding, number of centres, geographic location), number of participants, baseline characteristics of participants (age, gender, average duration of hypertension, proportion of ever treatments, average body mass index, diabetes prevalence, smoking prevalence, average creatinine, baseline clinic blood pressure) and outcome events (reduction in systolic and diastolic blood pressure, number of patients achieving blood pressure targets, number of adverse and serious adverse events, total number of all-cause dropouts).
In cases of inconsistency during data extraction, a third researcher will adjudicate the discrepancies. For information that is unclear or not obtainable from the literature and its supplementary materials, the authors will be contacted via email for clarification.
The dosage of antihypertensive medications is defined as follows: standard doses are predetermined based on the 2018 Chinese Hypertension Guidelines1 and the WHO Collaborating Centre for Drug Statistics Methodology (https://www.whocc.no/atc_ddd_index/). A low dose is considered less than the standard dose, such as half, one-third or a quarter of the standard dose.
Given the similar antihypertensive effects of the five major classes of antihypertensive drugs, to control the confounding effects from inconsistencies in total DDD and ensure data comparability and transitivity, the DDD for low-dose combination drugs should match those of the control or placebo, with the placebo DDD set at 1.
Definition of treatment time
According to hypertension guidelines,1 blood pressure should be gradually reduced to target levels within 4–12 weeks after medication. Therefore, data extraction will be limited to the period between 4 and 12 weeks after medication, selecting the longest duration of data within this range if multiple outcomes were reported.
Risk of bias assessment
The risk of bias will be assessed using the Cochrane Risk of Bias Tool 2.0,24 independently conducted by two researchers, with disagreements resolved by a third researcher. This tool evaluates randomised controlled trials across five domains: randomisation process, deviations from intended interventions, missing outcome data, measurement of the outcome and selection of reported results. For crossover trials, an additional assessment for bias arising from carryover and period effects is included. The outcomes are categorised into low, moderate or high risk of bias. As systolic and diastolic blood pressure, blood pressure control rate, AEs, SAEs and all-cause dropout rates are relatively objective indicators, detection bias is the same for all six outcomes; thus, we will assess risk of bias only once.
Quality of the evidence
This network meta-analysis will employ the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) framework, which assesses the evidence quality of all pairwise comparisons and overall rankings within the network across five dimensions: study limitations, imprecision, inconsistency, indirectness and publication bias. The quality of evidence is categorised into four levels: high, moderate, low and very low. The initial evidence level for randomised controlled trials is high, with potential downgrading on the basis of the following criteria:
Study limitations
Based on the Cochrane risk of bias assessment for randomised controlled trials, within a comparison group, studies with low, moderate, and high bias risk were scored 0, −1, and −2, respectively. A composite weighted score is calculated from the proportion of studies at low, moderate and high risk across network comparisons, with rounding to assess for downgrading.
Imprecision
Following the GRADE guidelines,25 imprecision is assessed by examining the optimal information size and whether the 95% CIs for effect differences between the intervention and the control group include significant benefit or harm. If the predicted effect is insufficient for clinical relevance, the quality is downgraded by one level.
Inconsistency
The quality is downgraded by one level in the presence of significant inconsistency (p<0.05) in inconsistency tests. For heterogeneity, evidence quality is downgraded by one level if I² exceeds 50% in comparisons. A maximum of two levels can be downgraded for this domain.
Indirectness
A comprehensive assessment will be conducted based on the differences among participants, interventions, outcome measurements, and the indirectness of evidence comparison. If the assumptions for indirect comparisons are not met or uncertain, evidence quality is downgraded by one level.
Publication bias
Publication bias will be assessed comprehensively through search strategies, funnel plots, Egger’s test results and trim-and-fill methods for each outcome comparison. If significant publication bias is present, the quality is downgraded by one level.
Evaluation of synthesis assumptions
All six outcomes (reduction in systolic/diastolic blood pressure, rate of achieving target blood pressure, and proportion of AEs, SAEs and all-cause dropout after at least 4 weeks of treatment) will be quantitatively synthesised.
Pairwise comparison
The DerSimonian-Laird random-effects model or the Mantel-Haenszel fixed-effects model will be used in traditional pairwise meta-analyses. For continuous outcomes, the weighted mean difference (WMD) and its 95% CI are summarised. For binary outcomes, the OR and its 95% CI are summarised. The I² statistic is calculated to assess for heterogeneity, as a measure of the proportion of overall variation that is attributable to between-study heterogeneity.
Indirect and mixed comparisons
We will employ a random-effects model within the frequentist approach to analyse primary outcomes. The WMD with its 95% CI for continuous outcomes and the OR with its 95% CI for binary outcomes will be used as summarised effects.
For all treatment regimens, we will estimate their probabilities of being in each possible rank for each indicator. Additionally, we will rank all treatment regimens using the surface under the cumulative ranking curve (SUCRA), which represents the probability of being the most effective treatment in the absence of outcome uncertainty, with a score of 1 indicating certainty of being the best treatment and 0 indicating certainty of being the worst treatment.
Assessment of inconsistency, transitivity and heterogeneity
To assess the overall inconsistency of the network, we will employ the design-by-treatment interactions method.26 For local inconsistency, we will use the loop-specific inconsistency test to assess for inconsistency within each closed loop.27 Additionally, we will employ the node-splitting method28 and heat maps to assess the local inconsistency of the model by separating evidence on particular comparisons into direct and indirect evidence.
For transitivity assessment in the network meta-analysis, we will evaluate the comparability of clinical variables across studies that may influence the final outcome.
We will assess the global heterogeneity of the network using I2 statistics. For local heterogeneity, we will employ both I2 statistics and prediction interval plots29 to evaluate the level of uncertainty in local effect estimates. In prediction interval plots, heterogeneity-induced uncertainty is defined as the inconsistency between the CI of the treatment outcome and its prediction interval.
Assessment of study contributions and publication bias
Contribution plots will be used to assess the contribution of each pairwise direct comparison to the mixed comparisons or indirect comparisons in the final estimation of the network model,29 thereby assisting in the evaluation of the overall quality of evidence in the network.
Comparison-adjusted funnel plots will be employed to visualise potential publication bias in small and large study results.27 Additionally, Egger’s test and trim-and-fill analysis will be used for quantitative assessment of publication bias.
Sensitivity and subgroup analyses
A sensitivity analysis will be conducted via a Bayesian framework with a random-effects model.30 Similarly, WMD with its 95% CI for continuous outcomes and OR with its 95% CI for binary outcomes will be used as summarised effects, using SUCRA to rank all treatment regimens. Within the Bayesian framework, an effect model will be constructed using Markov chain Monte Carlo methods, and model convergence will be assessed using trace and density plots, Gelman-Rubin-Brooks plots and the potential scale reduction factor. The number of Markov chains, amount of adaptation iterations and amount of simulation iterations will be adjusted based on convergence. The type of model (consistency or inconsistency) will be determined based on the consistency assessment. Model fit will be evaluated using total residual deviance, which indicates good fit if it approximates the number of data points.
To assess whether the results are impacted by study characteristics (effect modifiers), subgroup analyses will be conducted according to sample size, quality of study and sponsorship, if the data allow.
Statistical analysis
The network meta-analysis will be conducted using R (V.4.3.2), with statistical analysis facilitated by the R packages ‘gemtc’ and ‘netmeta’. Additionally, STATA V.18.0 MP Parallel Edition will also be employed.
Patient and public involvement
None.
Ethics and dissemination
Since this study relies solely on published literature, no ethics approval is necessary. The results will be submitted to a peer-reviewed journal.
Ethics statements
Patient consent for publication
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
XZ and TL are joint first authors.
XZ and TL contributed equally.
Contributors XXZ designed this meta-analysis and together with TL wrote the first draft of the protocol. YNH provided technical guidance for the statistical analysis. WLF and XGL provided suggestions for revising the manuscript. WHJ funded this work and provided critical review of the manuscript. All authors read and approved the final manuscript. WHJ accepts full responsibility for the finished work and the conduct of the study, had access to the data and controlled the decision to publish.
Funding The study will be supported by the Key Research and Development Program of Hunan Province (no: 2022SK2029). The funders had no role in considering the protocol design nor in the writing of the report and in the decision to submit the article for publication.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Author note XGL and WHJ are joint last authors.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.