Efficacy and safety of patent haemostasis-guided strategies for preventing radial artery occlusion following transradial coronary procedures: a protocol for a systematic review, network meta-analysis, and exploratory component network meta-analysis

Abstract

Introduction

Patent hemostasis (PH) is important for preventing radial artery occlusion after transradial coronary procedures, but the relative efficacy and safety of different PH-guided strategies remain uncertain. This study aims to compare and rank these strategies and, if data permit, explore the potential contributions of their key components.

Methods and analysis

Nine databases (PubMed, Embase, CINAHL, the Cochrane Library, Web of Science, China National Knowledge Infrastructure, Wanfang Database, VIP Database and the Chinese Biomedical Literature Database) will be searched from inception to 1 July 2026 for randomised controlled trials involving adults undergoing coronary intervention and diagnosis via transradial access. Supplementary searches will include trial registries, academic search engines and reference lists. Two reviewers will independently conduct study selection, data extraction, network node assignment, component coding and risk-of-bias assessment. Pairwise and frequentist network meta-analyses will be performed using R v4.4.1 under a random-effects framework. Heterogeneity, transitivity, inconsistency, small-study effects, certainty of evidence, subgroup effects and sensitivity will be assessed. Exploratory component network meta-analysis will be conducted using an additive model if data and network structure permit.

Ethics and dissemination

This study will analyse published aggregate data and will not involve participant recruitment, clinical intervention or collection of biological samples; therefore, ethical approval is not required. The findings will be disseminated through peer-reviewed publication and other relevant academic channels.

Trial registration

PROSPERO CRD420261436288.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• Strategy nodes will be defined by procedural characteristics rather than haemostatic device names alone.

• If sufficient data are available, an exploratory component network meta-analysis will assess the potential contribution of individual components of patent haemostasis-guided strategies.

• This protocol is reported in accordance with established guidance for systematic review protocols, and the completed review will be reported in accordance with guidance for network meta-analyses.

• The use of aggregate study-level data may limit more detailed exploration of patient-level effect modification.

• Differences in trial design, haemostasis protocols and outcome assessment may influence the certainty of the evidence.

Introduction

Transradial access (TRA) has become a widely used vascular approach for coronary angiography (CAG) and percutaneous coronary intervention (PCI).^{1 2} Compared with conventional transfemoral access, TRA is associated with a lower risk of access-site bleeding, earlier patient ambulation, improved comfort and more efficient perioperative management.^{3 4} Despite the widespread adoption of TRA in coronary diagnostic and interventional procedures, radial artery occlusion (RAO) remains an important clinical issue that has yet to be fully resolved.⁵ Although frequently asymptomatic, RAO may preclude future use of the radial artery for repeat coronary intervention, coronary artery bypass grafting, arteriovenous access creation and other vascular procedures.⁶

Haemostasis strategy is considered an important modifiable factor affecting the risk of RAO. Although conventional radial artery compression is effective for achieving haemostasis, excessive or prolonged compression may impede radial arterial flow, thereby increasing the risk of local blood stasis, thrombus formation and RAO.⁷ Patent haemostasis (PH) was therefore developed to achieve effective haemostasis while preserving antegrade radial artery flow, with patency confirmed by objective methods such as the modified Barbeau test, pulse oximetry waveforms or Doppler ultrasound.⁸ Previous evidence suggests that PH can reduce the incidence of RAO. Accordingly, it is regarded as a key principle in postprocedural haemostasis management after transradial procedures.⁹

Current clinical practice and research have adopted increasingly diverse approaches to implementing PH. Several modified PH approaches have been developed beyond standard PH-guided compression. These approaches include short-duration compression or rapid-deflation PH strategies, individualised PH protocols based on mean arterial pressure (MAP) or pressure-guided compression, PH combined with ipsilateral ulnar artery compression (IUAC) to maintain radial artery flow and adjunctive PH strategies incorporating haemostatic dressings, including chitosan- or kaolin-based materials.^10–13 Despite differences in implementation, these strategies share the common goal of preserving radial artery patency while achieving effective haemostasis, thereby reducing the risk of RAO and related complications. However, these approaches differ in patency confirmation methods, compression intensity, compression duration, procedural complexity, patient experience and clinical feasibility. Therefore, PH should no longer be viewed merely as a single haemostatic principle, but rather as a group of clinically implemented strategies with distinct procedural characteristics and different risk-benefit profiles. In this context, operational characteristics refer to the specific procedural features that determine how patent haemostasis is achieved and maintained, rather than merely to the name of the haemostatic device used. An individual strategy may incorporate one or more of these features. These features form the basis for classifying strategies in this review, with detailed operational definitions provided in table 1.

Table 1. Operational definitions of eligible PH-guided haemostasis strategies.

Strategy category	Operational definition
Standard PH-guided compression	Compression with objective confirmation or maintenance of radial artery patency, such as by the modified Barbeau test or Doppler ultrasound, without other major strategy modifications.
Short-duration or rapid-deflation PH	PH-guided compression in which the main intervention feature is shortened compression, rapid deflation, or early device removal.
MAP-guided or pressure-guided PH	Individualisation of compression intensity according to MAP or other pressure parameters to achieve PH.
IUAC-facilitated PH	PH-guided compression combined with IUAC.
Haemostatic material-assisted PH	PH-guided compression combined with chitosan, kaolin or other haemostatic materials.

Other eligible PH-guided strategies not covered by these categories will be classified according to their defining operational feature or combination of features. Clinically and methodologically comparable variants will be grouped within the same node, whereas additional nodes will be created only for clinically distinct strategies.

A recent systematic review and meta-analysis found that adjunctive haemostatic agents shortened the time to achieve haemostasis but did not significantly reduce the overall incidence of 24-hour RAO.¹³ However, that review focused on the addition of haemostatic agents rather than comparisons among distinct PH-guided strategies. Another recent systematic review and network meta-analysis has compared radial artery haemostasis devices for the prevention of RAO, offering evidence on the relative performance of different devices, materials and compression methods after transradial procedures.¹⁴ A key limitation of that review is that interventions were primarily categorised according to haemostatic devices or compression duration rather than distinct PH-guided strategies. Because procedural protocols may differ substantially even when the same device is used, device-based classifications may obscure the effects of clinically distinct PH-guided strategies and their key components. Thus, existing evidence remains insufficient to answer a more clinically decision-oriented question: which complete PH-guided haemostasis strategy provides the greatest benefit in preventing RAO while achieving a better overall balance among bleeding risk, haemostasis duration and patient acceptability?

Based on this rationale, the present study aims to reorganise existing evidence from randomised controlled trials (RCTs) from the perspective of PH-guided haemostasis strategies, according to predefined operational definitions. A network meta-analysis (NMA) will be conducted to compare the relative efficacy and safety of conventional compression, standard PH-guided compression and various PH-related modified strategies in preventing RAO and improving related outcomes. In addition, because these PH-guided strategies often consist of several shared key components, an exploratory component network meta-analysis (CNMA) will be further performed if sufficient data are available.¹⁵ This analysis will explore the potential contributions of different components to RAO prevention and safety outcomes. This study is expected to provide more clinically interpretable evidence for the selection and optimisation of PH-guided radial artery compression strategies after TRA and to inform the development of standardised haemostasis protocols and the design of future studies.

Methods

This systematic review protocol is reported in accordance with the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) guidelines.¹⁶ The completed PRISMA-P checklist is provided in online supplemental table S2). The findings of this study will be reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension statement for network meta-analysis (PRISMA-NMA).¹⁷ A schematic overview of the study design and analytical framework of this protocol is shown in figure 1. The study protocol has been registered in the International Prospective Register of Systematic Reviews (PROSPERO), with the registration number CRD420261436288. The overall review process commenced on 29 June 2026 and is expected to be completed by 1 January 2027, consistent with the review timeline registered in PROSPERO. Any subsequent amendments to the protocol will be updated accordingly in the PROSPERO registry.

Figure 1. Overview of the study design and analytical framework. Abbreviations: CAG, coronary angiography; CNMA, component network meta-analysis; d, days; DBP, diastolic blood pressure; h, hours; IUAC, ipsilateral ulnar artery compression; MAP, mean arterial pressure; mm Hg, millimetres of mercury; NMA, network meta-analysis; PCI, percutaneous coronary intervention; SBP, systolic blood pressure; SpO₂, peripheral oxygen saturation; TRA, transradial access.

Eligibility criteria

Inclusion criteria

Eligibility criteria were defined according to the participants, interventions, comparators, outcomes and study design framework, as follows:

• Participants: Adult patients aged ≥18 years who underwent CAG and/or PCI via TRA.

• Interventions: Eligible interventions will include PH-guided post-procedural radial compression or haemostasis strategies. Each study arm will be assigned to a single primary network node based on the core procedural characteristics of the reported strategy rather than solely on the name of the haemostatic device. Eligible PH-guided strategies will include, but are not limited to, the five main categories presented in table 1.

All strategy features will be coded separately for each study arm. In the strategy-level NMA, each arm will be assigned to one primary node using the prespecified operational definitions, based on the feature that constitutes the primary randomised difference and the stated purpose of the intervention; co-occurring features will be retained in the arm-level strategy description. If two or more features are introduced together and no single defining feature can be identified, the combination will be treated as a distinct node when clinically interpretable and compatible with network connectivity; otherwise, the comparison will be summarised narratively.

• Comparators: Comparators may include conventional compression or compression without confirmed patency, standard PH-guided compression or any other eligible strategy. Conventional compression or compression without confirmed patency refers to post-procedural compression without an explicit requirement for objective confirmation or maintenance of radial artery patency. Because all eligible study arms will be treated as network nodes, studies must compare at least two eligible radial compression or haemostasis strategies.

• Outcomes: The primary outcomes of this review will be 24-hour RAO (24 hours RAO) and 30-day RAO (30 d RAO). RAO will be defined according to the criteria used by the individual trialists, generally indicating the postprocedural absence of detectable flow in the accessed radial artery. The specific definition, diagnostic method and assessment timing used in each study will be extracted. Secondary outcomes will include time to achieve haemostasis, the incidence of bleeding events, the incidence of access-site haematoma, pain score and patient comfort.

• Study design: Only RCTs reported in Chinese or English will be eligible.

Exclusion criteria

• Studies conducted exclusively in special populations in whom radial haemostasis, PH implementation or RAO assessment may differ substantially from routine practice, such as patients with severe coagulation disorders, dialysis access-related populations or severe peripheral vascular disease.

• Studies enrolling mixed populations will be excluded if data for eligible patients undergoing transradial coronary procedures cannot be separately extracted.

Search strategy

A comprehensive literature search will be conducted from database inception to July 1, 2026. To maximise the sensitivity of the main database searches, we will use broad and inclusive search terms, as the terminology used to describe different PH-guided haemostasis strategies may vary across titles, abstracts and indexing records. The detailed search strategies are provided in online supplemental table S1.

• Databases: The following English and Chinese electronic databases will be searched: PubMed, Embase, CINAHL, the Cochrane Library, Web of Science, China National Knowledge Infrastructure (CNKI), Wanfang Database, VIP Database and the Chinese Biomedical Literature Database (CBM).

• Registers: ClinicalTrials.gov and the WHO International Clinical Trials Registry Platform (WHO ICTRP) will be searched to identify ongoing, unpublished, or recently completed clinical trials.

• Other methods: Grey literature will be searched using Google Scholar and Baidu Scholar. In addition, the reference lists of included studies and relevant reviews will be manually screened to identify additional eligible studies.

study selection

Two reviewers will independently screen all retrieved records. After duplicates are removed using EndNote X9, the same software will be used for title and abstract screening and subsequent full-text assessment of potentially eligible studies. Disagreements will be resolved through discussion or, if necessary, consultation with a third reviewer. Reasons for exclusion at the full-text stage will be recorded. The study selection process will be presented in a PRISMA flow diagram (figure 2).¹⁸

Figure 2. The PRISMA flow diagram. CBM, Chinese Biomedical Literature Database; CINAHL, Cumulative Index to Nursing and Allied Health Literature; CNKI, China National Knowledge Infrastructure; ICTRP, International Clinical Trials Registry Platform; RCT, randomised controlled trial.

Data extraction

A standardised data extraction form will be developed and pilot-tested before formal data extraction. Two reviewers will independently extract data from each included study and cross-check the results for accuracy and completeness. Network node assignment, grouping of clinically comparable strategy variants, and CNMA component coding will also be performed independently by the two reviewers according to the prespecified operational definitions. Any disagreements will be resolved through discussion or, if necessary, consultation with a third reviewer. The following categories of data will be extracted:

• General study information: first author, year of publication, country or region, study design, number of study arms and sample size.

• Participant characteristics: age, sex, relevant comorbidities, inclusion and exclusion criteria, and baseline radial artery status when reported.

• Procedural characteristics: type of coronary procedure; sheath size; access site; antithrombotic or anticoagulation regimen; and heparin dose.

• Radial compression or haemostasis strategy details: haemostatic device or material used, compression method, timing of sheath removal, initial compression pressure or air volume, pressure adjustment method, compression duration, deflation protocol and device removal time.

• PH-related information: whether PH was objectively confirmed or maintained, the method used to assess radial artery patency, and the timing of patency assessment.

• Strategy classification information: the procedural characteristics required for network node assignment and exploratory CNMA component coding will be extracted according to the prespecified operational definitions.

• Outcome data: For each prespecified outcome and assessment time point, event counts and numbers analysed will be extracted for dichotomous outcomes, and means, SD and numbers analysed for continuous outcomes; other summary statistics will be recorded as reported. Outcome definitions, measurement methods or scales, assessment timing, and the diagnostic method for RAO will also be recorded.

Missing data

For missing, unclear or incomplete data, corresponding authors will be contacted by email for clarification. If the required data cannot be obtained, available data will be used where appropriate. Studies with insufficient data for a specific outcome will be excluded from the corresponding quantitative synthesis but may be included in the narrative synthesis.

Statistical analysis

All statistical analyses will be performed using R software v4.4.1, with the meta, metafor and netmeta packages. Analyses will be conducted separately for each prespecified outcome. Given the anticipated clinical and methodological heterogeneity, a random-effects model will be used as the primary analytical model.

For dichotomous outcomes, risk ratios with 95% CIs will be used, and corresponding absolute effects will also be presented where feasible. For continuous outcomes, mean differences with 95% CIs will be used when measurements are sufficiently comparable, whereas standardised mean differences with 95% CIs will be used when the same underlying construct is assessed using different scales or scoring methods. Units and scoring directions will be standardised where necessary. Other reported summary statistics will be converted to means and SD where appropriate using established methods.^{19 20} Where available, validated context- and instrument-specific minimal important differences identified from published evidence will be used to interpret continuous outcomes. For outcomes without an established minimal important difference, clinical relevance will be assessed by considering the magnitude and precision of effect estimates and the certainty of evidence. Before quantitative synthesis, the comparability of outcome definitions, measurement methods, and assessment timing will be assessed. The 24-hour RAO and 30-day RAO outcomes will be analysed separately. If the exact target time point is unavailable, the reported assessment closest to 24 hours or 30 days, respectively, will be selected, with the actual timing and diagnostic method recorded. RAO outcomes will be pooled only when their definitions, assessment methods, and timing are considered sufficiently comparable; otherwise, they will be analysed separately or summarised narratively. For time to achieve haemostasis, data will be converted to a common time unit where possible, and clinically incompatible definitions will not be pooled. Scale directions for pain and patient comfort will be standardised to ensure consistent interpretation.

For direct comparisons represented by at least two studies, pairwise meta-analyses will be conducted. A frequentist NMA will then be performed to compare all eligible strategies within a connected evidence network. Correlations among comparisons from multi-arm studies will be appropriately accounted for. To avoid excessive node fragmentation, nodes will be defined according to the prespecified strategy categories rather than individual device brands or minor protocol variations. Arms sharing the same defining operational feature or combination of features will be grouped when clinically and methodologically comparable; minor variations will create separate nodes only when they are expected to modify treatment effects. Sparse but connected nodes will be retained when relative effects are estimable and interpreted cautiously, whereas disconnected or non-comparable strategies will be analysed in separate subnetworks or summarised narratively. Network plots will be used to display the evidence structure, and relative treatment effects will be presented in league tables where appropriate. If the network is sufficiently informative, P-scores will be used to rank competing strategies. Rankings will be interpreted together with the magnitude and precision of effect estimates, clinical relevance, and certainty of evidence. When data permit, global inconsistency will be assessed using the design-by-treatment interaction model and local inconsistency will be explored using node-splitting or side-splitting methods where applicable.^21–23 If important inconsistency is identified, potential clinical and methodological sources will be explored. When at least 10 studies are available for a given outcome, comparison-adjusted funnel plots will be used to explore possible small-study effects or publication bias.²⁴ Studies that meet the eligibility criteria but cannot be included in the quantitative synthesis because of insufficient, unusable, or clinically non-comparable data will be summarised narratively.

If sufficient data are available, an exploratory CNMA will be conducted to evaluate the potential contribution of prespecified components of PH-guided haemostasis strategies. For each study arm, the presence or absence of objective PH confirmation, short-duration compression or rapid deflation, pressure-guided compression, IUAC and haemostatic material assistance will be coded according to the reported intervention protocol. Accordingly, an arm containing multiple strategy features will be represented by the complete set of its coded components, independently of its primary node assignment in the strategy-level NMA. The analysis will use an additive model, assuming that the effect of a combined strategy can be decomposed into the contributions of its individual components.²⁵ CNMA will be undertaken only when component definitions are sufficiently clear, the component network is adequately connected, and severe component collinearity is absent; its findings will be interpreted as exploratory.

Risk of bias assessment

Two reviewers will independently assess the risk of bias for each relevant result included in the synthesis using the Cochrane Risk of Bias 2 (RoB 2) tool.²⁶ Assessments will be conducted across the five domains of RoB 2, and the overall risk of bias will be judged as low risk, some concerns or high risk. Disagreements will be resolved through discussion or, if necessary, consultation with a third reviewer.

Assessment of evidence quality

The certainty of evidence for each pairwise meta-analysis result will be assessed using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach,²⁷ whereas confidence in the network meta-analysis findings will be assessed using the Confidence in Network Meta-Analysis (CINeMA) framework across its six domains.²⁸ For both approaches, certainty or confidence will be rated as high, moderate, low or very low. Two reviewers will independently perform these assessments, with disagreements resolved through discussion or consultation with a third reviewer.

Heterogeneity assessment

Clinical and methodological heterogeneity will be assessed by comparing relevant study and intervention characteristics. For direct comparisons including at least two studies, statistical heterogeneity will be assessed using the I² statistic and Cochran’s Q test. In the NMA, between-study heterogeneity will be evaluated using the estimated between-study variance (τ²). If substantial heterogeneity is identified, potential sources will be explored through subgroup or sensitivity analyses where data permit.

Assumption of transitivity

Transitivity will be assessed by comparing the distribution of potential effect modifiers across treatment comparisons.¹⁸ These may include coronary procedure type, radial access site, sheath size, baseline radial artery status, antithrombotic or anticoagulation regimen, heparin dose, and RAO assessment method and timing. Important imbalances will be considered when interpreting indirect comparisons and explored through subgroup or sensitivity analyses where appropriate.

Subgroup analysis

If sufficient data are available and network connectivity is maintained, subgroup NMAs will be conducted to explore potential sources of heterogeneity and the influence of important effect modifiers related to the transitivity assumption. Prespecified subgroup analyses will be based on coronary procedure type (CAG only, PCI only or mixed CAG/PCI populations), RAO assessment method (Doppler-based or non-Doppler assessment), and radial access site (distal or conventional/proximal radial access).

Sensitivity analysis

Sensitivity analyses will be conducted to assess the robustness of the findings by excluding studies at high risk of bias, studies with uncertainty in network node classification, and studies using non-Doppler-based RAO assessment methods. Where feasible and network connectivity is maintained, analyses will also be repeated using alternative clinically plausible node definitions. Fixed-effect and random-effects models will also be compared where appropriate. Robustness will be evaluated according to changes in the direction and magnitude of the effect estimates.

Patient and public involvement

Patients and/or the public were not involved in the design, conduct, reporting or dissemination plans of this systematic review.

Ethics and dissemination

This study will conduct a secondary analysis of published RCT data and will not involve direct participant recruitment, clinical intervention, or collection of biological samples; therefore, formal ethical approval is not required. The original studies included in this review should have obtained appropriate ethical approval and informed consent. The findings will be disseminated through academic publication and other relevant channels.

Discussion

This protocol conceptualises PH as a complex intervention framework defined by a common therapeutic goal and multiple combinable procedural components, and accordingly reorganises the existing evidence. Existing evidence will therefore be synthesised at both the strategy and component levels. To our knowledge, this will be the first systematic review and network meta-analysis to define network nodes primarily according to core procedural characteristics rather than haemostatic device names and to compare complete PH-guided haemostasis strategies. If the available evidence permits, an exploratory component network meta-analysis will also be conducted. These two approaches will provide distinct but complementary clinical information: the NMA will compare complete haemostasis strategies that can be directly implemented in clinical practice, whereas the CNMA will explore the potential contributions of key components shared across strategies. The review is expected to provide more clinically interpretable evidence on the relative efficacy and safety of PH-guided strategies and to identify priorities for future research; however, its conclusions will depend on the amount, connectivity and certainty of the available evidence.

Several limitations are anticipated. First, the use of aggregate study-level data may limit our ability to adequately assess whether individual patient characteristics modify the effects of different strategies. Second, PH is not a fully standardised intervention, and variations in procedural implementation across studies may challenge the transitivity assumption and the consistency between direct and indirect evidence. Third, some novel or complex strategies may be supported by only a limited number of studies, potentially reducing the robustness of the findings. Finally, the exploratory CNMA is based on an additive model, and its interpretation depends on clearly defined components, sufficient network connectivity, and the absence of substantial collinearity among components.

Data availability statement

Data sharing not applicable as no datasets generated and/or analysed for this study.