Document resource
Background Drug-coated balloons (DCBs) are an established treatment for in-stent restenosis and are increasingly used in selected cases of de novo coronary lesions to deliver antiproliferative therapy without permanent stenting. While clinical trials and meta-analyses support their efficacy and safety, patterns of real-world institutional adoption remain uncertain. The latest UK national audit data enable assessment of whether DCB utilisation trends reflect clinical indication alone or are also influenced by institutional procedural exposure and practice patterns.Methods A cross-sectional hospital-level analysis was performed using the 2020 Annual Summary Report from the National Institute for Cardiovascular Outcomes Research (NICOR) National Audit of Percutaneous Coronary Interventions (NAPCI). Hospitals reporting percutaneous coronary intervention (PCI) activity in three DCB-relevant contexts were included: no-stent PCI, PCI for restenosis, and PCI with stent deployment. DCB utilisation was defined as the proportion of procedures using DCBs within each context. National utilisation was estimated using indication-specific volume-weighted means with 95% confidence intervals (CIs) derived by bootstrap resampling. Associations between institutional procedural volume and DCB use were assessed using log-transformed beta regression. Consistency of DCB use across indications was examined using correlation analyses adjusted for procedural volume, with additional sensitivity analyses excluding low-volume centres for robustness.Results Across the national cohort, DCB utilisation varied by clinical context, with volume-weighted use highest in restenosis PCI (49.7%, 95% CI 43.3–55.3) and no-stent PCI (46.2%, 95% CI 40.3–51.8), and substantially lower in PCI involving stent deployment (10.9%, 95% CI 9.38–12.5). After adjustment for procedural exposure, greater institutional volume within DCB-eligible PCI contexts was independently associated with higher DCB utilisation across all indications, with the strongest association observed in restenosis PCI (log-volume coefficient 0.395, p<0.001), and more modest but significant associations in no-stent PCI (0.217, p=0.043) and PCI with stent deployment (0.311, p<0.001). At the hospital level, DCB use demonstrated strong consistency across indications, with a high correlation between no-stent and restenosis PCI (r=0.81, p<0.001) and moderate correlations between stented and non-stented contexts (r=0.50–0.53, all p<0.001); these relationships persisted after volume adjustment and strengthened in higher-volume centres. DCB use was moderately clustered by procedural volume, with approximately one-third of DCB-eligible procedures performed by the top decile of hospitals.Conclusions DCB utilisation in UK PCI broadly reflects current evidence, with frequent use in restenosis and no-stent procedures and more selective use alongside stent deployment. Beyond indication, institutional procedural exposure is consistently associated with higher DCB use, and stable cross-indication practice patterns suggest that institutional strategy contributes to the implementation of this technology. These findings highlight systematic variation in DCB adoption and provide a foundation for future work linking institutional practice to patient-level outcomes.