Back to 2025 Abstracts
Stroke/Death Ratio As A Measure To Gauge Reporting Of Outcomes Following Carotid Revascularization
Yoel Solomon, MD
1, Filippo Mantovani
1,
Sai Divya Yadavalli, MD1, Andrew Sanders, MD
1, Sara Allievi, MD
2, Christina L. Marcaccio, MD
1, Kirstein D. Dansey, MD
3, Kristina A. Giles, MD
4, Gert J. de Borst, MD
5, Marc L. Schermerhorn, MD
1.
1Department of Surgery, Division of Vascular and Endovascular Surgery, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA,
2Department of Vascular Surgery, Fondazione IRCCS Cà Granda Ospedale Maggiore Policlinico, Milan, Italy,
3Department of Surgery, Division of Vascular Surgery, University of Washington, Seattle, WA, USA,
4Department of Vascular Surgery, Maine Medical Center, Portland, ME, USA,
5Department of Vascular Surgery, University Medical Center, Utrecht, Netherlands.
OBJECTIVES: We hypothesized that strokes following carotid revascularization may be underdiagnosed or underreported in registry and administrative data. To evaluate this, we compared the ratios of strokes to deaths across different data sources—registry, administrative data, and retrospective/non-randomized trials—with those reported in randomized controlled trials (RCTs).
METHODS: We identified patients undergoing CEA and tfCAS for carotid stenosis between 2011 and 2022 from the Vascular Quality Initiative (VQI), National Surgical Quality Improvement Program (NSQIP), and the National Inpatient Sample (NIS). We did a scoping review using Medline, Embase, Cochrane, and Web of Science databases for RCTs, other national-level registry studies, and non-randomized/retrospective studies (>100 patients in each arm) reporting outcomes for CEA and/or tfCAS. We extracted rates of any stroke and death regardless of symptoms status occurring either during hospitalization or within 30 days post-procedure and calculated the ratio of stroke events per death (stroke/death ratio) in each study. We then estimated average stroke/death ratios for each data type.
RESULTS: Stroke/death ratios following both CEA and CAS were lower in VQI, NSQIP and NIS (in-hospital: CEA: 3.7, 4.7, 2.7; CAS: 1.4, 2.1, 0.8) and (30-day: VQI and NSQIP; CEA: 2.4 and 3.5; CAS: 1.3 and 2.9) when compared with ratios from RCTs for intention to treat outcomes (CEA: 1.5-7.3; [overall 4.6] CAS: 2.6-30 [7.3]) or per protocol outcomes (CEA: 1.0-10 [3.1]; CAS:3.0-20 [7.4]). Lower ratios were also observed in other larger registries (CEA: 0.8-5.4 [2.6]; CAS: 0.9-12 [2.1]) and non-randomized/retrospective studies (CEA: 1.0-12 [2.7]; CAS: 0.5-16 [3.4]).
CONCLUSIONS: We found that stroke/death ratios were lower in registry, administrative, or retrospective data compared with RCTs. These differences may be attributed to distinctions in patient characteristics between RCTs and real-world populations, or to factors such as inconsistent postoperative stroke evaluations by neurologists, inaccuracies in self-reporting, or coding errors. Given that deaths are well-captured across various data sources, the stroke/death ratio could serve as a new metric for assessment of the thorough capture of stroke events.
Back to 2025 Abstracts
