Abstract 96: Independent Adjudication Of Get With The Guidelines Thrombectomy Imaging And Angiography Data Reveals Major Discrepancies

Abstract

Introduction: The Get with the Guidelines - Stroke (GWTG-Stroke) registry has been collecting endovascular thrombectomy (EVT) data on acute stroke interventions since 2015. The key variables associated with EVT, solely based on local site documentation, have never been independently adjudicated. We conducted a detailed analysis of single center EVT data as entered in the GWTG-Stroke registry. Methods: Consecutive EVT cases entered into both GWTG-Stroke and an independent research database at a large academic comprehensive stroke center were sampled from 2020-2022. For each case, the following EMR and PACS imaging variables related to EVT efficacy and safety outcomes were compared between GWTG-Stroke and core lab independent readings, including: site of target occlusion (STO), first-pass time (FPT), time mTICI >=2b50 first documented (reperfusion time), final mTICI score (0, 1, 2a, 2b50, 3; FTICI), presence of post-intervention hemorrhagic transformation (HT), and subtype/extent of HT. Results: The GWTG-Stroke registry EVT-imaging data variables document only 6/30 (20%) of the common data elements (CDEs) recommended by NINDS and 6/34 (18%) recommended by the FDA. Of the 80 cases sampled, 29 (36%) had discrepancies between GWTG-Stroke recorded data and independent core lab findings. In 4 cases (5%), reperfusion time was incorrect. In another 4 cases (5%), FTICI was incorrect, even when using the gross 2b50/3 categories. In 1 case (1.3%) STO was incorrect. In 2 cases (2.5%) patient data was not reported to GWTG-Stroke. In 21 cases (26%), HT was documented as not present, when in fact it was. Of those, 3 cases (4%) were PH2, while 18 cases (23%) were IPH of PH1 or less, SAH or SDH. Conclusions: Detailed analysis of the GWTG-Stroke registry on EVT for acute ischemic stroke reveal major discrepancies in numerous variables. In addition, the majority of variables recommended by NINDS and FDA for routine collection in thrombectomy procedures are not captured in GWTG-Stroke. Even the minority of recommended thrombectomy CDEs currently captured in GWTG-Stroke further contain subject level discrepancies in imaging and angiography outcomes when centrally adjudicated.