O-020 Recan or rescan? Validation of the halt score and refinement to HALT+
Bibliographic record
Abstract
Background Early recanalization (ER) prior to endovascular thrombectomy (EVT) for acute large vessel ischemic strokes can significantly impact patient management during acute stroke triage. However, there are no widely adopted diagnostic guidelines on the optimal approach for evaluating ER. The HALT score was previously proposed as a clinical tool to predict ER prior to EVT, however, it did not include clinical exam change, which is thought to be associated with early recanalization. Methods Consecutive stroke patients referred to a Comprehensive Stroke Center for EVT evaluation from 2019 to 2023 were retrospectively identified. Patients who were unable to achieve arteriotomy within 1.5 hours of stroke diagnosis were included. Patient demographics, clinical information, and neuroimaging findings were recorded. ER was defined as recanalization of the initial vascular occlusion seen on either CT or catheter angiography that resulted in a decision not to perform the EVT procedure. The previously proposed HALT (hyperlipidemia, atrial fibrillation, location of vascular occlusion, and thrombolysis) score was validated using receiver-operating characteristic (ROC) curve analyses. We then optimized score value assignments for the location of vascular occlusion to include distal vessels, and we added NIH stroke scale improvement to create the HALT+ score. Finally, we assessed the performance of the HALT+ score using ROC curve analyses. Results 343 patients were included. The original HALT score was a robust predictive tool for ER (c-statistic 0.82 [95%CI 0.77 to 0.87], p<0.001). Only 2 in 138 patients (1.4%) with low HALT scores (0–2) experienced ER, compared to 19 of 53 patients (35.8%) with high HALT scores (6–8). After optimizing the weighting of scores associated with the location of vascular occlusions and including NIHSS improvement (6 points or more) as a predictor variable, the HALT+ score was created (figure 1A). It achieved excellent predictive ability (c-statistic 0.91 [95%CI 0.87 to 0.95], p<0.001; +0.09 compared to the original HALT score, p<0.001; figure 1B). Only 1 in 173 patients (0.6%) with low HALT+ scores (0–3) experienced ER, compared to 20 of 26 patients (76.9%) with high HALT+ scores (8 or higher). Conclusions This external validation study confirmed the ability of the HALT score for identifying stroke patients unlikely to achieve ER prior to EVT. Created with minor modifications and additional inclusion of NIH stroke scale improvement, the HALT+ score was significantly superior to HALT as a robust tool to reliably identify patients both unlikely and likely to experience ER prior to EVT. Disclosures S. Rewinkel: None. S. Amin: None. H. Chen: None. D. Lockwood: None. D. Kim: None. R. Priest: None. G. Nesbit: None. J. Liu: None. M. Horikawa: None. W. Clark: None. R. Laursen: None. D. Gandhi: 1; C; Focused Ultrasound Foundation, InSightec, LTD, MicroVention, Inc., NIH Clinical Center, University of Calgary. M. Colasurdo: None.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.006 | 0.014 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".