PDF: GWTG-IHS-MT
(2026)Objective
To evaluate presentation, quality metrics, treatment utilization, and outcomes of in-hospital stroke (IHS) compared with community-onset stroke (COS) in the mechanical thrombectomy era (2016-2023) using the Get With The Guidelines-Stroke registry.
Study Summary
• Adjusted in-hospital mortality was more than double (aOR 2.27, 95% CI 2.18-2.36) with lower odds of discharge home (aOR 0.46) and independent ambulation (aOR 0.52)
• MT rates more than doubled in both groups over the study period (IHS 4.47% to 9.54%; COS 2.35% to 5.55%) and defect-free care converged across IHS (88.4%) and COS (86.0%)
Intervention
Observational comparison of in-hospital stroke vs community-onset stroke cohorts (no interventional arm); evaluated process measures, reperfusion treatment rates (IV tPA and mechanical thrombectomy), and clinical outcomes.
Inclusion Criteria
Adults aged 18 to over 90 years with a final diagnosis of ischemic stroke recorded in the GWTG-Stroke registry between January 2016 and December 2023, from US hospitals contributing at least one IHS case during the study period.
Study Design
Arms: In-Hospital Stroke (IHS, symptom recognition after hospital arrival) vs Community-Onset Stroke (COS, symptom recognition before arrival)
Patients per Arm: IHS: 191,355 (3.8%); COS: 4,805,037 (96.2%); Total: 4,996,392 from 2,542 US hospitals
Outcome
• Lower discharge home (aOR 0.46, 95% CI 0.45-0.48) and lower independent ambulation at discharge (aOR 0.52, 95% CI 0.50-0.53)
• Recognition-to-CT longer in IHS (median 51 vs 18 minutes, P<0.001)
• IV tPA use lower in IHS (6.46% vs 12.02%) but recognition/door-to-needle <=60 min comparable (82.2% vs 84.5%)
• MT rates rose significantly in both groups (P<0.001 Cochran-Armitage); IHS had 80% higher odds of catheter-based treatment vs COS (OR 1.80, 95% CI 1.77-1.83)
• Compared with 2006-2012 IHS cohort, defect-free care improved from 60.8% to 88.4% and recognition-to-needle <=60 min from 19.7% to 82.2%
Clinical Question
In the mechanical thrombectomy era (2016-2023), how do in-hospital stroke (IHS) and community-onset stroke (COS) compare with regard to clinical presentation, GWTG-Stroke quality metrics, reperfusion treatment utilization, and in-hospital outcomes?
Bottom Line
In the MT era, in-hospital stroke continues to present with greater neurological severity, longer recognition-to-imaging times, and substantially worse in-hospital outcomes than community-onset stroke (mortality more than doubled after adjustment), despite marked improvements in guideline-adherent care and a rising MT rate that is consistently higher in IHS than COS.
Major Points
- Large contemporary retrospective cohort from the AHA Get With The Guidelines-Stroke registry, 2016-2023, including 4,996,392 ischemic stroke admissions from 2,542 US hospitals.
- IHS comprised 3.8% (n=191,355) of admissions; COS comprised 96.2% (n=4,805,037).
- IHS patients were sicker at baseline with higher cardioembolic etiology (33.1% vs 23.7%), atrial fibrillation (20.7% vs 16.6%), CAD/MI (26.6% vs 20.2%), HF (14.3% vs 9.0%), and worse premorbid function (independent ambulation 85.3% vs 91.4%).
- IHS presented with greater severity: NIHSS >20 in 14.5% vs 7.9%; NIHSS 0-4 in 43.3% vs 60.4%.
- Recognition-to-CT time was markedly longer for IHS (median 51 vs 18 minutes, P<0.001).
- IV tPA use was lower in IHS (6.46% vs 12.02%), but among treated patients, recognition/door-to-needle <=60 min was similar (82.2% vs 84.5%).
- MT/catheter-based treatment increased over time in both groups and was consistently higher in IHS (4.47% to 9.54%) than COS (2.35% to 5.55%); each calendar year was associated with an 11% increase in odds of receiving catheter-based treatment (OR 1.11, 95% CI 1.11-1.12).
- Adjusted in-hospital mortality was significantly higher for IHS (aOR 2.27, 95% CI 2.18-2.36); discharge home (aOR 0.46) and independent ambulation at discharge (aOR 0.52) were significantly lower.
- Compared with the pre-MT era IHS cohort (Cumbler et al, 2006-2012), defect-free care improved from 60.8% to 88.4%, recognition-to-needle <=60 min from 19.7% to 82.2%, intensive statin therapy from 20.8% to 86.6%, and dysphagia screening from 65.6% to 85.3%.
- Defect-free care converged between IHS (88.4%) and COS (86.0%) in the contemporary era, with composite achievement exceeding 95% in both.
- Persistent gaps in IHS care include early antiplatelet administration (85.2% vs 92.9%) and antithrombotics on discharge (71.6% vs 81.3%).
Study Design
- Study Type
- Retrospective multicenter cohort study using prospectively collected registry data (GWTG-Stroke)
- Randomization
- No
- Blinding
- Not applicable (observational registry analysis)
- Sample Size
- 4996392
- Follow-up
- Index hospitalization only (no long-term follow-up captured)
- Centers
- 2542
- Countries
- United States
Primary Outcome
Definition: In-hospital mortality, discharge to home, and independent ambulation at discharge (composite primary outcomes assessed individually)
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| COS - lower mortality, higher discharge home, higher independent ambulation (reference) | IHS - higher mortality, lower discharge home, lower independent ambulation | - (Mortality aOR 2.27 (2.18-2.36); Discharge home aOR 0.46 (0.45-0.48); Independent ambulation aOR 0.52 (0.50-0.53)) | <0.001 for all three outcomes |
Limitations & Criticisms
- Observational registry design - residual unmeasured confounding and treatment selection bias cannot be excluded.
- Voluntary nature of GWTG-Stroke participation limits generalizability and may bias toward higher-performing centers.
- Potential under-ascertainment or misclassification of IHS events due to reliance on registry documentation.
- Substantial missing data for NIHSS and several time-based metrics (denominators for recognition/door-to-needle <=60 min were limited to patients with complete timestamps).
- Registry extract did not permit consistent quantification of transfer-in admissions, precluding transfer-restricted sensitivity analyses.
- No long-term functional outcomes captured (e.g., 90-day mRS); outcomes limited to index hospitalization.
- Era-based comparison with Cumbler et al (2006-2012) is descriptive only and subject to evolving registry definitions, abstraction practices, and changing care standards over time.
- Eligible denominators for quality metrics varied between cohorts and eras, complicating direct comparisons (particularly for IVT and DVT prophylaxis measures).
- Analytic window ended December 31, 2023; later data were excluded due to incomplete abstraction at project initiation.
- MT-specific outcomes (recanalization, 90-day functional independence) were not directly evaluated for IHS vs COS in this cohort.
- Multiple comparisons across many baseline and process measures with very large sample sizes - statistical significance does not necessarily imply clinical significance.
Citation
Nouh A, Olson D, Liang H, Koriesh A, Ali M, Bashir MMI, Mukerjee P, Man S, Fonarow GC. In-Hospital Stroke Quality and Outcomes in the Mechanical Thrombectomy Era: Get With The Guidelines-Stroke, 2016 to 2023. Stroke. 2026;57:00-00.