PDF: RACECAT
(2022)Objective
Whether direct transport of stroke patients with suspected LVO to a comprehensive stroke center (CSC) improves outcomes compared with transport to the nearest primary stroke center in nonurban areas.
Study Summary
• Direct transfer strategy delayed initiation of IV thrombolysis in eligible patients.
Intervention
Cluster-randomized, crossover trial conducted in Catalonia, Spain, from March 2017 to June 2020. Emergency medical service (EMS) sectors alternated every 7 months between two prehospital triage strategies for suspected acute ischemic stroke with LVO within 4.5 hours of onset: (1) direct transport to CSC bypassing PSC, or (2) transport to the nearest PSC with secondary transfer if needed. Primary endpoint: 90-day modified Rankin Scale (mRS) score distribution.
Inclusion Criteria
Adults with suspected acute ischemic stroke, symptom onset within 4.5 hours, prehospital RACE scale ≥5, and located in nonurban areas >30 min from a CSC.
Study Design
Arms: Direct to CSC vs. Nearest PSC with transfer as needed.
Patients per Arm: Direct CSC: 613; PSC: 617.
Outcome
• mRS 0–2 at 90 days: Direct CSC 43.0% vs PSC 42.6%.
• Mortality at 90 days: Direct CSC 17.8% vs PSC 16.4% (NS).
• IV thrombolysis use: Lower in direct CSC (43.7%) vs PSC (52.3%); P=0.002.
• Time from onset to IVT: Longer in direct CSC (median 167 min) vs PSC (median 122 min).
• EVT rates: Direct CSC 49.8% vs PSC 40.5%; P<0.001.
• Symptomatic ICH: Similar between groups.
Bottom Line
Direct transport did not improve 90-day outcomes vs local stroke center transport (adjusted OR 1.03; 95% CI 0.82-1.29). Direct transport increased thrombectomy rates (48.8% vs 39.4%) but reduced IV tPA rates (47.5% vs 60.4%), suggesting these effects offset each other. Stopped early for futility.
Major Points
- No difference in 90-day mRS: adjusted OR 1.03 (0.82-1.29). Stopped for futility at 2nd interim analysis.
- Direct transport increased thrombectomy: 48.8% vs 39.4% (OR 1.46; 1.13-1.89).
- But reduced IV tPA: 47.5% vs 60.4% (OR 0.59; 0.45-0.76).
- Time trade-off: tPA delayed by 35 min (155 vs 120 min) but groin puncture shortened by 56 min (214 vs 270 min).
- Mortality identical: 27.3% vs 27.2% (HR 0.96; 0.78-1.18).
- 64.6% of local stroke center patients were emergently transferred for thrombectomy.
- First RCT comparing mothership vs drip-and-ship in suspected LVO.
- 1,401 randomized (cluster, spatial-temporal); Catalonia, Spain; 6 thrombectomy + 22 local centers.
- No subgroup benefited from direct transport (age, sex, tPA eligibility, RACE score — all NS interaction).
- Results reflect a high-performance Catalan system — may not generalize to systems with longer transfer times or lower baseline thrombectomy rates.
Study Design
- Study Type
- Multicenter, population-based, spatial-temporal cluster-randomized trial with blinded endpoint assessment
- Randomization
- Yes
- Blinding
- Open-label; blinded 90-day mRS assessment by certified telephone assessors. Cluster randomized by 12-hour time slots, stratified by territory and day of week.
- Sample Size
- 1401
- Follow-up
- 90 days
- Centers
- 28
- Countries
- Spain (Catalonia)
Primary Outcome
Definition: Distribution of mRS at 90 days (ordinal shift, mRS 5+6 collapsed)
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| mRS median 3 (IQR 2-5) | mRS median 3 (IQR 2-5) | - (0.82-1.29) | NS (futility stopped) |
Limitations & Criticisms
- Results apply only to nonurban patients in a high-performance Catalan system with efficient drip-and-ship.
- 44% of local stroke center patients had no vessel imaging at first hospital.
- Stopped early for futility — may have been underpowered for subgroups.
- Cluster randomization with inadvertent mismatch in first 350 patients.
- Cannot establish transport time thresholds that might favor direct transport.
- Workflow reflects European model with mandatory neurology contact — may not replicate elsewhere.
Citation
JAMA. 2022;327(18):1782-1794.