Background The Stroke 90 Project was implemented to reduce delays to stroke thrombolysis and involved 7 hospitals and 2 ambulance services in the Avon, Gloucester, Wiltshire and Somerset regional network. Interventions included a direct to CT (DtoCT) protocol for paramedics to transport patients directly to the CT scanner. Coincidentally, there were severe winter pressures on all participating emergency departments during this period.

Methods Comparison of data from 2 groups across all 7 hospitals: preintervention (n=136) and postintervention patients (n=215) thrombolysed from August 2012 to January 2013. The χ² test, t tests, multiple and linear regression were used for analysis.

Results Ambulance transport times were 56.8 min for preintervention versus 57.5 min for postintervention patients (p=0.78). 11.7% of preintervention patients received thrombolysis within 90 min of call for help versus 23.7% of postintervention cases (p=0.0135). 44% of postintervention patients entered the DtoCT pathway and achieved a mean reduction in door to CT time of 17 min (95% CI 11.5 to 21.5; p<0.0001) and a 19 min reduction in door to needle time (95% CI 10.8 to 26.8; p<0.0001). CT to needle times were 43.8 min preintervention and 42.1 min postintervention (p=0.57).

Conclusions The DtoCT pathway was successful in reducing delays to thrombolysis and should be implemented routinely. The call to door and CT to needle times were not improved by our interventions and further work is required to streamline these. Factors beyond the control of most hospitals may play a role in delaying treatment, but local changes can be implemented to mitigate this.