There are a number of research questions generated by the PASTA trial because of the unexpected combination of primary and secondary outcomes:
- To explain the cost-effectiveness results, we have proposed a novel mechanism whereby clinicians used additional pre-hospital information to moderate their treatment threshold during difficult decisions, but proving this would require a different study design, confirming guideline-compliant treatment for individual patients.
- We also considered whether or not the paramedic’s pre-departure checklist may have generally reinforced adherence to acute care guidelines. This approach has been adopted in other clinical settings and could be further assessed as an additional quality improvement tool for stroke and other emergency conditions.
- Future studies could consider the value of immediate and structured paramedic feedback to improve performance of specific actions during pre-hospital assessment.
- A wider aim would be to understand the reasons why a significant proportion of paramedics randomised to the intervention group did not engage with study training.
As more information becomes available from trials and audit reports about how IAT effectiveness and efficiency varies between patient groups and service configurations, the WP2 DES/ITEMS model can be further improved by:
- retrospective validation of the model output after a planned service reconfiguration, which would help to define limits of uncertainty for cost-effectiveness
- building in ambulance parameters that reflect resource availability and a potential redirection bias towards CSCs
- feedback from stakeholders using the ITEMS interface to understand barriers to and facilitators of wider implementation
- adding new parameters that might have a significant impact on pathways, such as the introduction of a new point-of-care diagnostic for LAO.
The programme outputs reflect the multilevel interactions between health-care monitoring systems, service configurations, clinician behaviours and evidence for specific treatments. These require mapping and parallel evaluation to maximise our understanding of the real-world consequences following introduction of complex interventions. It will be crucial that all future research in this area and any resulting care recommendations consider the whole emergency stroke pathway and the wider population of suspected stroke patients, so that specific gains do not inadvertently lead to a clinical or economic disadvantage that is unacceptable to services or patients.