**3. Results**

A total of 2830 EMS professionals undertook the training programme, and 69.5% completed the baseline questionnaire while 53.2% answered the three-month follow-up (Figure 1). The majority were males (76.7%) with a mean age of 35.8 years (SD = 6.3) and more than 10 years of experience in EMS (45.4%). Emergency medical technicians accounted for 90.3% of the staff, followed by nurses (6.4%) and then physicians (3.2%). Most of them worked as care providers (98.3%) and only 1.7% did so in the dispatch center. Meanwhile, 65% had received previous training on strokes. The satisfaction survey (Kirkpatrick level 1) at the end of the training was completed by 2668 (94.3%), scoring at least 4 out of 5 in all items.

**Figure 1.** Flowchart of emergency medical services (EMS) professionals' participation.

#### *3.1. Learning Assessment (Kirkpatrick Level 2)*

After the OTI, there was a significant increase in 80% of the questions related to the recognition of the signs and symptoms of a stroke (Table 1), especially for those addressing location-specific signs, as well as the treatment. Two generic questions about strokes were non-significant but scored high prior to the training. The questions regarding SCs, prehospital management, and prehospital stroke assessment increased significantly, with most scores above 85%. Changes were observed in stroke recognition, with improved identification in 5/6 of the clinical scenarios.



1 The smile, raise the arm, talk, stroke, call 911 fast mnemotechnic (i.e., RAPID) is a Catalan stroke assessment tool, equivalent to the FAST mnemotechnic in English-speaking countries. 2 Rankin Comorbidity (i.e., RANCOM) is a Catalan comorbidity scale for prehospital assessment of stroke patients. \* *p* < 0.05.

#### *3.2. Transfer to the Clinical Setting (Kirkpatrick Level 3)*

The Minimum Data Set (MDS) records available in the SITREM ® register were analyzed for 17,135 patients in the study period (Table 2). Immediately after the training, we found a significant increase in the registration of patient identification (ID) codes, glycaemia, systolic and diastolic blood pressure (SBP/DBP), and RACE scores, ye<sup>t</sup> notification of the time of the onset of symptoms decreased. For the 2015–2016 follow-up, only the patient ID and RACE records continued to increase, in contrast to the time of symptom onset (TSO), SBP and DBP. There were significant di fferences for all items in the 2017–2018 follow-up period: All items increased (i.e., TSO, glycaemia, ID, SBP, DBP, and RACE notification), and the frequency of anticoagulant treatment notification decreased. Overall, 71.5% of the items increased from baseline to the last follow-up. Only the registration of TSO and the notification of anticoagulant treatment diminished consistently over time.

Compliance with the RACE scale upon SC activation increased continuously over time. Starting from 60.9% immediately after training, compliance rose to over 85% by the 2017–2018 follow-up, reaching a 96.3% compliance level in 2018.



#### *3.3. Impact on Prehospital Stroke Code (Kirkpatrick Level 4)*

#### 3.3.1. Stroke Code Activation

Activation of SCs increased over time. At baseline (2014 Q1), 9.2 codes were activated daily (n = 834), which increased immediately after training (2014 Q4) to 10.7 (n = 965), maintaining at 9.9 in both 2015 (n = 2635) and 2016 (n = 3635). In 2017, 10.6 codes were initiated (n = 3888), which reached a daily maximum of 11.4 (n = 4187) in 2018.

#### 3.3.2. Time Performance in Stroke Code

The main differences in time performance were the overall time of prehospital care (Table 3), which increased from 48.9 to 53.6 min (*p* = 0.015). This extra time was mostly due to increased on-site clinical care time (*p* = 0.034) prior to transfer to hospital, which increased from 21.5 to 24.3 min. However, there were no changes in activation, alert, response, or transfer times.


**Table 3.** Differences in time performance over time.

1 Alert time: The time between the start and the end of a call at the dispatch center. 2 Activation time: The time from the start of a call to the allocation of clinical resources at the dispatch center. 3 Response time: The period from the resource allocation to the arrival of the EMS team at point of care. 4 Clinical care time: The time from the arrival of the EMS team at the point of care to the start of transfer; on-site care is provided. 5 Transfer time: The time of transportation from the point of care to arrival at the receiving center. 6 Overall time: The sum of all previous times. CI, confidence interval; SD standard deviation. \* *p* < 0.05.
