With an annual incidence of 299.4 per 100 000 residents in Canada, strokes represent a growing burden on our healthcare system associated with our population growth and ageing [1]. In 1993, Dr Gomez introduced the “time is brain” concept to emphasize that timely stroke recognition by witnesses is crucial to provide function and life-saving acute interventions such as thrombectomy and thrombolysis to a higher proportion of eligible patients [2]. Since 2014, the Heart and Stroke Foundation of Canada (HSFC) conducted numerous FAST (Facial drooping, Arm weakness, Speech difficulties and Time) stroke awareness campaigns across Canada. Unfortunately, according to surveys conducted by the HSFC two months after the beginning of the first campaign, an alarming proportion (43.5%) of Canadians were still unable to recognize facial drooping, arm weakness or speech difficulties as signs of stroke [3]. More recently, in 2022, Rioux and colleagues reported that the three public awareness campaigns conducted in 2016–2019 resulted in a 26% improvement in identification of any FAST stroke sign, with a lower performance in men, retired individuals and those with a lower socioeconomic status [4]. This 26% improvement in early recognition by patients is motivating, but does it translate in a better identification of strokes eligible to thrombolysis or thrombectomy? On one hand, we could hypothesize that a better and faster recognition of stroke symptoms could lead to an optimal treatment of stroke. On the other hand, more stroke mimics transported by paramedics could increase inappropriate visits in already overcrowded emergency departments.

In this issue, Brissette and colleagues designed a protocol with the aim to address this very relevant question[5]. They assessed the impact of the five FAST bilingual campaigns held in 2015–2019 on calls to the emergency medical services (EMS) of Laval and Montreal (Quebec, Canada) for any stroke symptoms within the first 5 h and with a Cincinnati Prehospital Stroke Scale (CPSS) 3/3 performed by a qualified paramedic. After five campaigns conducted by the HSFC, for an average of 9 weeks each, calls for any symptoms of stroke increased by 28% (p < 0.001) compared to 10.1% for headaches (p = 0.012), and calls for symptoms’ onset < 5 h increased by 61% (p < 0.001). The amount of calls for any symptoms of stroke plateaued after three campaigns and no individual campaign resulted in an increase in calls for symptoms’ onset < 5 h nor CPSS 3/3. 61% of “faster calls” is an impressive improvement, but it represents an average of 1.2 additional daily calls, a small number for an EMS agency covering 2.5 million inhabitants. Among the 16.1 daily calls for any suspected stroke, the median proportion of calls for stroke < 5 h was 15.4%. Even after five campaigns, despite important clinical symptoms, patients are still delaying to call 911. Of course, this remains hypothetical as we have no data on the final diagnoses, management, and outcomes of these patients.

Time is critical to enable rapid access to stroke reperfusion therapy. Despite being controversial and subjected to change with the growing literature on this subject, the authors used a 5 h cut-off. This could be a limitation, but as the authors explained, this is the provincial cut-off used by the EMS dispatchers to classify stroke as acute or recent. Also, the severity of the symptoms was not included in the collected data set, except the CPSS 3/3 which corresponds to a National Institutes of Health Stroke Scale (NIHSS) of 3–12. This information would have been of interest, knowing that thrombectomy can result in significant improvement in neurological symptoms events even up to 24 h after onset depending on the mismatch of the clinical deficits and the infarct volume on perfusion computed tomography [6].

Are public health campaigns the key to improve these statistics? As the authors showed us, a single campaign is likely insufficient to induce large scale behavior change in stroke recognition. However, multiple campaigns are likely associated with increased stroke-related calls to EMS. Data from England, Australia and New Zealand also point toward similar conclusions: in general, the FAST public awareness campaigns led to an increase in ambulance dispatch for stroke suspicion, a rise in stroke symptom recognition and even an impact on thrombolysis activity [7,8,9,10]. Furthermore, a systematic review conducted on various health campaigns also concluded that they produce a consistent advantage in behavior change [11, 12].

But how can we improve? In their analysis of repeated surveys, Rioux and colleagues concluded that future FAST campaigns should especially target men, retired people, and individuals with a lower socioeconomic status [4]. This is no different than what others have concluded before, socioeconomic status and geography are significant barriers to equitable access to stroke care in our country. In future health communication campaigns, we must make sure that we address this inequity. Furthermore, this study by Brissette and colleagues showed a better detection of stroke features, but prehospital care stakeholder must reinforce the message that stroke is a time-critical disease.