For patients with ST-Elevation myocardial infarction (STEMI), time-to-treatment is an important modifiable determinant of survival [1, 2]. Timely primary percutaneous coronary intervention (PCI) is considered the preferred method of reperfusion compared to fibrinolytic therapy, given its lower STEMI mortality rate [3]. However, about 25% of Canadian residents do not live within a one-hour drive of a PCI-capable centre and often have to be transferred from a non-PCI capable (referral) hospital [4]. Consequent lengthening in receipt of PCI therapy has a demonstrably negative impact on clinical outcomes [5, 6].

Delays in providing timely reperfusion therapy can occur at several points along the treatment pathway but take place frequently at the referral hospital [7, 8]. Guidelines for STEMI recommend that the time between arrival to, and transfer from a referral hospital to a PCI-capable hospital, also referred to as the door-in-door-out (DIDO) time, should be ≤30 min, and has been widely adopted as an important metric for quality of STEMI care [1, 9]. However, several studies have noted that this benchmark is rarely achieved [7, 8, 10‐12].

In light of this, there is a clear need to develop new strategies or refine pre-existing ones that are responsible for prolonged DIDO times. One such suggestion is having emergency medical services (EMS) personnel administer a 12-lead electrocardiogram (ECG) prior to arrival at the hospital (pre-hospital ECG) for patients suspected of having a STEMI, although the impact on reducing DIDO times has not been well studied [13].

Utilizing a population-based cohort of STEMI patients in Ontario, Canada, we aimed to identify potentially modifiable factors to improve DIDO time. A secondary objective was to examine the impact of DIDO times on 30-day mortality rates adjusted for important confounders. Our findings could have important public health and policy implications, providing ‘real world’ evidence for areas in the STEMI system that should be targeted for improvement.

In the present study, we utilized a population-based cohort of STEMI patients in Ontario, Canada to identify potentially modifiable factors that can improve DIDO times and the impact of DIDO times on long-term mortality. Only 20% of STEMI patients who were transferred from a referral hospital to a PCI-capable centre for primary PCI met the DIDO time of ≤30 min. Age was the strongest patient-level predictor of DIDO times. Transport to hospital by EMS in combination with receiving a pre-hospital ECG was the strongest process of care measure associated with DIDO times. In particular, this group had substantively shorter symptom to reperfusion times compared to those who transported themselves to hospital, driven largely by significantly shorter symptom to door and DIDO times. The mortality gradient for DIDO times observed in our study is important as it demonstrates poor long-term outcomes for those who experienced longer DIDO times. The crude 30-day mortality rates in patients transferred for primary PCI was higher than those who got fibrinolytic therapy, in particular for those with DIDO times over 90 min.

Our finding that achievement of benchmark DIDO time was low (20%) is consistent with previous studies [7, 8, 10‐12]. A study using national registry data from the United States reported that only 11% of patients achieved DIDO benchmarks, and other studies reported rates of ~ 10% [8, 10, 11]. Canadian studies have reported higher rates than American-based studies, one reported 14% in Quebec [7, 10]. Several factors could account for the disparities in rates including varying characteristics of the study population, regional differences in policies and design of the STEMI system and/or changes in trends over time [7, 8, 10‐12]. The latter may be especially pertinent in this case given that our work is more recent than other cited studies.

The guidelines indicate that 90% of patients transferred from a referring hospital should meet the 120-min time-to-treatment standard [1]. Less than 50% of transferred patients in our study met the FMC-to-balloon time benchmark. We observed stark differences in achievement of FMC-to-balloon time benchmark across DIDO status. Most transferred patients with DIDO time > 30 min do not have absolute contraindications for fibrinolytic therapy [18, 19]. In our and previous studies, about half of transferred patients had a DIDO time > 60 min, with few of these individuals receiving timely PCI [7, 8, 10‐12]. Delayed transfer of STEMI patients with DIDO time > 60 min may not be a superior strategy to the timely use of fibrinolytic therapy [1, 3, 20]. One potential option to address prolonged DIDO times might be to increase the use of fibrinolytic therapy at referral hospitals, followed by transfer for catheterization post-lytic therapy [21].

Elderly age was found to be one of the strongest predictors of DIDO time in our study which is in light with previous studies [7, 10]. Elderly individuals may be less likely to present with classic STEMI symptoms (e.g., retrosternal chest pain) and/or more likely to have ECGs which are complex to interpret, that may delay diagnosis and recognition of a STEMI. Our findings highlight there is a need to bring awareness to the complexity and non-traditional nature of the symptoms with which these elderly patients sometimes present in order to improve DIDO times for this group. Prior studies have shown the disparities in time to reperfusion for STEMI patients in the elderly [22, 23]. Better software that can accurately read difficult ECGs, enhanced training for medical personnel and public awareness campaigns might improve DIDO times.

We observed that patients transported by EMS personnel, whether in association with a pre-hospital ECG or not, resulted in shorter symptom-to-door times compared to those self-transported patients. However, DIDO times were found to be significantly shorter compared to patients who self-transported only amongst those who received a pre-hospital ECG, suggesting that while EMS transport, irrespective of pre-hospital ECG status affects symptom-to-door time, for DIDO times it is the receipt of pre-hospital ECG that makes a difference. Provision of a pre-hospital ECG was not a universally adopted or mandated practice in Ontario in 2012 [9]. However, the findings of the present study suggest that policies which mandate and fund equipping ambulances with this technology and training EMS personnel to administer these ECG’s might have a significant impact on reduction of DIDO times and mortality.

Ideally, provision of a pre-hospital ECG should allow for a quicker diagnosis, thus facilitating the activation of STEMI protocols, result in more direct transfers to PCI-capable centres or mitigate delays in the referral ED [24]. Notably, our study found that a fair proportion (11.1%) of individuals received a pre-hospital ECG and were still initially transferred to a non-PCI capable hospital. Contributing factors may include diagnostic uncertainty, incorrect interpretation of the ECG and/or a clinically unstable patient. It is also plausible that some ambulances continue to bring patients to the nearest community hospital, regardless of ECG findings because the PCI-capable centre is further away, which would identify an area to target for improvement. Increased clarity in the guidelines around when EMS personnel should bypass the closest centre in favor of a more distant PCI-capable hospital, which is integrated within a STEMI protocol, may in part help alleviate this problem. In 2015, a STEMI bypass protocol in Ontario enables paramedics to bypass local hospitals and transport patients with STEMI directly to PCI-capable centres if arrival at the PCI centre will be ≤60 min from the first medical contact [9]. These guidelines came into effect in the province in 2017 and future studies should examine whether new guidelines have reduced the number of patients with STEMI-positive ECG’s being transferred to non-PCI capable hospitals. Our findings show that the major delay in reperfusion times occurs at the referral hospital, in line with previous studies’ findings [10, 25]. However, it should also be recognized that it may be difficult in EDs to provide care within guideline recommendations for every patient, especially for those with atypical presentations.

The present study aids in generalizability of findings to other populations. The fact that we examined important process of care measures including receipt of pre-hospital ECG’s, which have not been extensively studied enabled us to assess the impact of current STEMI management recommendations and guidelines on DIDO times so that we can ascertain a more complete picture of the STEMI care system and identify areas for improvement. Additionally, the present study found DIDO times to be an independent predictor of 30-day mortality, with the observed gradient strengthening the conclusion, and suggests the adverse impact of long DIDO times extends beyond the hospital visit itself. Previous studies have observed similar gradients for in-hospital mortality only or have been underpowered to make inferences regarding longer-term mortality outcomes adjusted for important confounders [10].

However, the present study presents with some important limitations. First, information on key time intervals was ascertained through retrospective chart review and recorded times could not be independently validated. Secondly, we were unable to collect information on the distance to the referral hospital or between the referral hospital and the PCI-capable centre or the details of patient symptoms as they were not available or well documented. Lastly, the administrative STEMI codes we used have not been validated. However, they were applied consistently across all records and those who were identified as STEMI cases from administrative data had their diagnosis verified via chart review by a trained cardiology research nurse, making it unlikely that non-STEMI cases were included in the study.

In conclusion, our findings suggest that benchmark DIDO times were not achieved for a majority of transferred STEMI patients in the province. Patient age and process of care measures, namely symptom-to-FMC time and receipt of a pre-hospital ECG, were identified to be strong independent predictors of DIDO time. Despite provision of pre-hospital ECGs, a fair proportion of patients were still transferred to non-PCI capable hospitals, suggesting that policies and system-level changes aimed at bypassing non-PCI capable hospitals in favor of PCI-capable hospitals for those with STEMI positive ECGs could have an important impact on outcomes.