Primary angioplasty is the best treatment for ST-segment elevation myocardial infarction (STEMI) patients.1–4 However, even in the most developed countries, with a national hospital network equipped with catheterization laboratories and highly skilled teams working 24/7, primary percutaneous coronary intervention (pPCI) is rarely achieved within 120 min of symptom onset. Efforts therefore need to be made to enable better access to pPCI. For this purpose, it is essential to audit, monitor and assess coronary networks. Primary angioplasty should be conducted within 12 hours of symptom onset, but the greatest benefits are achieved if pPCI is performed within two hours.1,2,4,5

Total ischemic time, defined as time from symptom onset to reperfusion, is a well-established prognostic factor in STEMI patients.6-9 It is divided into two main periods, time from symptom onset to first medical contact (FMC) (patient delay), and time from FMC to pPCI (system delay).4,10

The guidelines suggest door-to-balloon time (time from arrival at a pPCI center to beginning of pPCI) as a measure to assess the hospital's performance in STEMI treatment.1,4,11 However, other studies have reported that reducing this time does not positively impact mortality11 and that delays at other stages also influence infarct size and mortality in STEMI patients.12,13

Studies have reported that longer system delay is associated with higher mortality and morbidity rates in STEMI patients.7,14–16

Stent for Life (SFL) is an initiative by the European Association of Percutaneous Cardiovascular Interventions and the European Society of Cardiology designed to improve the treatment of STEMI patients and to reduce STEMI-related mortality. SFL aims to increase the number of STEMI patients treated by pPCI in countries that join the initiative and to ensure that centers are able to perform pPCI 24/7.17 Portugal has been part of this initiative since 201118 and currently 18 centers in mainland Portugal perform pPCI procedures 24/7.

Various factors influence system delay. Importantly, failure to contact the emergency medical services (EMS) and off-hours presentation lead to longer system delay.19

In the last decade, pPCI rates in Portugal were among the lowest in Western Europe, though in recent years the procedure has been performed more frequently in Portuguese hospitals, suggesting that the country's participation in the SFL initiative has had a positive impact.20

This paper aims to assess changes in system delay since Portugal joined SFL, to identify factors that influence this time, and to identify areas of intervention designed to improve care of STEMI patients.

The sample comprised 838 patients who underwent pPCI between 2011 (T0) and 2015 (T4). Patient characteristics over the years are summarized in Table 1. In the last year (T4), patients included in the analysis were older (p=0.048) and had a higher prevalence of diabetes compared to previous years (p=0.032). Although the percentage of patients who called the national medical emergency number (112) did not change, the number of patients admitted to a pPCI center through the EMS tended to increase. By contrast, there was a downward trend in the proportion of patients transferred from local hospitals without pPCI facilities. Of the 319 patients with suspected STEMI who called 112, only 169 (53%) were in fact transported by the EMS (Table 1).

Considering only the patients transferred (n=379), there were differences between the time points in those transferred by EMS from another hospital to the pPCI center (inter-hospital transport: 21% vs. secondary transport: 79%) (p<0.001). The percentage of patients using secondary transport decreased throughout the survey (data not shown).

Table 2 presents the characteristics of system delay (FMC-to-ECG time, door-to-balloon time, system delay and total ischemic time) for the different time points. No differences were found between the four time points. The percentage of patients with FMC-to-ECG time <10 min was less than 50% at all time points. Door-to-balloon time <60 min decreased non-significantly from 56.4% to 48.8%. Two hundred and twenty-two patients (26%) had a system delay of ≤90 min and only 70 patients (8%) had total ischemic time of ≤120 min.

To identify potential factors predicting system delay, univariate and multivariate logistic regression models were used to analyze a set of variables that could influence the outcome (Table 3). The variables T4, female gender, age ≥75 years, Central and Alentejo regions, not calling 112, not arriving by their own means of transport to a pPCI unit, and attending a center without pPCI or other healthcare unit after FMC (patient transferred) were identified as potential factors predicting longer system delay in univariate logistic regression analysis (Table 3). Admission during off-hours periods (nights or weekends) was not significantly associated with longer system delay. In multivariate logistic regression analysis, only age ≥75 years, Central region, not calling 112 and attending a center without pPCI remained statistically significant predictors of longer system delay. Medians and interquartile range of system delay are presented in Figure 1 for each category of these predictive factors with p-values from multivariate logistic regression.

pPCI or other healthcare unit after FMC (patient transferred). Results are presented as medians and interquartile range.'> pPCI or other healthcare unit after FMC (patient transferred). Results are presented as medians and interquartile range.' title='Variables that impact system delay. (A) Age; (B) region; (C) calling 112; (D) attending a center without pPCI or other healthcare unit after FMC (patient transferred). Results are presented as medians and interquartile range.'/>

Figure 1.

Variables that impact system delay. (A) Age; (B) region; (C) calling 112; (D) attending a center without pPCI or other healthcare unit after FMC (patient transferred). Results are presented as medians and interquartile range.

(0.18MB).

Figure 2A shows that there were no significant changes in the number of patients calling 112 over the different time points of the survey (T0-T4) (p=0.739). When stratified according to whether or not patients called 112 (Figure 2B), differences between the groups were found for FMC-to-ECG time (p<0.001), system delay (p=0.001) and total ischemic time (p<0.001). It can also be seen that 32.9% of patients calling 112 had a system delay ≤90 min, which is significantly more than those who did not (22.5%). Similarly, considering as a variable whether the patient was transferred or was taken direct to the pPCI center by the EMS, differences were found for FMC-to-ECG time, system delay and total ischemic time (Figure 3). It can be seen that 33.1% of patients transported directly to a pPCI center by the EMS had a system delay ≤90 min vs. 13.2% of patients who were transferred (p=0.001), which corresponds to 20% more patients who were within the recommended system delay times.

ECG: electrocardiogram; FMC: first medical contact; T0: time zero, 2011; T1: time one, 2012; T2: time two, 2013; T3: time three, 2014; T4: time four, 2015.'> ECG: electrocardiogram; FMC: first medical contact; T0: time zero, 2011; T1: time one, 2012; T2: time two, 2013; T3: time three, 2014; T4: time four, 2015.' title='(A) Changes in numbers of patients calling for the different time points of the survey; (B) differences in system delay between patients who called and did not call 112. ECG: electrocardiogram; FMC: first medical contact; T0: time zero, 2011; T1: time one, 2012; T2: time two, 2013; T3: time three, 2014; T4: time four, 2015.'/>

Figure 2.

(A) Changes in numbers of patients calling for the different time points of the survey; (B) differences in system delay between patients who called and did not call 112. ECG: electrocardiogram; FMC: first medical contact; T0: time zero, 2011; T1: time one, 2012; T2: time two, 2013; T3: time three, 2014; T4: time four, 2015.

(0.11MB).

pPCI center by the EMS. ECG: electrocardiogram; EMS: emergency medical services; FMC: first medical contact.'> pPCI center by the EMS. ECG: electrocardiogram; EMS: emergency medical services; FMC: first medical contact.' title='Characterization of system delay according to whether the patient was transferred or was taken directly to the pPCI center by the EMS. ECG: electrocardiogram; EMS: emergency medical services; FMC: first medical contact.'/>

Figure 3.

Characterization of system delay according to whether the patient was transferred or was taken directly to the pPCI center by the EMS. ECG: electrocardiogram; EMS: emergency medical services; FMC: first medical contact.

(0.07MB).

System delay is one of the most important timings for assessing the quality of a system, but also for determining which reperfusion strategy should be used.23 Current guidelines suggest pPCI as the ideal reperfusion strategy in STEMI patients, unless it cannot be offered within the recommended timeframes, in which case fibrinolytic therapy should be considered. There is, however, still considerable debate concerning how much delay is acceptable when making this decision, or when a combination of the two reperfusion methods should be preferred.24,25 There is evidence that pPCI loses its advantage over fibrinolysis for longer delays after symptom onset.26

Our results suggest that there were no significant improvements in system delay between the beginning of SFL in Portugal (T0) and 2015 (T4). Furthermore, the study suggests that the aims stated in the current STEMI treatment guidelines1,2 have not been achieved in Portugal: system delay was over 115 min at all the time points of the study, and only 33.1% of STEMI patients transported directly to a pPCI center by EMS had a system delay of 90 min or less. The situation is very different in central and northern Europe. In daily practice in the Netherlands, almost all STEMI patients can be transported to a pPCI center within 60 min of FMC (which is usually the emergency call),27 while in Sweden the median FMC-pPCI time is 70 min.28

Univariate analysis of our results revealed that six factors influence system delay: age ≥75 years, female gender, Central and Alentejo regions, and attending a center without pPCI were associated with longer system delay, whereas the variables calling 112 and arrival by their own means of transport at a pPCI center were associated with shorter system delay. However, the multivariate model showed that only the variables age ≥75 years, Central region, and attending a center without pPCI were predictors of longer system delay, whereas calling 112 was predictive of shorter system delay.

In view of these results, it is important to look at the way elderly patients are handled when they enter the health system. This segment of the population presents specific characteristics that contribute to greater system delay. Low literacy levels, slowness and frailty affect how these patients are handled and transported in and between hospitals, leading to further delays.

The Central region presents significantly longer system delays than other regions of the country. One of the reasons for this result may be the greater distances in this region between the patient's location at the time of symptom onset and the pPCI center. Geographic adjustments in the STEMI network and improvements in transport (direct to a pPCI center and transfer) may lead to more equitable access to pPCI.

The fact that there were more patients who called 112 with system delay ≤90 min than those who did not may have a simple explanation: in the former group, the number of patients who arrive directly at a pPCI center and who spend no time in inter-hospital transfer is likely to be much higher. The fact that in our study patients who called 112 had significantly shorter system delay may be closely related to the SFL initiative in Portugal, as since the beginning of this initiative, the EMS started to contact pPCI centers directly, to transmit ECGs wirelessly, and to deliver STEMI patients directly to catheterization laboratories. This is confirmed by the fact that, in this study, more patients transported by the EMS had an FMC-to-ECG time of ≤10 min than those not transferred by the EMS. A recent study in the USA demonstrated that prehospital wireless electrocardiogram transmission reduced system delay.29 In Denmark, a study conducted between 1999 and 2009 investigated the impact of a gradual introduction of field triage for pPCI and associated outcomes. Among patients transported by the EMS from the scene of the event, the proportion who were triaged directly to a pPCI center in the field increased from 33% to 72%.30 Current guidelines confirm the importance of field triage, stating that the delay between FMC and diagnosis should be reduced to 10 min or less.1

In our study, STEMI patients who initially attended a center without pPCI had significantly longer system delay than those who arrived directly at a pPCI center. Comparison of the variables patient transferred vs. direct to pPCI center by EMS, as they relate to FMC-to-ECG time ≤10 min, system delay ≤90 min, and total ischemic time ≤120 min, suggests that attending a center without pPCI is a strong predictor of prolonged system delay. Significantly more patients transported directly to a pPCI center by the EMS had a system delay of ≤90 min. In the worst-case scenario (patient transferred), prolonged delay is closely related to time spent in the first hospital (door-in-door-out) and with inter-hospital transfer. Some studies suggest that door-in-door-out time is associated with patient management in the emergency room of the non-pPCI hospital.16,31

Although information campaigns to raise awareness of MI have been conducted, many patients still do not call the EMS and arrive at the hospital by their own means.5 However, in some cases admission to a center without pPCI may be through the EMS. This does not always imply a system failure; one study suggested that these delays are often due to diagnostic uncertainty instead. Of patients with suspected STEMI who called 112, only 53% were in fact transported by an EMS ambulance. In the other cases, the national referral center for emergency patients (CODU) did not activate EMS transportation for two main reasons: there were no EMS vehicles available, or the patients were misdiagnosed. In both cases, the patients were transported by the fire department and not by EMS ambulance, as should have occurred. In a Canadian study, patients requiring pPCI and undergoing inter-hospital transfer had longer symptom-call times, lower ECG ST-elevation scores, and more protocol-negative ECGs at presentation.32

Regarding transferred patients, our results suggest a positive trend, with a significant decrease being observed over the course of the study. Nevertheless, there are some actions that can still be implemented to improve the outcomes of transferred patients, as suggested in a five-year study in the USA, in which a program of rapid triage, transfer, and treatment of STEMI patients implemented in a rural area reduced in-hospital mortality and produced progressive improvements in door-to-balloon time.33

Raising public awareness, including strengthening and improving campaigns to publicize the onset of MI symptoms and to encourage people to call the 112 emergency number, could help reduce system delay.

Our results indicate that system delay did not decrease over the course of the SFL initiative, but this does not mean that the initiative was unsuccessful. Over little more than a decade, the use of pPCI tripled in Portugal.34 The first centers to perform pPCI were located in the largest urban centers, which were provided with more than one angiography room. In these urban centers, secondary transport was practically non-existent and few patients called 112. Thus, system delay was estimated on the basis of door-to-balloon time only. The spread of pPCI to peripheral centers and more frequent 112 calls led to an increase in system delay, mainly due to increased use of secondary transport. At the same time, primary transport by EMS also increased, which also increased system delay. In fact, although the system has become more efficient, the clock now starts ticking as soon as FMC occurs, before arrival at the hospital.

In the intermediate stage of such an initiative, while the system is still adapting and expanding, conventional quality indicators such as patient delay, system delay and door-to-balloon time are not sufficiently sensitive to assess how the initiative is developing. The percentages of patients who call 112, who go directly to secondary hospitals and who are transported by EMS may be more sensitive and earlier indicators that can be used to measure a positive evolution.

Our study has some limitations. Data were only collected on patients treated by pPCI, so they cannot be generalized to all STEMI patients. In addition, the survey only covered one month in each year, which means the possible effects of seasonal factors on the results were not addressed. For this reason, future surveys should collect data continuously.

In conclusion, this study showed that system delay did not change significantly during the study period. However, it revealed that the variables age ≥75 years, attending a center without pPCI, and Central region were significantly associated with prolonged system delay, whereas calling 112 was clearly associated with shorter system delay. Based on these factors, objective measures can be taken to reduce system delay and to improve clinical outcomes in Portuguese STEMI patients. However, efforts to improve outcome should not simply address a single quality measurement but should instead embrace a broader spectrum of procedures in MI care.

The authors state that they have no funding to declare.

The authors have no conflicts of interest to declare.