Feasibility of implementing reperfusion strategies in STEMI codes in a mountainous population remote from a PCI-capable center

INTRODUCTION

Acute myocardial infarction (AMI) is a medical emergency that requires a rapid response to minimize cardiac damage and improve patient survival. Initial recognition and early treatment based on the optimal reperfusion strategy are key to survival; however, implementing this protocol in rural and mountainous regions is a major challenge.1-4

Primary percutaneous coronary intervention (pPCI) is recommended in all cases provided it can be performed within 120 minutes (ideally in less than 90 minutes).4 If not contraindicated, fibrinolysis is the treatment of choice when this time frame cannot be guaranteed. Contraindications to fibrinolytic therapy may be classified as absolute or relative and should be evaluated on an individual basis. When fibrinolysis is contraindicated, primary angioplasty should be prioritized whenever feasible.1,4 The STEMI code in Catalonia was implemented as a regional health care network in June 2009 designed to organize the management of patients with suspected ST-segment elevation myocardial infarction (STEMI).5,6

Several authors have linked delays in reperfusion treatment to the type of infarction, the timing of symptom onset, and complications.7-9 Other studies suggest that a distance > 50 km to a PCI- capable center is associated with higher mortality rates vs early fibrinolysis.10,11 Other experiences, such as sharing patient data during prehospital care, including electrocardiograms (ECG) and the use of nighttime helicopter transfer to the PCI-capable center have been associated with shorter diagnosis-to-treatment times; however, a reduction in mortality has not been demonstrated.12 However, few studies have analyzed the factors causing delays in mountainous and hard-to-access geographic areas.

The main aim of this study was to determine the factors associated with delays in reperfusion treatment and those associated with mortality in patients treated under the STEMI code in a European mountainous area.

METHODS

Study design

We conducted an observational, retrospective, and quantitative study that included all STEMI code activations in the Alt Pirineu-Aran territory (Catalonia, Spain) from January 2015 through December 2020. The main sources of information were the STEMI code registry of the Department of Health of the Government of Catalonia5,6 and the database of the medical Emergency Medical Services (EMS) of Catalonia, which were cross-referenced to obtain a comprehensive overview. Because the data were derived from preexisting registries and analyzed anonymously, informed consent was deemed unnecessary. The study was approved by the Instituto Universitario de Investigación en Atención Primaria (IDIAP) Jordi Gol Ethics Committee, code CEIm 22/238-P. The SAGER guidelines were followed regarding potential sex and gender bias.

Study setting

Alt Pirineu-Aran is a mountainous region comprising 6 counties and represents 18% of Catalonia’s territory, yet it is home to less than 1% of its population. Population density is extremely low (12.6 inhabitants per km²), and most towns are located between 500 and 800 meters above sea level, far from a specialized hospital center.

This regional health care system faces several challenges. Prehospital care is provided by the EMS, a public service that operates 24 hours per day and provides coverage throughout the entire territory. Each county is served by 1 advanced life support unit, and the region has access to 1 medicalized helicopter, 1 of the 4 operating in Catalonia. Alt Pirineu–Aran includes 4 county hospitals, all of which are non–PCI-capable centers (figure 1).

Figure 1. Alt Pirineu-Aran Health Region. H03: Tremp medicalized helicopter. H: county hospitals. Unit call signs are identified with A for Alt Pirineu; 1st number indicates the type of unit (7, basic life support; 4, advanced life support + nurse; and 6, advanced life support + physician); 2nd and 3rd numbers indicate the location of the units on the map.

Hospital Universitario Arnay de Vilanova (Lleida, Spain) serves as the reference center for pPCI for Alt Pirineu-Aran, with the exception of the Cerdanya basic health area, where STEMI code patients are transferred to centers in the Barcelona metropolitan area (table S1).

Definitions and inclusion criteria

The definitions of “delay” used in the study were more than 10 minutes for fibrinolysis administration and more than 120 minutes for pPCI, measured from the time of ECG acquisition. These criteria were established on the basis of former studies and are consistent with current European clinical practice guideline recommendations.1-4

According to EMS protocols, a 90-minute transfer threshold—from ECG acquisition to arrival at the receiving hospital—is used to allow adequate time to perform pPCI and ensure compliance with the 120-minute target. To determine whether patients had a transfer time of less than 90 minutes to a PCI-capable center, a geographic analysis based on distance and estimate travel-time maps was performed.

Patients were included if the STEMI code was activated in Alt Pirineu–Aran and they were attended by EMS during the study period, as well as those who died after prior activation of the STEMI code.

Incomplete cases or those with coding errors were excluded, as were patients transferred to Toulouse (France) from Vall d’Aran.

Study variables

The variables analyzed included demographic, clinical, and logistical data. The primary time intervals assessed were symptom onset to first medical contact, time from first medical contact to ECG, and ECG to initiation of reperfusion treatment. The type of reperfusion strategy (fibrinolysis or pPCI) was recorded as well. Other relevant variables included the location of STEMI code activation, distance to the PCI-capable center, mode of transport to the PCI-capable center (ambulance or helicopter), and type of transfer (primary: direct care and transfer by a medicalized EMS ambulance; secondary: interhospital transfer; or delayed primary: initial assessment by a primary care physician or nurse-staffed ambulance followed by transfer to a medicalized ambulance or medical helicopter). Acute-phase complications and all-cause mortality at first medical contact, and at 24 and 48 hours, and at 15 days were also recorded.

Statistical analysis

The descriptive statistical measures used were absolute and relative frequencies for qualitative variables; mean and standard deviation for quantitative variables with normal distribution; and median and interquartile range for the remaining non-normally distributed quantitative variables, according to the Shapiro-Wilk test.

We analyzed a total of 4 binary outcome variables: use of fibrinolysis as the initial treatment, delays in fibrinolysis (> 10 minutes from ECG acquisition), delays in pPCI (> 120 minutes from ECG acquisition), and mortality. Furthermore, we evaluated associations between each outcome variable and patient- and care-related characteristics using the chi-square test for qualitative variables (or Fisher’s exact test when expected frequencies were < 5), the Mann-Whitney U test for non-normally distributed quantitative variables, and the Student t test otherwise. In addition, we analyzed the importance of variables for treatment delay and mortality using the Boruta algorithm for variable selection. Only variables not rejected by this algorithm were selected for subsequent multivariable analyses to reduce the risk of overfitting. A conditional inference classification tree was constructing using a Monte Carlo test with a minimum terminal node size of 3.

All statistical analyses were performed using R statistical software (R Foundation for Statistical Computing, Austria). P values < .05 were considered statistically significant.

RESULTS

During the study period, a total of 24 125 STEMI codes were activated across Catalonia. The study analyzed 225 cases occurring in Alt Pirineu-Aran, representing less than 1% of the total. Four patients were excluded for not meeting STEMI code criteria (1 pulmonary thromboembolism, 2 coding errors, and 1 duplicate case) (figure 2).

Figure 2. Flow diagram of STEMI codes. AMI, acute myocardial infarction; pPCI, primary percutaneous coronary intervention; EMS, Emergency Medical Services of Catalonia.

The mean age of the 221 included patients was 64.7 years (range, 27–96). Of these, 72.4% were men and 67.4% resided in the study area. All STEMI codes were activated after ECG acquisition at first medical contact, either at a county hospital (51.6%), by EMS at the patient’s home or in a public setting (28.9%), or at a primary care center (19.5%). The median time from first medical contact to ECG acquisition was 6 minutes, and from pain onset to ECG acquisition, 90 minutes (table 1).

Table 1. Clinical characteristics and care times of activated STEMI codes and comparison according to therapeutic decision

The incident location had a mean altitude of 838 meters and was located at distances ranging from 81 km to 257 km from the PCI-capable center (median, 131 km). Overall, 76.5% of patients were situated more than 90 minutes from the PCI-capable center (median estimate transfer time, 106 minutes). Air advanced life support was used in 52.5% of the transfers (table 1).

Differences were observed in the time from ECG acquisition to pPCI according to mode of transport (Kruskal-Wallis; P < .001). The median time was 183 minutes for ground ambulance, 138 minutes for helicopter transport, and 140 minutes for combined transport (table 1). Pairwise comparisons using the Mann–Whitney U test showed significant differences compared with ground transport after adjustment for the false discovery rate.

We observed marked variability in the annual frequency of STEMI code cases, with a particularly high number in 2019, and in the proportion of first-assistance fibrinolysis performed from a minimum of 10.5% in 2020 to a maximum of 34.6% in 2017. The number of cases attended in 2020 (the year of the COVID-19 pandemic) was slightly higher than in 2015–2017 (38 cases [17.2%]) but lower than in 2019, which recorded 57 cases (25.8%) (table 1).

Delays in treatment

In 91.5% of patients in whom the therapeutic decision was to perform fibrinolysis at first medical contact, the time between ECG acquisition and treatment exceeded the recommended threshold (> 10 minutes).

When the therapeutic decision was to perform pPCI, the time between ECG acquisition and pPCI exceeded the recommended threshold (> 120 minutes) in 79.6% of cases (table 2).

Table 2. Factors associated with delay in reperfusion treatment

Among patients located more than 90 minutes from a PCI-capable center who did not receive fibrinolysis at first medical contact, the reason for withholding treatment was not documented in 87 cases (68.5%). Although 76.5% of patients were situated more than 90 minutes from the PCI-capable center, a substantial underuse of fibrinolytic treatment was observed. Of note, patients with a longer time interval from symptom onset to ECG acquisition were more likely to receive fibrinolytic treatment earlier.

Interhospital transfers, incidents without participation of air advanced life support, and nighttime incidents located more than 90 minutes from the PCI-capable center showed greater delays in performing pPCI. Exclusive use of ground ambulance for transfer to the PCI-capable center was associated with a 96.9% rate of delay and emerged as the primary variable in the classification tree for delay between ECG acquisition and pPCI. Among helicopter transfers, delays were more frequent in interhospital transfers (85.4%). In primary or delayed primary transfers, delays were more common when the incident location was more than 90 minutes from the PCI-capable center (65.1%) compared with locations less than 90 minutes away (23.1%) (figure 3).

Figure 3. Classification tree for delay in primary percutaneous coronary intervention (pPCI). Three factors were significantly associated with delay: mode of transport (ambulance [A], helicopter [H], or combined use), type of activation (primary [P], delayed [D], or interhospital [S]), and estimated road travel time from the event location to the PCI-capable center in minutes (> 90 minutes < 90 minutes).

Mortality

A total of 18 patients (8.1%) died within the first 15 days. Mortality was significantly higher in patients who experienced a major event (asystole, intubation, shock, or ventricular fibrillation) within the first 24 hours. Age was significantly associated with mortality only at the 48-hour and 15-day follow-up. There were no significant differences in 15-day mortality between patients treated with fibrinolysis and those directly transferred for pPCI. Increased mortality was associated with treatment delays, particularly in the early mortality subgroup (< 24 hours); however, these differences did not reach statistical significance (table 3).

Table 3. Factors associated with mortality according to time from clinical course prior to death

DISCUSSION

In our study, only 20.4% of patients who underwent pPCI received treatment within 120 minutes. Only 4 of the 47 patients treated with fibrinolysis received therapy within 10 minutes.

Several studies, such as the STREAM,13 have shown that prehospital fibrinolysis followed by early pPCI may offer results similar to direct transfer for pPCI if patients are treated within the first 3 hours from symptom onset. In our analysis, we identified underuse of fibrinolytic therapy, along with insufficient documentation of the reasons for withholding treatment.

In contrast to the studies by Stopyra et al.9,14 conducted in North Carolina (United States) in which 60.5% of patients underwent pPCI within ≤ 90 minutes, only 20.4% of patients in our study achieved reperfusion within a broader threshold (< 120 minutes). These findings underscore longer treatment delays in our region and support more frequent consideration of fibrinolytic therapy.

Although Aboal et al.10,11 reported higher mortality rates associated with delayed pPCI, we did not observe a significant association in our study, despite a substantial lower proportion of patients undergoing pPCI within the target time (20.4% vs 42%). In addition, fibrinolysis was more frequently used in patients located more than 50 km from the PCI-capable center (66.7%), reflecting structural and logistical constraints specific to the Alt Pirineu-Aran region that influence treatment selection and reperfusion times.

In line with former evidence regarding mortality, the study results reinforce the clinical importance of time to care as a potential factor associated with worse prognosis, which is particularly relevant in early mortality (< 24 hours).

Compared with the study by Carol et al.,15 in which 58% of patients underwent pPCI after more than 120 minutes, our data show a substantial higher proportion (79.6%). Factors associated with delay, such as intubation, initial shock, and nighttime care, were consistent between the 2 studies; however, other variables, including left bundle branch block, were not. A notable finding of our study is that the care delivered by the EMS was associated with shorter treatment times compared with county hospitals.

Patient origin, geographic location, sex, number of resources involved, time required for therapeutic decision-making, and mode of transport influenced treatment times as well. Hakim et al.16 suggest that helicopter transport is less effective than ground transport, especially for distances of less than 50 km. In our study, 71% of patients transferred by helicopter (all located more than 50 km from the PCI-capable center) received pPCI after more than 120 minutes, which may be associated with the lack of a helipad at the reference PCI-capable center, requiring additional ground transfer, and service hours, mainly daytime during the study period. In constrast to other studies conducted during the COVID-19 pandemic,17,18 our series did not demonstrate a reduction in case volume or prolongation in alert time. However, 2020 was the year with the lowest use of fibrinolysis (4 cases, 10.5%).

Several studies indicate that delays in STEMI recognition and code activation have a direct impact on reperfusion time. One study highlights that training significantly improves these times,19 suggesting that lack of training or skills among professionals may generate delays in care and lower use of fibrinolysis. Our study shows appropriate timing in ECG acquisition and emergency recognition by teams. Delayed primary transfers included the highest percentages of patients with optimal reperfusion times (table 2).

Our findings underscore the need to review and optimize action protocols, particularly in non-PCI-capable centers and geographic areas with greater delays. It is essential to optimize professional response to suspected STEMI, improve therapeutic decision-making time, and explore innovative solutions such as triage systems and STEMI code detection with technological support, optimization of air transport, and greater coordination between PCI and non-PCI capable centers.

Limitations and strengths

This study has limitations due to the small population size in the rural areas studied, which required extending the study period to 6 years to obtain the collected sample. This extension implies that data underwent several changes in record systems (from paper to digitized format) and organizational modifications (implementation of triage in emergency departments and nighttime helicopter flights). The sample of 221 patients remains limited and may have affected the statistical power to detect significant differences between analyzed variables. Similarly, distances from the incident location and estimated average travel times under optimal conditions were used, without accounting for any potential delays due to traffic or other unforeseen factors.

CONCLUSIONS

The results of this study demonstrate the need to increase the use of fibrinolysis in areas distant from a PCI-capable center to reduce reperfusion delays. Documentation of the reasons for not performing fibrinolysis should be improved, as this limits interpretation regarding therapeutic appropriateness. Joint initial care by primary care and emergency teams, as well as delayed primary transfers, reduce reperfusion times and avoid interhospital transfers. Finally, the low number of deaths during the study period prevents multivariate analysis and only allows identification of the variables or characteristics associated with mortality described in the results.

Despite these limitations, the study provides a comprehensive analysis of variables and describes in detail how the investigated population is managed and transferred, something unprecedented in this context. In addition, cross-referencing and thorough review of 2 databases provide valuable information to assess treatment delays. Therefore, this study is a solid basis for future advances in improving STEMI code management in rural areas.

FUNDING

This study was funded by the Provincial Council of Lleida through “The strength of municipalities” project and by IRBLleida through project PP10851 of the Alt Pirineu-Aran Intramural Research Program (IREP).

ETHICAL CONSIDERATIONS

This study was approved by the IDIAP Jordi Gol Ethics Committee, code CEIm 22/238-P. The SAGER guidelines were followed regarding potential sex and gender bias.

STATEMENT ON THE USE OF ARTIFICIAL INTELLIGENCE

ChatGPT was used to improve the wording of some paragraphs of the article. After using this tool, the authors reviewed and edited the content as necessary and take full responsibility for the final version.

AUTHORS’ CONTRIBUTIONS

M. Navarra Llorens was responsible for study conception and design, overall supervision, and manuscript drafting. M. Martínez Alonso conducted the statistical analysis, interpreted the results, and critically reviewed the content. Y. Azeli was responsible for clinical analysis and critical review. S. Ferrandis Barrés collected data and contributed to the discussion and critical review. M. Canelles Seix collected data and reviewed the manuscript. L. Duch Grau participated in data collection, manuscript review, and final approval. A.M. Forradelles Rey collected data and performed critical review. M.A. Martínez Momblan supervised the project and critically reviewed the manuscript. X. Jiménez-Fàbrega supervised the project and provided intellectual contributions to the discussion. All authors approved the final text.

CONFLICTS OF INTEREST

None declared.

ACKNOWLEDGMENTS

We thank Mar Franch Casanovas and Francisco Iturbe Recasens for their collaboration in data collection, and Isidre Felip for his advice in drafting the manuscript.

SUPPLEMENTARY DATA

REFERENCES