Paramedic-Delivered Fibrinolysis in the Treatment of ST-Elevation Myocardial Infarction: Comparison of a Physician-Authorized versus Autonomous Paramedic Approach

Abstract

Background: For those patients who receive fibrinolysis in the treatment of ST-elevation myocardial infarction (STEMI), early treatment, i.e., within 2 hours of symptom onset, confers the greatest clinical benefit. This rationale underpins paramedic-delivered fibrinolysis in the prehospital setting. However, the current New Zealand approach requiring paramedics to first gain physician authorization, has proved inefficient and time consuming, particularly due to technological failings. Therefore, this study aimed to trial a new autonomous paramedic-delivered fibrinolysis model, examining the impact on time-to-treatment, paramedic diagnostic accuracy and patient outcomes. Methods: Utilizing a prospective observational approach, over a 24-month period, paramedics identified patients with a clinical presentation and electrocardiogram features consistent with STEMI, and initiated fibrinolysis. These patients were compared to a historic cohort who received fibrinolysis by paramedics within the same regions but following physician authorization. The main outcome measures were pain-to-needle (PTN) time and accuracy of paramedic diagnosis. A secondary end-point was 30-day and 6-month mortality and hospital length of stay (LOS). Results: A total of 174 patients received fibrinolysis (mean age, 64 years, SD ± 11.2). Median PTN time was 87 minutes (IQR = 58) for the historic cohort (n = 96), versus 65 minutes (IQR = 31) for the experimental cohort (n = 78), (p = 0.007). Autonomous paramedic diagnosis showed a sensitivity of 96% (95% CI 89–99) and specificity of 91% (95% CI 76–98). A significant reduction in both 30-day mortality and hospital LOS was observed among the experimental cohort (p = 0.04 and <0.001, respectively). No significant difference was observed between groups in terms of 6-month mortality. Conclusions: Prehospital fibrinolysis provided autonomously by paramedics without direct physician oversight is safe and feasible. Moreover, this independent approach can significantly improve time-to-treatment, resulting in short term mortality benefit and reduced hospital LOS.

Key words:

• prehospital
• STEMI
• paramedic
• fibrinolysis

Introduction

Throughout New Zealand, fibrinolysis remains the primary reperfusion strategy for many patients suffering ST-elevation myocardial infarction (STEMI), particularly outside metropolitan settings where onsite interventional cardiology services do not exist, and where transport for percutaneous coronary intervention (PCI) is seldom possible within mandated timeframes (1). Timely provision of fibrinolysis confers the greatest clinical benefit in terms of morbidity and mortality (2, 3). Internationally, paramedic-delivered prehospital fibrinolysis programs have been most effective in addressing this time-to-treatment imperative, especially compared to the treatment’s provision in-hospital (4, 5). In New Zealand, current paramedic programs utilize a physician-authorised telemetry-based model. Here, when paramedics attend a patient with ischemic symptoms, they acquire and transmit the patient’s 12-lead electrocardiogram (ECG) to the receiving hospital for physician assessment. The paramedics then proceed through a fibrinolysis contraindication checklist with the patient before phoning the physician, who authorizes them to administer the treatment (or not) following full case discussion. This is the most commonly used approach by ambulance providers internationally (6).

However, these telemetry-based systems are not only costly, but also time-consuming and problematic, particularly due to technological failings (7–9). For example, a retrospective audit of a paramedic fibrinolysis program in the regions of Northland and Hawke’s Bay, New Zealand, revealed that over a six-year period transmission issues occurred a third of the time, and that 28 out of 124 patients (22%) could not receive fibrinolysis from paramedics because of complete transmission failure (7). A possible solution is to remove ECG transmission and physician consultation from the process. Instead, paramedics assess the patient and make an independent clinical decision as to whether fibrinolysis is indicated or not. This autonomous paramedic model of care has already been trialed in England, Wales, and the Netherlands, where it proved both safe and effective (8, 10, 11). Therefore, the aim of this study was to trial the same model in the New Zealand setting, and to examine the impact on time-to-treatment, the accuracy of paramedic diagnosis and the impact on patient outcomes, compared to the physician-authorised telemetry-based approach. Internationally, this research is the first to undertake such a comparison.

Methods

Design and Setting

We utilized a prospective analysis of differences study design to compare two groups of STEMI patients, both of which received fibrinolysis from paramedics but by two different models of treatment authorization. The first group, an historic control cohort (referred to as the pre-implementation group), received fibrinolysis from paramedics under the physician-authorised telemetry-based model. The second group, our prospective experimental cohort (referred to as the post-implementation group), received fibrinolysis from paramedics who made an autonomous clinical decision under protocol guidance, without physician oversight and authorization. The difference between these two groups was the “intervention”: the implementation of a new fibrinolysis protocol that permitted this independent model of paramedic treatment provision.

This study was undertaken with St John, the country’s main emergency ambulance provider, which covers approximately 95% of the country. Two geographic regions were included, Northland and Hawke’s Bay. From a national population of 4.792 million, both regions service a combined population of 328,000 and share similar key geographic and demographic features (12). Each region had previously implemented paramedic fibrinolysis programs utilizing the physician-authorised telemetry-based model, allowing a direct comparison between the two treatment authorization processes. In both regions the main hospital is a medium-sized secondary-level facility without interventional cardiology services. The study was conducted over a 24-month period (May 2015–May 2017) and enrolled 91 intermediate and advanced life support level paramedics, all of whom had been enlisted in the previous fibrinolysis programs.

Study Population

The pre-implementation group included all patients with STEMI who received paramedic-delivered fibrinolysis under the previous physician-authorised telemetry-based model, occurring from October 2008 to April 2015. The post-implementation group included all patients with STEMI who received fibrinolysis by autonomous paramedics in line with protocol criteria over the trial’s 24-month period. Those patients transported by paramedics over this same time-period who did not receive fibrinolysis in the field but had a hospital diagnosis of STEMI were also investigated.

Sample Size

To measure all outcomes, based on a previous New Zealand paramedic prehospital fibrinolysis study involving 100 patients by Ranchord et al. (5), and an a priori power analysis using their data, 108 patients were required for this study: 54 in each of the two observed groups (13). This assumed a sensitivity of 70%, an accuracy measure of ±10% and an alpha level of 0.05, denoting a 95% confidence interval.

Criteria for Diagnosis of STEMI and Fibrinolysis Drug Regime

Diagnostic criteria for STEMI were per Cardiac Society of Australia and New Zealand (CSANZ) guidelines (1). However, for the post-implementation phase, a new or presumed new left bundle branch block (LBBB) was considered a contraindication to treatment. The fibrinolysis drug regime utilized throughout both study phases included single bolus intravenous (IV) tenecteplase, IV bolus enoxaparin (for patients < 75 years of age only) and subcutaneous enoxaparin, plus oral clopidogrel and aspirin.

Measures

Patient Characteristics

To describe the patient sample population, key demographic data was collected, as well as clinical characteristics on initial presentation and cardiovascular disease (CVD) risk factors. The time interval from symptom onset to receipt of 111 emergency call was also measured.

Time-to-Treatment

Mandated treatment-related target time intervals (“needle times”) were measured, per CSANZ Guidelines (1), and compared between groups. These included:

1. Pain-to-needle (PTN) time and call-to-needle (CTN) time target: <60 minutes

2. Emergency medical services contact-to-needle (ETN) time target: <30 minutes

3. First diagnostic STEMI ECG to needle time target: <20 minutes

Accuracy of Paramedic Diagnosis and Protocol Application

Sensitivity and specificity, positive predictive value (PPV) and negative predictive value (NPV) of the accuracy of paramedics’ clinical diagnoses were determined. This was possible only in the post-implementation phase of the study. Three independent cardiology consultants reviewed all cases after the fact, to determine accuracy of diagnosis and protocol application.

Patient Clinical Management and Outcomes

The following data was collected for both observed groups and analyzed to determine patient outcomes:

1. Clinical complications in the field, i.e., cardiogenic shock, compromising arrhythmia (not including cardiac arrest), and cardiac arrest

2. Inappropriate fibrinolysis, i.e., no rise in cardiac biomarkers identified

3. Coronary angiography performed; initial findings

4. Medical management only during hospital admission

5. PCI performed (including rescue and non-urgent PCI)

6. Emergent coronary artery bypass grafting (CABG) from the cardiac catheterization laboratory (CCL), and elective CABG performed later during admission

7. PCI procedure complications

8. For those who underwent rescue PCI, the proportion receiving the treatment within guideline-mandated timeframes

9. In-hospital major bleed (excluding intra-cranial hemorrhage (ICH))

10. ICH

11. In-hospital reinfarction

12. Thirty-day and six-month mortality

13. Hospital length of stay (LOS)

Hospital LOS (measured in bed days) is an indicator of public health system costs. It can also indicate that other in-hospital treatment interventions were necessary.

Intervention

Autonomous Paramedic Fibrinolysis Protocol

The treatment indication criteria were similar across both groups, with two main exceptions. Firstly, there were three additional 12-lead ECG criteria required in the new autonomous paramedic protocol, designed to mitigate the risk of erroneous ECG interpretation of STEMI. These were:

• automated interpretation of STEMI

• normal QRS width (≤0.12 s) OR right bundle branch block

• heart rate <130 beats/minute.

A positive automated interpretation of STEMI was mandated to provide decision support. If the treating paramedic disagreed with the automated interpretation (whether to treat or not to treat – termed “discordance”), fibrinolysis was to be withheld.

The second main exception related to treatment contraindications. In the new protocol, patients with a LBBB or aged ≥85 years were excluded from receiving fibrinolysis. All other treatment indications/contraindications were as per CSANZ Guidelines (1).

Data Analysis

All analyses were performed using the Statistical Package for the social sciences software program version 24. Descriptive analysis included frequencies, means and standard deviations, or medians and interquartile ranges. Continuous measures were tested for normal distributions. The rates of True Positive and True Negative, and False Positive and False Negative cases were identified, and the sensitivity, specificity, PPV and NPV of paramedic diagnosis and protocol application were calculated. Outcome variables were initially compared using chi-square tests for independence or independent-sample t-tests. Comparison of medians was made using Mann-Whitney U tests. All tests were 2-sided, and p values < 0.05 were considered statistically significant.

Ethics Approval

Ethical approval for the study protocol was obtained from New Zealand Health and Disabilities Ethics Committee—Northern A (14/NTA112) and from the Auckland University of Technology Ethics Committee (15/03).

Results

Patient Characteristics

A total of 174 patients were included in the study—96 patients in the pre-implementation group and 78 patients in the post-implementation group. Table 1 reports on patient demographics and CVD risk factors. These were similar overall between the two observed groups. Most patients were male (combined average age of 64 years, SD ± 11.2), of European ethnicity and residing in more socioeconomically deprived areas (as measured by the New Zealand Deprivation Index). Smoking and a family history of acute coronary syndrome (ACS) were the most prominent risk factors identified. Clinical characteristics of patients on initial presentation were similar between the two groups (Table 2).

Table 1. Demographics and cardiovascular disease risk factors: Comparison of the pre- versus post-implementation group

Variable	Pre-implementation group (n = 96) n (%)	Post-implementation group (n = 78) n (%)	Test statistic	p-value
Sex—Male	68 (71)	53 (68)
Female	28 (29)	25 (32)	χ^2(1) = 0.169	0.68
Mean age in years (±SD)	63 (±11.1)	64 (±11.5)	t (172) = 0.21	0.82
Ethnicity—European	(84)	64 (82)
Māori	12 (13)	11 (14)
Pacific peoples	3 (3)	3 (4)	FET	0.90
NZDep score 1–5	15 (16)	19 (24)
NZDep score 6–10	81 (84)	59 (66)	χ^2(1) = 1.569	0.21
CVD Risk factors—HTN	52 (54)	51 (65)	χ^2(1) = 1.366	0.24
Diabetes	17 (18)	10 (13)	χ^2(1) = 0.456	0.50
Hyperlipidaemia	47 (49)	35 (45)	χ^2(1) = 0.148	0.70
Increased BMI	34 (36)	31 (40)	χ^2(1) = 0.184	0.66
Current smoker	61 (63)	54 (69)	χ^2(1) = 0.394	0.53
Family history of ACS	67 (70)	48 (61)	χ^2(1) = 0.966	0.32
IHD	36 (38)	23 (29)	χ^2(1) = 0.901	0.34
Previous AMI	10 (11)	12 (15)	χ^2(1) = 0.564	0.45
Previous PCI	13 (14)	9 (11)	χ^2(1) = 0.028	0.86
Previous CABG	3 (3)	3 (4)	FET	1.00

χ² = chi-square; SD = standard deviation; t = t-test value; FET = Fisher’s exact test; NZDep = New Zealand Socioeconomic Deprivation Index; CVD = cardiovascular disease; HTN = hypertension; BMI = body mass index; ACS = acute coronary syndrome; IHD = ischemic heart disease; AMI = acute myocardial infarction; PCI = percutaneous coronary intervention; CABG = coronary artery bypass grafting.

The shading is to show that those particular groupings were all analyzed together.

Table 2. Clinical characteristics on initial presentation: Comparison of the pre- versus post-implementation group

Clinical characteristic	Pre-implementation group (n = 96) n (%)	Post-implementation group (n = 78) n (%)	Test statistic	p-value
Mean HR—beats/min (±SD)^a	70 (±10.5)	71 (±11.4)	t (172) = 0.21	0.65
Mean SBP—mmHg (±SD)^a	128 (±27.7)	127 (±28.7)	t (172) = 0.25	0.79
Mean SpO2—% (±SD)^a	96 (±2.6	97 (±2.8)	t (172) = 0.46	0.64
Mean GRACE 2.0 score (±SD)^b	122 (±38.9)	115 (±34.8)	t (172) = 1.22	0.22
Killip class II–IV^b	8 (8)	6 (7)	χ^2(1) = 0.023	0.87
Infarct type^b—Anterior	38 (39)	35 (45)	χ^2(1) = 0.494	0.48
Inferior	47 (49)	36 (46)	χ^2(1) = 0.135	0.71
Other	5 (5)	4 (5)	FET	1.00
Mimic	6 (7)	3 (4)	FET	0.73

HR = heart rate; SD = standard deviation; t = t-test value; SBP = systolic blood pressure; SpO₂ = percentage of arterial oxygen saturation; GRACE = Global Registry of Acute Coronary Events; χ² = chi-square; FET = Fisher’s exact test.

^a HR, SBP, and SpO₂ recorded on first paramedic assessment in the field.

^b GRACE score, Killip class and infarct type recorded on patient presentation at hospital.

There was no significant difference between groups in the time between patient symptom onset and phoning for ambulance assistance (median time: pre-, 34 minutes, 95% CI 28–101, [IQR 65], versus post-implementation, 26 minutes, 95% CI 22–97, [IQR 48], p = 0.41, two-tailed). The combined median time for this sub-interval was 28 minutes (95% CI 25–92).

Inappropriate Fibrinolysis

Of the 174 patients who received fibrinolysis across both groups, nine patients (5.1%) showed no rise in cardiac biomarkers, six from the pre-implementation group and three from the post-implementation group (6.2% pre- versus 3.8% post-implementation, p = 0.73). Retrospective review of these cases showed that six were inconsistent with myocardial ischemia. Only one patient, from the pre-implementation group, developed complications from fibrinolysis. These nine cases were excluded from analysis of time-to-treatment and patient outcomes. Table 3 provides a diagnostic summary of these cases.

Table 3. Inappropriate fibrinolysis cases and final diagnoses

Case and final diagnosis	N	(%)
Pericarditis	3	(33.3)
Upper respiratory tract infection with nonspecific ST-elevation	1	(11.1)
Cardiomyopathy/apical ballooning syndrome	2	(22.3)
Costochondritis with early repolarisation	1	(11.1)
Upper respiratory tract infection with previous left ventricular aneurysm	1	(11.1)
Intra-cranial hemorrhage with raised intracranial pressure	1	(11.1)
Total	9	(100)

Time to Treatment

Table 4 compares the treatment time intervals between the two groups. On all measures, a significant time improvement was demonstrated in favor of the post-implementation group. This includes a 22-minute reduction in median PTN time (p = 0.007). In addition, the ETN time goal of <30 minutes was achieved in only 28% of pre-implementation cases versus 93% of post-implementation cases, a greater than three-fold increase.

Table 4. Comparison of key treatment time intervals in minutes (median values and interquartile ranges): Pre- versus post-implementation group

	Pre-implementation group (n = 90)			Post-implementation group (n = 75)
Treatment time interval	Median	[95% CI]	IQR	Median	[95% CI]	IQR	p-Value
From pain-to-needle (PTN) time^a	87	[82, 138]	58	65	[60, 110]	41	0.007
From call-to-needle (CTN) time^a	52	[49, 62]	30	36	[33, 42]	10	<0.001
From EMS contact-to-needle (ETN) time^a	37	[34, 45]	30	20	[17, 23]	8	<0.001
From first diagnostic STEMI ECG to needle time^a	24	[24, 31]	24	16	[14, 17]	9	<0.001

EMS = emergency medical services; STEMI = ST-elevation myocardial infarction; ECG = electrocardiogram; CI = confidence interval; IQR = interquartile range. All comparisons between groups were made using Mann-Whitney U tests.

^a Excludes cases of inappropriate fibrinolysis (n = 9).

Accuracy of Paramedic Diagnosis and Protocol Application

In the post-implementation phase, 116 patients were assessed for their suitability to receive fibrinolysis by autonomous paramedics. Seventy-eight of these patients were deemed eligible and received the treatment; the remaining 38 were deemed ineligible and did not receive the treatment. Of the 78 patients who received fibrinolysis, three cases (3.8%) were later identified as inappropriate treatment decisions. The overall accuracy of paramedic diagnosis and protocol application is presented in Table 5.

Table 5. Accuracy of autonomous paramedic determination of patient eligibility for fibrinolysis (post-implementation group, n = 116)

Accuracy value	(%)	[95% CI]
Sensitivity	(96)	[89–99]
Specificity	(92)	[79–98]
Positive predictive value (PPV)	(96)	[89–99]
Negative predictive value (NPV)	(92)	[79–97]

CI = confidence interval.

Among the 38 excluded cases, 28 patients presented with a contraindication to fibrinolysis, the most common being uncontrolled hypertension (14/28), defined as a systolic >180 mmHg and/or a diastolic >110 mmHg. In all these 28 cases, the patient had a hospital diagnosis of STEMI. A further seven patients presented with a STEMI-mimic on ECG, the most common being left ventricular hypertrophy (4/7). The remaining three patients were misdiagnosed.

Patient Clinical Management and Outcomes

Out-of-hospital complications and coronary angiographic findings were similar between both observed groups. This included most patients presenting with single vessel disease. Table 6 summarizes ongoing clinical management and outcomes across both groups. Thirty-day mortality was significantly lower in the post-implementation group. For patients admitted to hospital, those in the post-implementation group experienced a significantly shorter hospital LOS (median = 4 days, IQR = 2) compared to the pre-implementation group (median = 5 days, IQR = 3, p = <0.001). Across both groups, for those patients who underwent rescue PCI (n = 40), the average time from arrival at the referring hospital to balloon inflation was 182 minutes (SD ± 13).

Table 6. Ongoing clinical management and outcomes: Comparison of the pre- versus post-implementation group

Variable	Pre-implementation group (n = 90) n (%)	Post-implementation group (n = 75) n (%)	Test statistic	p-value
Hospital treatment:
Medical management only	22 (24)	10 (13)	χ^2(1) = 3.231	0.07
Rescue PCI	20 (22)	20 (26)	χ^2(1) = 0.440	0.50
Non-urgent PCI	42 (46)	42 (56)	χ^2(1) = 1.426	0.23
PCI procedure complications	4 (5)	4 (5)	FET	1.00
Emergent CABG from CCL	1 (1)	–	–	–
Elective CABG later during admission	5 (6)	4 (5)	FET	1.00
In-hospital major bleed (excluding ICH)	–	–	–	–
ICH up to discharge	2 (2)	–	–	–
In-hospital reinfarction	–	–	–	–
Mortality at 30 days	8 (9)	1 (1)	FET	0.04
Mortality at 6 months	10 (11)	2 (3)	FET	0.06

MI = myocardial infarction; χ² = chi-square; PCI = percutaneous coronary intervention; FET = Fisher’s exact test; IABP = intra-aortic balloon pump; CABG = coronary artery bypass grafting; CCL = cardiac catheterization laboratory; ICH = intracranial hemorrhage; ACS = acute coronary syndrome. Table excludes cases of inappropriate fibrinolysis (n = 9).

Discussion

This study is the first to compare two distinct models of prehospital paramedic-delivered fibrinolysis. Authorization of treatment was the main distinction between the two models, and this study has demonstrated significant time-saving in treatment delivery in favor of the autonomous paramedic model. Compared to the physician-authorised telemetry-based approach utilized in the pre-implementation phase, a 22-minute reduction in median PTN time was achieved in the post-implementation group. Moreover, significantly more patients received fibrinolysis within benchmark timeframes, as stipulated by CSANZ Guidelines (1).

The ETN time goal of <30 minutes was achieved in only 28% of pre-implementation cases versus 93% of post-implementation cases, a greater than three-fold increase. Similarly, there was a significant reduction in 30-day mortality and hospital LOS in favor of the autonomous paramedic model. Of note, there were no significant differences in demographic features, clinical presentation or CVD risk factors between the two observed groups that would account for these improvements in both treatment times and outcomes. In addition, no notable practice changes were identified in either the prehospital or in-hospital domains between the two study phases that may have impacted on our study results.

For those STEMI patients who reside in areas where timely access to a PCI-capable center is not feasible, paramedic-delivered fibrinolysis offers the most viable option for early reperfusion, particularly when compared to the treatment’s provision in-hospital (14–16). This study has demonstrated that a simplified and protocol-based autonomous paramedic model of care can overcome technological limitations, reduce delays to treatment delivery, and result in improved patient outcomes. These results are consistent with previous studies from England, Wales and the Netherlands (8, 10, 11).

Although the risks associated with fibrinolysis are relatively low, when administered to patients with existing contraindications there is an exponential increase in the likelihood of harm (17–19). Therefore, it is crucial that paramedics accurately determine patient eligibility for treatment. Across both groups in our study there were nine inappropriate treatment cases, although only three were part of the autonomous paramedic cohort. Of these three patients, the first two were later diagnosed with pericarditis while the third was diagnosed with apical ballooning syndrome. None developed complications from fibrinolysis.

These results serve in part to validate the study’s new fibrinolysis protocol, which included additional ECG criteria, designed to mitigate the risk of misdiagnosis. In applying this protocol our autonomous paramedics demonstrated highly accurate STEMI diagnosis and treatment decision-making. Moreover, these results are consistent with previous evidence showing that accuracy of paramedic ECG interpretation of STEMI can be comparable to emergency physicians (20, 21).

A potential limitation of our new protocol was the exclusion of a new or presumed new LBBB as a STEMI equivalent. However, in the out-of-hospital setting, this presentation is often difficult to determine as frequently there is little to no access to patient medical records. In addition, these criteria are both less sensitive and specific for the diagnosis (22, 23). Regardless, this study has demonstrated that New Zealand paramedics are able to interpret 12-lead ECGs and determine patient eligibility for fibrinolysis with a high degree of accuracy. These results support broadening the ECG criteria for STEMI within national paramedic treatment protocols, something which has already recently occurred (24). Of note, no patients were encountered in either phases of the study with a new or presumed new LBBB, as identified by either the attending paramedics or the receiving emergency department staff.

Both study cohorts failed to consistently achieve the mandated PTN target time of <60 minutes. A significant improvement in this treatment sub-interval was achieved in the post-implementation group, but the target time of <60 minutes was still only achieved in 44% of cases. For both cohorts, the main contributing delay was failure of patients to promptly seek ambulance assistance upon symptom onset (combined median time delay of 28 minutes, 95% CI 20–92). In the New Zealand context, compared to previous reports, these findings represent a moderate improvement (25). However, further public awareness and education campaigns are needed to improve reperfusion times to within guideline targets.

Approximately 24% of patients (40/165) failed to achieve reperfusion following fibrinolysis and subsequently underwent rescue PCI out of region. For this sub-group, our study showed prolonged delays at the referring hospital before transfer to the CCL and balloon inflation (average door-to-balloon time 182 minutes, SD ± 13). Therefore, considering that revascularisation should ideally occur within 120 minutes from time of symptom onset to achieve myocardial salvage, rescue PCI within our study was universally performed beyond the period of optimal benefit. In addition, current CSANZ guidelines recommend immediate patient transfer to a PCI-capable center for those patients who remain symptomatic and who fail to show >50% ST-segment resolution or new ST depression at 60 minutes post-fibrinolysis (1).

Considering that a large proportion of this 60-minute period will likely occur prior to hospital arrival among those patients treated by paramedics, prompt recognition of failed fibrinolysis, anticipation of time delays and arranged rapid transfer to the CCL are essential elements for ED staff to consider. Moreover, recently developed national guidelines propose direct transfer to a PCI-capable hospital for all patients who receive out-of-hospital fibrinolysis, regardless of whether reperfusion is achieved or not (26). This strategy aims to reduce time delays to rescue PCI and ensures patients are delivered directly to hospitals better equipped to deal with potential complications.

Limitations

This study was not randomized and there were several differences in the fibrinolysis protocol between groups. However, both patient groups were demographically similar and had similar clinical characteristics on initial presentation, so were considered comparable. A cost efficacy analysis was beyond our study’s remit, but it would be of interest to compare costs of the two models of treatment authorization. Pain-to-needle time may be subject to recall bias. However, it was adopted as one of the primary outcome measures so that our study aligned with other similar studies presented in the literature. Moreover, all other relevant treatment timeline metrics are provided for comparison. It may also be argued that our autonomous paramedic participants were more adept at determining patient eligibility for fibrinolysis due to their experience under the previous physician-authorised program. However, using the same paramedics in both arms of the study provided equivalence in all other aspects of clinician participation. Nevertheless, our results may not be truly representative of the wider New Zealand paramedic workforce.

Paramedics in the study’s post-implementation phase may not be considered truly autonomous in their clinical decision-making, due to the protocol requirement of a positive automated interpretation of STEMI. However, this criterion was added to provide decision support. Paramedics made their own decision first based on ECG changes, then examined the automated interpretation. If there was discordance fibrinolysis was withheld. Therefore, an automated interpretation may be likened to the presence or absence of ischemic symptoms, as just one contribution to the fuller picture of patient assessment, which may or may not lead to confirmation of STEMI.

Conclusions

The provision of fibrinolysis in the out-of-hospital setting by autonomous paramedics without physician oversight is a safe and feasible strategy for the treatment of STEMI. Moreover, this approach provides a significant time-saving advantage for treatment delivery compared to the more commonly utilized physician-authorised telemetry-based model, resulting in improved 30-day mortality rates and reduced hospital LOS. Continued public education campaigns which encourage ACS patients to call for ambulance assistance immediately after symptom onset are required, to help bring PTN times within guideline targets. Further investigation into causes of hospital-based delays for urgent referral of STEMI patients to PCI centers in New Zealand is warranted.