The safety and efficacy of implementing the PreHEART decision support tool by the Emergency Medical Service for patients with undifferentiated chest pain: a prospective randomized open blinded end-point study (PreHeart-3 study)

Every year, more than 100,000 patients with chest pain are evaluated by the
emergency medical services (EMS) in the Netherlands. These patients undergo
triage for potentially life-threatening conditions for which early diagnosis
and treatment are essential. A major component of this triage consists of an
electrocardiogram (ECG) to identify individuals with and ST-elevation
myocardial infarction (STEMI). Notably, when a STEMI is identified, the patient
will be conveyed to the nearest hospital with percutaneous coronary
intervention (PCI) facilities for treatment of the culprit lesion. This care
path has been established and has demonstrated efficacy and cost-effectiveness.
However, STEMI only represents a minority of the cases presenting with acute
chest pain (7%), and approximately half of the patients with a myocardial
infarction will not show ST-elevation on their ECG (Non-STEMI or NSTEMI).

On the other hand, EMS contact often results in patient conveyance to the
emergency department (ED), but ambulance care can alternatively result in
patients not being conveyed at all. Non-conveyance is defined as *an ambulance
deployment as appropriate, where the patient after examination and/or treatment
on-scene does not require transport with medical personnel and equipment to the
healthcare facility*. Non-conveyance includes referral to the general
practitioner (GP - huisarts) and it can be due mainly to two reasons: a
patient*s refusal or the professional decision of EMS personnel. Often,
non-conveyance is a combination of these two categories. The current priority
to conduct research on non-conveyance is reflected on the Dutch National
Pre-hospital Research Agenda for EMS 2014-2018. For ambulance professionals,
the (non)conveyance decision-making process is challenging and multifactorial.
Competencies needed to decide on non-conveyance are marginally described, and
there is a limited number of standardized support tools available for general
and specific non-conveyance populations. This may compromise patient-safety.

Currently, the majority of patients with chest complaints (>75%) are conveyed
to the nearest ED for further evaluation, while acute care evaluation is an
increasing burden in the health care system. Overall, less than 10% of patients
hospitalized for chest pain ultimately receive the diagnosis of myocardial
infarction or other severe pathology. Moreover, at least 45% of the patients
transported to the ED are discharged after evaluation within hours without a
life-threatening condition (unnecessary ED visit)6 and approximately 6% are
further (re)transported to a hospital with PCI facilities for definitive
interventional care. Hence, the sequential pre-hospital and hospital approach
for patients with undifferentiated chest discomfort is probably inefficient for
a significant proportion of them, and there is much room for improvement.

A promising clinical tool has been the HEART risk score, which is an acronym
indicating the evaluation of History of the chest discomfort, ECG findings,
patient Age, the presence of cardiovascular Risk factors, and levels of
Troponin I. The HEART risk score has been clinically validated in multicenter
collaborations demonstrating a capacity to exclude short-term major adverse
cardiovascular events (MACE) with >98% of confidence. But notably, the utility
of the HEART score has been mostly evaluated in the hospital setting (i.e. the
ED), while only three studies have evaluated it in the pre-hospital setting.
Recently, we have validated an adaptation of this tool for prehospital
evaluation by EMS, named the preHEART risk score. This preHEART score, has
demonstrated a negative predictive (NPV) value of 99.4% for the occurrence of
MACE at 3 days in patients deemed as very-low risk.
It is still unknown whether prehospital implementation of the preHEART score as
a decision support tool for EMS to select a dedicated care path for patients
with undifferentiated chest pain is not only at least as safe but also provides
more efficient care as compared to current usual care.

Primary Objective:
1. To determine the non-inferiority of pre-hospital risk stratification and
care path selection through the preHEART algorithm as compared to usual care
with respect to the occurrence of unexpected serious adverse events (USAE) in
patients evaluated by EMS for undifferentiated chest pain.

Secondary Objective(s):

1. To determine the non-inferiority of pre-hospital risk stratification and
care path selection through the preHEART algorithm as compared to usual care
with respect to the occurrence of unexpected major adverse cardiovascular
events (UMACE) in patients evaluated by EMS for undifferentiated chest pain.
2. To determine the impact of pre-hospital risk stratification and care path
selection through the preHEART algorithm as compared to usual care in terms of
cost/resource utilization for the healthcare system in patients evaluated by
EMS for undifferentiated chest pain.
3. To determine the impact of pre-hospital risk stratification and care path
selection through the preHEART algorithm as compared to usual care, on patient
quality of life (QoL) of patients evaluated by EMS for undifferentiated chest
pain.

Due to the nature of the comparison, this study will be a pragmatic,
prospectively randomized, open blinded end-point (PROBE), non-inferiority
trial. This design addresses the problem of not being able to blind the EMS
personnel executing the study.
Patients requesting EMS for complaints of chest pain will be screened for in-
and exclusion criteria and if suitable, randomized in a 1 to 1 fashion to
either receive usual care (i.e. decision of (non)conveyance to the nearest ED
for further evaluation without support score) or undergo risk stratification
and (non)conveyance decision (dedicated care paths) with the preHEART score
algorithm (see Figure below).

All randomized patients will undergo follow-up at 3 days, 30 days, 6 months and
1 year.
The PreHEART Consortium
This project entails a continuum of health care collaborators. Therefore, it
will be carried out through a consortium (the preHEART Consortium) structure
involving prehospital (EMS), hospital (academic and regional hospitals with and
without PCI facilities), general practice institutions (huisartspraktijken) and
medical science experts (clinical, outcomes expertise, health economics, and
patient participation) to ensure seamless flow of the project and patients
through the areas needed to answer our research question.

The PreHEART consortium covers the full path of health care for chest pain and
follows the clinical development proposed by Hartnet Noord Nederland. The three
Northern Durch provinces will be involved. Two ambulance services (UMCG
Ambulancezorg and Kijlstra Fryslan) with a total of 70 ambulances, providing
emergency care in this region are key partners in this project and will be
recruiting the intended sample size (5150 patients). Furthermore, ten hospitals
are involved: the academic medical center (UMCG) and 9 regional hospitals. All
hospitals have an ED and coronary care, while three also have onsite PCI
facilities. The involvement of specialists from the two EMS, ten hospitals and
three general practices ensure that we will be able to recruit patients and
obtain accurate follow-up on their health care consumption. In particular, the
three general practices in the consortium spread through the area to provide
robust data and insights on patients referred to the GP under the conditions of
this proposal.
The recently concluded preHEART3 study (Sagel et al. 2019, submitted) provides
a relevant background for the feasibility and efficacy of the proposed
intervention. The results of that study highlighted the benefit expected from
the establishment of a working consortium as described in this subsection. All
parties have agreed to participate in the consortium and to perform the study
to achieve the stated goal.

Subjects with undifferentiated chest pain will be randomized between two
different care arms. In the intervention arm, subjects will be evaluated with
additional information provided by a support tool to reach a (non-) conveyance
decision and in the control arm, the decision is based without this support
tool. No treatment elements are added to these programs for research purposes.
In the intervention arm (preHEART algorithm arm), (non-)conveyance decisions
will be made based on the decision support tool (the preHEART risk score)
applied by the ambulance personnel.

The preHEART score has been developed and validated by our group based on the
original HEART score. The preHEART score is based on the scoring of 5 domains
with a possible range from 0 to 10 points (see Methods - Study Procedures).
Based on the obtained preHEART score, patients will be stratified into 3 groups
and a dedicated care path will be followed for their management:
• Very-low risk group (* 38%*) (preHEART score 0-3 points): patients will be
informed about their very low risk assessment and it will be suggested not to
be conveyed to the ED. If the patient agrees, he/she will not be transported to
the ED and will be advised to make an appointment to see their primary care
physician (huisarts) for evaluation of their symptomatology in an elective
setting within 24 hours. Such evaluation will also contemplate non-cardiac
etiologies of the undifferentiated chest discomfort presented.
• Intermediate risk group (* 55%*) (preHEART score 4-7 points): patients will
be transported to the nearest ED for further evaluation.
• Very-high risk group: (* 7%*) (preHEART score 8-10 points): patients will be
directly conveyed to the ED of a hospital with PCI capabilities (not
necessarily the nearest) for further evaluation.
* Of note: the approximate percentages are based on our prior observational
data available in the 75% of patients transported to the hospital.

Participation in the proposed study conveys a very low risk of adverse clinical
events due to misdiagnosis of a serious condition in patients stratified as low
preHEART risk. At the same time, these patients may benefit from the being
reassured and allowed to remain at their residence, as well as from having to
undergo unnecessary diagnostic testing that potentially carries discomfort and
risk of false positive findings.
Patients with a high preHEART risk may carry the burden of anticipation of
unfavorable outcomes based on their initial trial tag. However, the benefit is
also expected from a prompter recognition and treatment of ACS. This is paired
with less unnecessary diagnostics (due to an already high probability of the
condition) and avoidance of unnecessary transports between hospitals (as they
will be delivered in a facility with PCI capabilities). An additional overall
burden for all participants will be the need to fill in an informed consent
form and the follow-up questionnaires at 3-time points after the event.

There will be no direct compensation for participation in the PreHEART3 study.
However, when a patient is not transported to a hospital they do not have to
pay from their "own risk" contribution, as regulated by Dutch law. This could
vary between 385 till 850 euro.