Safety of Nitroglycerin for Field Treatment of STEMI

Nitroglycerin is widely used for treatment of chest pain of suspected cardiac etiology in the field. While there are a few studies from the pre-reperfusion era that demonstrate some benefit of nitroglycerin administration in patients with myocardial infarction, there are no studies establishing its benefit and safety in patients with STEMI in the emergency department or prehospital settings.

It is unclear if the current practice of administration of nitroglycerin in all patients with suspected cardiac chest pain in the field is of benefit. Furthermore, the risk to the patient with STEMI has not been established. The American Heart Association 2010 International Consensus on Cardiopulmonary Resuscitation and Emergency Cardiovascular Care states that "there is insufficient evidence to determine the benefit or harm of initiating nitroglycerin treatment in the prehospital setting or ED." By quantifying the risk of nitroglycerin use for treatment of suspected STEMI in the field and the benefit to pain relief, one can determine if the continued use of this medication for cardiac chest pain in the field is appropriate.

This study will be conducted at Harbor-UCLA Medical Center, UCLA Ronald Reagan Medical Center, and Cedar-Sinai Medical Center. Participants will be identified in the emergency department at all sites. There will be no interventions performed. The study is limited to observational data gathering only.

Los Angeles County established regionalized cardiac care in 2006 with a total of 34 designated STEMI receiving centers (SRC) throughout the county. Harbor-UCLA, UCLA Ronald Reagan Medical Center, and Cedar-Sinai Medical Center are designated SRC. As such, patients with possible STEMI, identified via ECG software interpretation of acute MI in the field, are directed to these centers by EMS.

LA County protocol requires paramedic units to contact the Base Station for all patients with suspected STEMI. There is a dedicated mobile intensive care nurse (MICN) or Base Physician 24-hours per day answering the radio, providing on-line medical direction, and maintaining a log of all patients. Outcome data is collected and available for quality improvement (QI) purposes. From the 2013 QI data, the investigators expect that approximately 70% of these patients will have a culprit lesion on catheterization, confirming a STEMI.

The MICN or research assistant will identify study subjects. A study investigator will review the log on a weekly basis to ensure no patient is missed. Patients will be eligible for enrollment in the study if they are identified by paramedics as having a STEMI and transported to a study site. The patient will be identified by the MICN or research assistant at the time of base contact or arrival to the hospital with a prehospital diagnosis of STEMI. Patients will be excluded if they are under 18 years of age, if the primary complaint is trauma, if they are in cardiac arrest at the time of paramedic arrival or if they are transported to a non-study site.

The primary outcome is frequency of hypotension (SBP<100) on initial ED vital signs after treatment with nitroglycerin for suspected STEMI in the field. The onset of action of sublingual nitroglycerin is 2-5 minutes, with a duration of 60 minutes. Average transport times in LA County are approximately 15 minutes. Therefore, the primary outcome will be measured upon ED arrival, after completion of field care and prior to any ED intervention.

Secondary outcomes will be change in chest pain score for patients treated for STEMI in the field and in-hospital mortality. Chest pain is assessed on a 11-point Numeric Rating Scale (NRS). The initial pain score is assessed and documented by paramedics. The same NRS is used by the triage nurse upon ED arrival. A priori, a decrease of 1.39 will be considered the minimum clinically significant difference (MCSD) based on prior literature. Planned subgroup analysis for these outcomes will be patients requiring PCI for proximal right coronary artery lesion on catheterization.

After patient arrival, the MICN or research assistant will log patient and retain paramedic and base record for the study investigator. The following information will be obtained regarding patient EMS and ED phase: patient sequence number and medical record number, age, gender, race, initial prehospital vital signs (blood pressure, heart rate, pain scale), treatment with nitroglycerin in the field (number of doses), initial triage vital signs (heart rate, blood pressure, pain scale), ECG findings (location of ST elevations if present), disposition to the cath lab (yes/no). On follow up, a study investigator will enter whether percutaneous coronary intervention (PCI) was performed, the location of the culprit lesion(s) for patients treated in the cath lab, and final diagnosis and disposition for all patients. This information will be obtained via chart review and discussion with the cath lab team as required. There will be no follow-up of patients beyond hospital discharge. A second investigator will review a random 10% sample of records for key data elements.

During data entry, all information will be maintained in SherlockMD, a HIPAA-complaint electronic secure database with access only to the principal investigator. At study completion, de-identified data will be exported into a Microsoft Excel spreadsheet (Microsoft Corporation, Redmond WA). Statistical analyses will be performed using SAS 9.4 (SAS Institute, Cary, NC). The study outcomes will be presented as proportions with exact binomial confidence intervals. P-values for univariate associations will be calculated using Chi-square or Fisher's exact test, as appropriate. Adjusted outcome estimates will be determined using multiple logistic regression. A power analysis is not applicable to the primary outcome: proportion of hypotensive episodes related to nitroglycerin use for STEMI in the field, with no comparative group. For the secondary outcome of pain relief, the investigators determined a sample size of 73 patients per group to achieve 90% power to detect the minimum clinically significant difference of 1.39, assuming, base on prior literature, an average pain score of 6.6 and standard deviation of 2.58. Kappa statistic will be used to assess agreement between data abstractors.