Implementation of crashing patient protocol and its association with EMS provider witnessed cardiac arrest incidence: a retrospective cohort study Authors: Timothy Schiller, Stephanie Van Alsten, Joshua Mastenbrook, Michael Bentley, William Fales Published: 2025 ABSTRACT BackgroundEmergency Medical Services personnel encounter patients in extremis that continue to deteriorate to the point of cardiac arrest. EMS personnel in one urban county in Southwest Michigan encounter approximately 250 cardiac arrests a year, some of which are witnessed by EMS personnel. Crashing patient protocols have been developed in other areas in the country with results showing fewer EMS witnessed cardiac arrests. This retrospective before and after analysis aims to compare witnessed cardiac arrest data from 2021 to 2022 following the implementation of our local crashing patient protocol.Main textWe compared Cardiac Arrest Registry to Enhance Survival (CARES) data regarding cardiac arrests witnessed by 911 personnel from year of implementation (2022) and prior year (2021). We performed a chi square analysis to compare these data and set the alpha level at 0.05 to determine statistical significance. For the year 2021, the data revealed that 12.95% of cardiac arrest calls were witnessed by EMS responders, compared with 6.77% of cardiac arrest calls in 2022. This decrease in the proportion of cardiac arrests witnessed by EMS from 2021 to 2022 yielded a p-value of 0.0181, suggesting a statistically significant difference between the two years.ConclusionOur study revealed a significant decrease in the proportion of EMS witnessed cardiac arrests from 2021 to 2022. The Crashing Patient Protocol was initiated in our EMS System in 2022, suggesting there may be an association with this protocol implementation and a decrease in the number of EMS witnessed cardiac arrests.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12245-025-01059-z. Background Emergency Medical Services personnel often encounter patients in extremis, some of whom deteriorate to the point of out of hospital cardiac arrest (OHCA). While EMS-witnessed OHCA has a higher survival rate than OHCA not witnessed by EMS, prognosis remains poor worldwide. As an example, only 11.2% of patients who had an EMS-witnessed cardiac arrest survived to 30 days or hospital discharge in one study out of Singapore [1]. The 2022 CARES Annual Report cites a rate of 17.1% of patients surviving an EMS-witnessed cardiac arrest in the United States [2]. While there are variables that increase survival rates, such as initial rhythm of ventricular fibrillation as demonstrated in one Swedish study, survival rates remain low [3]. As such, there is considerable interest in identifying patients at risk for EMS-witnessed OHCA as this may provide initial insights into potential strategies for avoidance of cardiac arrest altogether. Some studies have attempted to further examine predictors of EMS-witnessed cardiac arrest, perhaps the most well-known of which is the initial OPALs study published in 1999 [4]. This study identified chest pain and dyspnea as important prodromal symptoms prior to cardiac arrest. A more recent study out of South Korea attempted to categorize EMS-witnessed OHCAs according to one or multiple prodromal symptoms to determine the relationship of symptomatology with survival rates, attributing prodromal “cardiac symptoms” to the best chance of survival [5]. While these studies laid important groundwork, they fell short of identifying individual patient-specific factors, limiting applicability to on-scene patient care. Extrapolating to a method focused on individual patient factors, one Australian study categorized EMS-witnessed OHCAs similarly to studies mentioned above but further examined risk according to vital signs and mental status. With this, they provided a potential real-time method of risk stratification for patients on scene [6]. Similarly, a study out of Taiwan reported a correlation between vital sign abnormalities and subsequent cardiac arrest in the prehospital setting [7]. Going one step further, a newer US study described an “EMS Modified Early Warning Score.” This study demonstrated increased risk of cardiac arrest with increasing number of abnormal vital signs including mental status, much like the two studies mentioned previously, but this study also proposed a way to score those abnormal vital signs in a novel way for potential use as a cardiac arrest prediction tool [8]. Finally, one group in Japan began acting in response to elevated cardiac arrest risk by adopting a novel protocol of initiating chest compressions prior to loss of central pulses in a specific subset of patients with other signs of circulatory compromise. They subsequently observed an increased survival rate after this intervention was implemented [9]. To our knowledge, this is the only published study proposing a protocol change prior to cardiac arrest that resulted in a statistically significant survival benefit in the prehospital setting. Although not a published study, an abstract presented as a poster at the 2018 National Association of EMS Physicians (NAEMSP) Annual Meeting by Pinchalk et al. also focused on early treatment initiation aimed at avoidance of cardiac arrest. This abstract proposed a novel “Prehospital Crashing Patient Critical Care Bundle” designed for use in patients with respiratory distress. This group noted a decreased rate of cardiac arrest witnessed after medical contact during the study period, although this difference was not statistically significant [10]. This care bundle gave rise to the Commonwealth of Pennsylvania’s “Crashing Patient/Patient in Extremis” protocol [11], which we used as a model for the protocol in our county. Given the paucity of literature on potential strategies for avoidance of cardiac arrest once an at-risk patient is identified, further studies are needed in order to positively impact survival of critically ill prehospital patients. We aimed to evaluate our “Crashing Patient Protocol” in order to address that gap in literature. The concept of our “Crashing Patient Protocol” is that EMS crews should defer the rapid “load and go” approach in favor of aggressive, on-scene resuscitation. Given the poor prognosis associated with cardiac arrest, any protocol changes that decrease the incidence of cardiac arrest can contribute significantly to improvement in prehospital mortality rates, thus providing an important consideration for EMS medical directors. On a more general level, it is important for EMS systems to follow-up on the impact of all protocol changes to ensure they provide benefit to the public. Main text Study design and setting This was a retrospective cohort study reviewing CARES data. The study took place in an urban county in Southwest Michigan with approximately 260,000 residents and involved three Advanced Life Support ambulance agencies and fourteen fire department based Medical First Responder services. We used CARES data to evaluate our EMS system and cardiac arrest statistics; no patient identifying information was involved. The study was determined to be exempt by our Institutional Review Board. Selection of participants Participants were selected by being over the age of 18 and having had a prehospital non-traumatic cardiac arrest in the county as recorded in our CARES dataset. They were further stratified by whether their cardiac arrest was witnessed by EMS or not. Exclusion criteria included patients under age of 18, cardiac arrest outside of the county, and cardiac arrest not in the prehospital setting. Inclusion and exclusion criteria were based on how CARES collects and reports data. Intervention A new system protocol, “Crashing Patient / Patient in Extremis,” was introduced into the county EMS system and implemented between 2021 and 2022 [12]. An online self-paced training course was created by the office of the medical director. All system EMTs and paramedics completed the online course, which was hosted on a platform routinely used by our system’s prehospital providers for education. The “Crashing Patient / Patient in Extremis” protocol directs EMTs and paramedics to initiate aggressive respiratory and circulatory resuscitation on scene for critically ill patients rather than attempting to move patients to the ambulance for such care. Patient movement can occur once critical stabilizing interventions have been performed. Please see the supplementary file, Crashing Adult State Format, to review the protocol in its entirety [12]. Measures/Outcomes The primary outcome was percentage of cardiac arrests witnessed by EMS in 2021 compared to 2022, after the implementation of the crashing patient protocol. Data from 2021 to 2022 were compared using a Chi-square test to determine if there was a statistically significant difference, defined as p < 0.05. Table 1 reveals non-traumatic prehospital workable cardiac arrest demographics for Kalamazoo County in 2021 and 2022. There were 278 OHCAs in 2021 and 251 in 2022. In 2021, 178/278 (64%) patients were male compared with 150/251 (60%) in 2022. The majority of the patients were Caucasian in both years (72.3% in 2021 vs. 75.3% in 2022). The arrest location was home/residence in 214/278 (77%) cases in 2021 and 202/251 (80.5%) in 2022. The first monitored rhythm was shockable in 54/278 (19.4%) cases in 2021 and in 49/251 (19.5%) cases in 2022 [13]. Table 1Prehospital cardiac arrest characteristics in Kalamazoo County 2021 vs. 2022 [13]Characteristic20212022Cardiac Arrest Cases278251Mean Age (years)62.163.3Sex (male)178 (64%)150 (60%)Race (Caucasian)201 (72.3%)189 (75.3%)Location of Arrest (Home)214 (77%)202 (80.5%)First Monitored Rhythm (shockable)54 (19.4%)49 (19.5%)Sustained ROSC103 (37.1%)80 (31.9%) Table 1Prehospital cardiac arrest characteristics in Kalamazoo County 2021 vs. 2022 [13]Characteristic20212022Cardiac Arrest Cases278251Mean Age (years)62.163.3Sex (male)178 (64%)150 (60%)Race (Caucasian)201 (72.3%)189 (75.3%)Location of Arrest (Home)214 (77%)202 (80.5%)First Monitored Rhythm (shockable)54 (19.4%)49 (19.5%)Sustained ROSC103 (37.1%)80 (31.9%) Prehospital cardiac arrest characteristics in Kalamazoo County 2021 vs. 2022 [13] For the year 2021 our county had 278 OHCAs. Of those, 36 (12.95%) were witnessed by 911 responders. In contrast, 17 (6.77%) of the 251 OHCAs in 2022 were witnessed by 911 responders (Fig. 1) [13]. There is a statistically significant difference, at the alpha = 0.05 level, between the proportion of cardiac arrest calls witnessed by 911 responders in 2021 and 2022 (p = 0.0181). The proportion of cardiac arrests witnessed by 911 responders was greater for the year 2021 compared to 2022. EMS-witnessed cardiac arrest proportion was chosen as a proxy measurement of protocol success due to being readily available in the CARES data set and a lack of another reliable way to determine the number of patients encountered by EMS who were in extremis. Furthermore, we would expect this outcome to result from benefits due to the protocol itself, due to shifts in EMS culture and due to training and education, all of which we wished to capture. Fig. 1Prehospital cardiac arrests within Kalamazoo County 2021 vs. 2022 Prehospital cardiac arrests within Kalamazoo County 2021 vs. 2022 Given that this is a retrospective before and after study, we cannot attribute causality of the implementation of the crashing patient protocol to the decrease in the proportion of EMS witnessed cardiac arrests, but rather an association. It is interesting to note there was a similar decrease in witnessed OHCAs that occurred in Pittsburgh, Pennsylvania, which boasts a population of 100,000 higher than our county, following the implementation of a similar crashing patient protocol. The generalizability of our results is likely limited. EMS systems with a similar patient population, limited provider turnover, system design, education, oversight, and protocol may be better positioned to realize similar results. Our ambulance agency paramedics and/or EMTs enter raw data from the field into each electronic patient care report (ePCR). One section of this report is dedicated to, at minimum, the National EMS Information System (NEMSIS) required fields for cardiac arrest reporting. Providers must select if the arrest was an EMS witnessed arrest. A mis-key or misunderstanding of the phrase could lead to inaccurate compiled data, thereby falsely increasing or decreasing the reported rate of EMS witnessed OHCAs. We did not see significant provider attrition during the study period and therefore would not except to have seen a large swing in the coding accuracy of the field. Conclusions Our study shows that EMS witnessed cardiac arrests decreased after the implementation of the “Crashing Patient / Patient in Extremis” protocol. Our study only involved one county in Southwest Michigan, and further studies should be completed to evaluate external validity. What our study does not tell us is why there was an improvement; we suspect it was due to the implementation of the “Crashing Patient / Patient in Extremis” protocol, but there may be other causes as well. One possible cause of decreased EMS witnessed cardiac arrests in 2022 when compared to EMS witnessed cardiac arrests in 2021 could be related to Covid 19. Data obtained by the Michigan Department of Health and Human Services shows us there were more deaths related to Covid 19 in 2021 when compared to 2022. There were a total of 8,154 deaths due to Covid 19 in 2022 when compared to 15,020 in 2021 [14]. It is possible that the cause for decreased witnessed cardiac arrests by EMS in 2022 was that the Covid 19 virus was less virulent and lead to fewer deaths that year compared to 2021. Our study revealed a statistically significant decrease in the proportion of EMS witnessed cardiac arrests from 2021 to 2022. The Crashing patient protocol was initiated in our EMS System in 2022, suggesting there may be an association with this protocol’s implementation and a decrease in the number of EMS witnessed cardiac arrests. A previous study done in Pittsburgh also showed improvement after implementation of a similar protocol. The state of Michigan has adopted a Crashing Patient protocol statewide as of 2023; future analysis should look at EMS witnessed cardiac arrest rates statewide to gain more insight with a larger patient pool into the association of this protocol change with the rate of EMS witness cardiac arrest. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1