By Kevin Waters and Floyd Salazar
In 2017, approximately 110 to 120 patients experienced an out-of-hospital cardiac arrest in our northern Colorado response system. Three survived neurologically intact. At the time, we did not have that data — and that was an important part of the problem.
Poudre Fire Authority (PFA) had always been an active part of the response to cardiac arrest. Our crews responded, started CPR, applied defibrillators and worked alongside UCHealth EMS, the hospital-based ambulance service providing advanced care and transport.
| READ NEXT: Turning data into decisions: How Henderson Fire is scaling QA/QI with AI
But by 2016, PFA was beginning to look more deliberately at its role in the larger EMS system and at how the system was performing as a whole. And around the same time, our shared medical director, Dr. Darren Tremblay, attended a presentation by Dr. Daniel Davis on systems-based cardiac arrest resuscitation. That presentation prompted Dr. Tremblay to look more closely at our own outcomes.
Early steps toward change
In June 2017, PFA created its first EMS battalion chief position. The move placed EMS within the command structure rather than alongside it, establishing authority over EMS operations, training and clinical quality.
Over the following year, we worked with Dr. Tremblay to build a small EMS leadership team that included a captain of EMS. Once that team was in place, the effort was no longer simply a training initiative. It had become a system redesign.
Shared medical direction made it possible to develop one clinical standard across two organizations, but the structure alone was not enough. Both organizations also had to be willing to treat cardiac arrest outcomes as a common responsibility. After all, a patient in cardiac arrest does not experience one organization and then another. The patient experiences one response system. Improving outcomes required us to think about the work the same way.
Creating one clinical standard
In fall 2018, we rebuilt our existing cardiac arrest protocols as a regional algorithmic standard that would apply consistently to every provider involved in the resuscitation. The new protocols went into effect in January 2019.
We relied heavily on work already being done in other high-performing EMS systems rather than trying to create something entirely new.
One principle guided most of the changes: Continuous, high-quality compressions are the foundation of a successful resuscitation, and anything that interrupts them has to be carefully considered.
Additionally, real-time CPR feedback became an expected part of the resuscitation. Compression quality could be evaluated objectively rather than relying primarily on perception.
We redesigned the defibrillation process to protect chest compression fraction. Crews pre-charged the monitor before rhythm checks so that a shock, when indicated, could be delivered quickly and compressions resumed with minimal interruption.
The airway approach changed for the same reason. The priority was oxygenation and ventilation, not the procedure used to achieve them. Endotracheal intubation moved into the background, the i-gel became the default advanced airway for cardiac arrest, and waveform capnography became an expected part of ventilation.
That was probably the most significant departure from how many providers had been trained and practiced up to that point. For years, endotracheal intubation had been viewed as an important marker of paramedic skill. Asking clinicians to use a supraglottic airway as the default required them to reconsider something that had been strongly reinforced throughout their careers. The shift in mindset from intubation to resuscitation was difficult — but it got easier as crews began seeing patients survive.
Reviewing every case
In March 2019, PFA and UCHealth EMS began reviewing 100% of out-of-hospital cardiac arrest resuscitations. We looked at compression rate, depth and fraction; the timing of rhythm checks; whether the monitor had been pre-charged; whether the shock decision was correct; airway management; and medication administration. That level of review surfaced issues that traditional chart review alone would not have identified.
One example involved the way some clinicians were disarming a charged monitor. Changing the energy setting instead of using the monitor’s disarm function meant the next shock could be delivered at the wrong energy. Nobody would have reported that because nobody knew it was happening. It became visible only when we reviewed the resuscitation data itself.
Initially, every case was reviewed face to face with Dr. Tremblay. That required a considerable investment of time, and it was not intended to be the permanent model. Early in the program, however, it helped establish a shared expectation that every resuscitation was worth learning from.
The process eventually developed into a standardized case analysis built directly from the monitor data. Each review includes CPR performance, the ECG, waveform capnography and the timing of key interventions. Producing a complete review takes approximately two hours. That remains a meaningful commitment of time. It has also become one of our most effective tools for showing providers what went well and where there was an opportunity to improve.
The purpose is not to turn every case into an investigation. It is to understand what happened, identify where the individual provider or the system could improve, and carry those lessons into the next resuscitation.
Letting the data shape the work
The protocols and review process were never treated as finished products. As recurring patterns emerged, we made additional changes. For example, we clarified post-return-of-spontaneous-circulation care after reviews identified inconsistent use of epinephrine infusions and missed opportunities for pacing. Later, the data showed that clinicians were generally recognizing when pacing was indicated, but procedural success was lower than expected. That led to regular hands-on monitor training focused on defibrillation, synchronized cardioversion, and pacing.
In 2021, we began participating in the Cardiac Arrest Registry to Enhance Survival (CARES). We entered our 2020 cases ourselves while a more sustainable process was being developed. That gave us a standardized way to track outcomes and, for the first time, something outside our own system to measure against.
Some of the most useful lessons came when the data challenged something we had already decided to do. For example, in 2024, our medical directors introduced a vector-change defibrillation protocol. Later that year, survival declined. When we looked more closely, we found that use of a second set of defibrillation pads had increased by approximately 40%, although only 17 patients had actually met the indications for their use.
Our concern was not the intervention itself. It was that crews were preparing for a possible next step before it was needed, and that additional activity appeared to be distracting from more fundamental parts of the resuscitation. As such, we discontinued the protocol and returned the focus to basic resuscitation quality.
That was an important lesson for us: Quality improvement has to include the willingness to reconsider your own decisions when the results do not support them.
Outcome changes
After implementing these changes, outcomes improved considerably.
| Year | Medical Arrest Resuscitations | Neurologically Intact Survivors | Survival Rate |
| 2017 | ~110 to 120 | 3 | ~2.5 to 2.7% |
| 2018 | 124 | 9 | 7.3% |
| 2019 | 126 | 18 | 14.2% |
| 2020 | 103 | 21 | 20.4% |
| 2021 | 114 | 16 | 14.0% |
| 2022 | 126 | 20 | 15.9% |
| 2023 | 120 | 21 | 17.5% |
| 2024 | 115 | 17 | 14.8% |
| 2025 | 104 | 18 | 17.3% |
Figures reflect medical cardiac arrest resuscitations. Traumatic arrests are excluded throughout.
Neurologically intact survival increased from less than 3% in 2017 to 7.3% in 2018. Following implementation of the regional protocols and full case review process, the survival rate increased to 14.2% in 2019 and reached 20.4% in 2020. Since then, survival has remained in the mid-to-high teens.
Participation in CARES also allows us to place those numbers in a national context rather than only measuring against ourselves. In 2025, our neurologically intact survival rate of 17.3% compared with a national CARES rate of 8.3%.
What other agencies can take from this
Every EMS system is different, but several parts of this work are transferable.
Align medical direction early, then create one standard: Shared medical direction made it much easier for two organizations to develop one clinical approach. Systems with separate medical direction may need to spend more time aligning expectations before they begin rewriting protocols or building a shared quality process.
The standard itself matters just as much. A shared protocol gave providers across both organizations the same clinical framework. Clinical judgment still has an important place, but unnecessary variation makes performance much harder to understand and improve.
Use work that has already been done well: We did not build this program from scratch. Much of the original protocol work was adapted from approaches already being used by experienced EMS physicians and systems elsewhere. That allowed us to focus more of our time on implementation, measurement and improvement. There is no particular value in creating something proprietary when good work already exists.
Review every case: A comprehensive review process allows patterns to emerge. A single issue may not appear significant in isolation. When the same issue appears repeatedly, it may point to a problem with training, equipment, protocol language or system design.
Reviewing every case also changes the relationship an organization has with its outcomes. The numbers stop being something reviewed occasionally and become part of the work itself.
Agencies also need to plan for what the process becomes over time. Face-to-face review of every case may not be sustainable indefinitely, but the process that replaces it still has to be detailed enough to hold a provider’s attention and useful enough to change practice.
Keep the review focused on learning: People need to be able to participate honestly in the process. That requires a just culture approach: shared accountability between the provider and the organization, focused on the next patient rather than on the last mistake. That does not mean avoiding accountability. Providers should be expected to follow established clinical standards, and organizations should be expected to give them clear protocols, effective training, appropriate equipment and systems that support good care.
The question is not only what went wrong, but also what went well. The review process should reinforce strong performance while helping providers and the organization learn from the things that did not go as intended.
It also means being willing to talk about the things nobody wants to bring up. The organized rhythm that got shocked by mistake is exactly the case the process exists to find.
A meaningful quality program will eventually show you something you did not expect. It may identify a training gap, a problem with a process or an initiative that is not producing the intended result.
Focus on the fundamentals: EMS has no shortage of new devices, techniques and ideas. Some will improve care. Others will not. Our largest gains came from doing basic things consistently well: high-quality compressions, timely defibrillation, effective oxygenation and ventilation, familiarity with equipment and careful review of every resuscitation.
Keep learning
The work continues to evolve, and it should. Protocols will change. Equipment will change. New evidence will emerge, and some ideas that appear promising will not work as expected. Our responsibility is to keep measuring, keep learning and keep making the system better for the next patient.
ABOUT THE AUTHORS
Kevin Waters is Senior Director of Prehospital Care at UCHealth EMS in Northern Colorado, overseeing roughly 400 personnel and 60,000 annual responses across Larimer and Weld counties. He has spent nearly four decades in emergency services, with a career focused on patient care, operational performance and the systems that support frontline providers. He previously served as Poudre Fire Authority’s EMS Battalion Chief.
Floyd Salazar is Director of UCHealth EMS in Northern Colorado and previously served as the organization’s Clinical Quality Manager and PFA First EMS captain during the period described in this article. A former fire officer and paramedic, Salazar has spent more than two decades in emergency services, with a career focused on clinical quality and evidence-based practice. He works with regional EMS agencies and medical directors to strengthen prehospital care through quality improvement and protocol development.
