The Olympic motto is citius, altius, fortius (Latin for “faster, higher, stronger”). Although the 2016 Summer Games are long behind us, isn’t a similar expectation always applied to “Team EMS”? We’re expected to be faster in our response times, higher in our emergency care standards and stronger in our knowledge.
Improving emergency care standards has helped EMS become more integral, effective and essential to a community’s health—but what about response times? Is faster really better? Is it better in all circumstances? Is it better in any circumstance? Is there a cost to a fast(er) response—and what is that cost?
Any discussion on EMS response times should be framed in the context of our mission: the relentless pursuit of optimal prehospital emergency medical care. In carrying out this mission, our responsibilities include the safety of our patients and the public, as well as other EMS professionals, and operating with fiscal accountability and unquestionable ethics.
But how are our EMS systems actually judged by our patients, their families and loved ones, our colleagues in emergency care, and our community leaders? You’ve likely never received a congratulatory phone call for getting the amiodarone or epinephrine dose correct. Far more likely, especially for those in an administrative or leadership position, the communications you get are either hoorays or harpoons for just two issues: how fast or slow EMS arrived on scene, or how nice or rude the EMTs and paramedics were in caring for the patient.
That’s the reality, and it’s important to acknowledge it: We have to practice strong, high-quality emergency medicine, and that sometimes includes being as fast as conditions and safety allow, such as responding to sudden cardiac arrest.
But sudden cardiac arrest response calls are few and far between. Sometimes we actually need to slow down, especially when clinical situations won’t, don’t and can’t change if an ambulance arrives ≥ 8 minutes after a 9-1-1 call.
The < 8-minute response time parameter that nearly all EMTs or paramedics are held to when responding using lights and siren—often putting their own safety at increased risk—dates back to research published in the Journal of the American Medical Association (JAMA) in 1979. Authored by data-driven EMS pioneers, the study indicated that response to out-of-hospital sudden cardiac arrest was distinctly improved if BLS (i.e., CPR) arrived within four minutes of EMS activation and ALS (i.e., defibrillation) arrived within eight minutes.1
Fast forward to 2017 and BLS now provides CPR and defibrillation, and has done so for several years. Consider that emergency medical dispatch routinely directs callers to perform CPR and retrieve AEDs, often from specific locations. Our clinical capabilities to respond have evolved, so why haven’t our expected response times?
Now think about the myriad other EMS treatment modalities beyond cardiac arrest response: It’s obvious we aren’t doing many of the things we did 20 years ago, so why are we applying a response time standard derived only from the sudden cardiac arrest response standard from 1979? Granted, it’s likely the most time-sensitive condition encountered by EMS, but isn’t it a condition encountered in < 1% of responses in nearly all EMS systems?
In 2011, the University of Oklahoma School of Community Medicine’s Department of Emergency Medicine published a white paper using evidence-based research to examine and assess several aspects of EMS system infrastructure for Oklahoma City’s and Tulsa’s EMS Authority (EMSA).
Thomas Blackwell, MD, FACEP, an experienced EMS physician and researcher who has studied response time standards and their clinical impacts, played a key role in authoring the white paper’s chapter on response time. Takeaway points included that key survival impacts aren’t at 8-, 9-, 10- and 11- minute response timeframes. Impacts on major trauma survival strictly related to response time occur only 4–5 minutes following EMS activation.2
From Research to Practice
Many EMS systems incorporate the Medical Priority Dispatch System (MPDS) or some other form of organized, scripted caller interrogation and pre-arrival instructions. But few systems actually index the personnel and apparatus activated, as well as whether red lights and siren (RLS) are used during that response.
Even when MPDS or an analogous platform exists in the emergency communications center, all calls often get the same apparatus configuration and RLS response. (See “No Need for Speed: Response and transport with red lights and siren,” by S. Marshal Isaacs, MD, FACEP; Carla Cash, MD; Osama Antar, MD and Raymond L. Fowler, MD, FACEP, FAEMS.) Sending the same apparatus configuration to all calls and always using RLS isn’t evidence-based or clinically focused, nor is it resource- or safety-conscious.
In November 2013, EMSA took significant steps to take the 2011 white paper research to the street, making several specific operational changes for response times—the time from call pick-up in the multiagency public safety emergency communications center until apparatus arrival on scene, verified by GPS data—for calls within the city limits of both metropolitan Oklahoma City and Tulsa:
- The Priority 1 ambulance response time standard, assigned for suspected life-threatening situations, was lengthened from 8:59 to 10:59;
- The Priority 2 ambulance response time standard, assigned for situations not likely to involve any acute, life-threatening situations, was lengthened from 12:59 to 24:59; and
- The usage of ambulance RLS was discontinued for Priority 2 responses.
After two years of data from ambulance response times was gathered (through October 2015), EMSA began an analysis of the data. Because averages can hide some seriously fast or slow outliers, all numbers discussed in this article will be 90% fractiles, which means that 90 out of 100 EMS responses were faster or equal to the number reported.
In Oklahoma City (see Table 1), although a full two minutes of additional time was allowed for Priority 1 response, the actual impact was increasing the 90% fractile of 11:56 in the 12 months prior to the response time changes to 13:10—a difference of one minute and 14 seconds.
Priority 2 responses, which essentially allowed for a doubling of response time, resulted in increasing the 90% fractile of 12:07 in the 12 months prior to the response time changes to 18:13—a difference of only six minutes and six seconds.
So, why weren’t Oklahoma City EMTs and paramedics taking the full time allowed by the updated response times? Because, in an EMS system with response time performance standards for both Priority 1 and Priority 2 calls, ambulance staffing and deployment always factors a Priority 1 performance. Although sophisticated call analysis software can reliably predict call locations in the near future using historical EMS data, it’s less reliable in predicting if those calls will be Priority 1 or Priority 2.
The same analyses for metropolitan Tulsa, located 100 miles northeast of Oklahoma City, yielded essentially the same numbers and impacts. (See Table 2.)
Trusting the clinical assessments of our EMTs and paramedics can go a long way to ensuring the appropriate use of red lights and siren. Photo courtesy EMSA
A very critical step in analyzing the operational impact of response times in an EMS system is evaluating not only response time, but the type of responding apparatus and/or crew configuration.
We’ve discussed the time involved in getting ambulances to patients, but what about the impact on responding fire departments and the impact on overall scene times?
The computer-aided dispatch (CAD) data that tracks both ambulance and fire department apparatus response and scene times in the same two-year period (Nov. 1, 2015–Oct. 31, 2016) was also carefully calculated. (Note that the data was partially incomplete as this article was written in October 2016.)
Including both Priority 1 and Priority 2 ambulance responses in metropolitan Oklahoma City, the 90% fractile total fire department on-scene time at EMS incidents increased only two minutes and 25 seconds. (See Figure 1.)
Even though fire department-based personnel were on-scene three minutes and 25 seconds before ambulance arrival, the increase in overall scene time was less than that increase due to efficiency in patient care. (See Figure 2.)
That’s a wonderful pearl of understanding for us because it conveys that our EMS professionals, while employed within multiple agencies, function as a true coordinated patient care team. Similar impacts in apparatus and response times were seen in metropolitan Tulsa. (See Figure 3 and Figure 4.)
Also, as a critically important aspect in how our EMS system functions in both metropolitan Oklahoma City and Tulsa, the larger fire departments don’t respond to all 9-1-1 calls for medical help, but instead respond to approximately 60% of such calls.
This rate of fire department response may sound surprising, but consider this: If an EMS system is sending all apparatus (fire engines and ambulance) to every call, how is it possible to send highly trained responders—strategically based in fire stations—to medical calls that are truly time sensitive, such as major trauma, cardiac arrest, strokes, etc.? They aren’t as available or strategically positioned if they’re overwhelmed with responding to less time-sensitive situations.
We’ve gotten wonderful dialogue and cooperation from partner fire departments, specifically their administrative and frontline leadership. When presented with the data, they get it. They understand they can’t be everywhere all the time. They say, “Tell us where we’re medically making a difference and we’re all-in on those calls.”
From a safety perspective, reducing the number of apparatus responding on calls and reducing the use of RLS are inherently safer.
A natural data comparison would be evaluating the impact on reducing emergency vehicle contacts, particularly those inolving injury to EMS personnel or the public. However, with well-trained emergency vehicle operators in our system, we were fortunate to already have historically low emergency vehicle contacts and we can’t honestly say that response time allowance changes and reduction in RLS use have made a statistically significant reduction in contacts.
To put this in full perspective, in the two years of responses we’ve been discussing, 229,667 uses of RLS were eliminated! As overheard by one Oklahoma City citizen, “A man or woman can sleep a bit later in the mornings now that ambulance sirens aren’t as predictable as a daybreak rooster!”
Consider that emergency medical dispatch routinely directs callers to perform CPR and retrieve AEDs. Why are we still responding like it’s 1979? Photo courtesy EMSA
Operational impacts and benefits are important, but can’t be analyzed in a vacuum. Were patients harmed by the changes? Were conditions unforgivably worsened by the minor response time additions?
These are remarkably important questions because the primum non necere (Latin for “first, do no harm”) doctrine applies to operational changes—at least in our eyes. We couldn’t, and wouldn’t, support changing response time allowances if there was clinical harm done in the process of rendering clinical help.
As in most EMS systems, the EMS System for Metropolitan Oklahoma City and Tulsa is getting busier year after year. The total number of 9-1-1 EMS calls in 2016 soared far above 210,000 for the first time.
With such high call volumes, it’s impossible for every patient encounter and all patient documentation to be reviewed by medical oversight. Therefore, we often use surrogate markers to evaluate impact, and think carefully about what makes for a reliable surrogate marker.
Every ambulance in our EMS system is staffed with a minimum of one EMT and one paramedic, so every patient gets assessed by at least one paramedic. The fact is, we trust our EMTs and paramedics, as well as their clinical assessments. (If we can’t and if we don’t, then there would be something fundamentally wrong.) We assume they work hard to perform accurate assessments, but clinical assessments nearly always involve some level of uncertainty. If doubts are present about a patient’s clinical stability, paramedics will trust their instincts and utilize RLS in their transport of the patient to the hospital. That became a very important surrogate marker for clinical impact in these new response times.
We looked at every individual MPDS code used in our EMS system and examined the response configuration (fire and ambulance vs. ambulance only), the response to scene modality (RLS vs. non-RLS), and the number of calls actually answered with patient contact. Then we looked at what happened for each of those patient contacts, particularly if the patient was transported to the hospital—specifically noting if RLS were used.
If doubts are present about a patient’s clinical stability, EMSA paramedics will trust their instincts and utilize RLS in their transport of the patient to the hospital. Photo courtesy EMSA
When deciding on high vs. low priority conditions, we chose a conservative but appropriate number of 10%. That means that if even 10 out of 100 returns to the hospital involve RLS, then we will designate it a Priority 1 condition, sending both fire apparatus and ambulance to the scene using RLS.
For each of our nearly 1,000 MPDS codes, we looked at three years of data in six-month blocks, examining both percentages and actual numbers. If the percent of scene to hospital was ≥ 10%, we looked to see if the number of patients involved was significant.
We defined “significant” as at least one RLS return per month of analysis per major metropolitan area served. For a six-month period, that would translate to a minimum of 12 patients being returned RLS.
We also wanted to be very careful in not reducing RLS at the cost of patient well-being, so we looked further to see if there was a proportional increase in RLS of at least 25% in every individual MPDS code, regardless of actual patient numbers.
These formulas were set before we started our analysis and haven’t changed since. Our reason: We can’t let the results dictate the formulas; we ethically must let the formulas guide our decisions.
Here’s an example of an individual MPDS code analysis: a one-year look in metropolitan Oklahoma City at 06-Charlie-01 (i.e., 06C01, or abnormal breathing not at a perceived life-threatening extent). (See Table 3.)
When comparing pre-November 2013 data to post-November 2013 data, the return to the hospital using RLS went from 4.67% to 4.15%. The takeaway here isn’t that by slowing down the response, the patient’s condition would get better, but rather that by applying the surrogate marker for patient condition using paramedic assessment and choice for RLS during hospital return transport, eliminating RLS responses for this particular MPDS code didn’t cause any patient clinical detriments.
To date, we haven’t found formula-derived reasons to change MPDS-driven Priority 2s established prior to November 2013 to Priority 1s. In fact, we have seen numerous validations that Priority 1s do predictably correlate with higher RLS hospital returns.
A truly unexpected benefits of these response time allowance changes that reduced RLS use is in the positive impacts on the mental health of EMS personnel. Many have spontaneously commented about the stress reduction achieved from eliminating the noise of the siren and the unpredictable reactions RLS create in the driving public.
To be clear, the sense of duty and the focus on arriving directly to the scene of Priority 2 medical incidents remain, just without as much stress getting there. We believe there’s a direct correlation between more accurately using RLS from the scene that results from more accurately using RLS to the scene.
In other words, when RLS response is used only when clinically necessary, there’s likely a reduction of tachycardia and hypertension in EMS providers while en route which may well translate to better abilities to see “the bigger picture” of a patient’s clinical care needs. That can further translate into less use of clinically unnecessary RLS back to the hospital.
Slowing down responses in less time-sensitivesituations for the sake of just doing it isn’t what this article is about. It’s about meeting the major responsibilities of our EMS system: providing optimal out-of-hospital emergency medical care; protecting the safety of our patients, the public, and our EMS professionals; and operating with fiscal accountability and unquestionable ethics.
As an integrated EMS system, we’ve been able to keep the focus on clinical care, continuing to refine our treatment protocols every two months. Reducing the use of RLS as described and sustained through careful, ongoing analysis, supports increasing safety in response without clinically compromising patients.
It’s also important to note that we haven’t reduced ambulance staffing during this time period—and we aren’t looking to do so in the future.
As you contemplate your own EMS system’s response time standards, remember to challenge traditions and seek data-driven answers. Continue to serve patients while not shortchanging the physical safety or mental health for you and your colleagues. This is about excellent care and excellent safety for everyone involved. Although emergency services aren’t Olympic sports, that’s a gold medal performance for Team EMS.
1. Eisenberg MS, Bergner L, Hallstrom A. Cardiac resuscitation in the community. Importance of rapid provision and implications for program planning. JAMA. 1979;241(18):1905–1907.
2. Blackwell TH: Response Time Standards. In Goodloe JM, Thomas SH (Eds.), EMS evidence-based system design white paper for EMSA. University of Oklahoma School of Community Medicine: Tulsa, Okla., pp. 18–29, 2011.