It’s been 35 years since the EMS world started talking about quality improvement as opposed to quality assurance. During that time we’ve been preaching the importance of focusing on systems rather than individuals, gathering data and using evidence.
Almost every EMS system has something with the word quality in it: a quality plan, a peer review QI committee or a quality improvement manager. Yet when you ask most EMS leaders what their “quality whatever” has made better, shoulders shrug and the subject changes. Somewhere along our path we seem to have forgotten the improvement part of quality improvement.
Around the same time EMS started talking about QI, Don Berwick, M.D., and some colleagues founded the Institute for Healthcare Improvement. They engaged a group of rock star statisticians from Associates in Process Improvement in Austin, Texas, and adopted their Model for Improvement as the vehicle for making healthcare better across America and the rest of the world. This simple yet powerful model holds the key to making things better.
How the model for improvement works
The first step is to write an AIM statement. Thousands of costly EMS ideas would be derailed if leaders just stopped and asked their team, “What are we trying to accomplish?”
Take my own example. A couple of years ago, some members of my clinical team wanted to change all of our cervical collars to a fancy new brand whose name shall remain anonymous. They excitedly strapped one on me in the day room exclaiming, “See how much better this is!”
When I asked them, “What are you trying to accomplish?” they said, “Better cervical immobilization.” That’s when I asked the second question in the model: “How will we know that a change is an improvement?” They looked at me as if I’d just asked them to calculate the core temperature of the sun using a nail file, a broken mirror and an out-of-juice C-battery.
What is the measure of inadequate spinal immobilization? The first one that comes to mind is the number of patients who were able to move their arms and legs before we cared for them who are now paralyzed due to something that happened during care/transportation. So I asked the clinical manager to run a report counting the number of patients each month who had their spinal cord transected during our care for the past two years. There weren’t any. In fact, no one could remember that happening in the past 20 years. How many complaints have we had from emergency physicians or nurses about inadequate spinal immobilization? None. How about from patients? None.
Management guru Peter Drucker said, “You can’t manage what you can’t measure.” Dr. Edward Deming, the father of performance improvement methods, used to say, “In God we trust, all others must bring data.” If you’re not able to measure (qualitatively or quantitatively) what you’re trying to improve, it’s impossible to know if you’ve made something better.
I ask this question regularly when visiting with EMS systems that want to add rapid sequence induction (RSI) to their protocols: How many patients per month in your system are unable to have their airway managed and suffer a worse outcome as a result? I’ve yet to have a single leader show me a graph with this data. If you can’t answer this question, then you have no business contemplating RSI.
The third question is where you brainstorm ideas for improvement based on your AIM and measurement criteria—but only after you have completed the first two steps! Too many changes in EMS start with this third step, often after folks return from the exhibit hall at the latest EMS conference.
One clue that an idea has skipped the first two questions is any statement that starts with, “We really need to get [fill in the blank].” Our industry is full of really cool solutions looking for problems, so this is the place to brainstorm improvement ideas. You’ll make better progress if you push yourself and your team to come up with at least three, but hopefully more, ideas. Too often we stop after one—or we craft an improvement project around the idea we’re most attached to. My favorite is, “If we did everything on the iPad Mini, the world would be perfect.”
Putting it into practice
Let’s put this model together with a real-world example from AMR’s Ventura County, Calif., operation.
Question 1: What are we trying to accomplish?
Answer: Measurably decrease suffering for the patients we serve.
Question 2: How will we know that a change is an improvement?
Answer: A higher percentage of our patient care reports will show a decrease in suffering.
It’s important to be specific about how, exactly, measurement will happen, so we will measure this by taking a random sample of 100 patient care reports each month and evaluating them for documentation of the nature and severity of suffering (pain, nausea, shortness of breath, anxiety, etc.); an intervention of some kind designed to decrease the suffering (CPAP, morphine, Zofran, etc.); and a post-intervention reassessment of the suffering. The numerator will be the number of patients in the monthly sample where the PCR demonstrates a reduction or elimination of suffering.
Question 3: What changes can we make that will result in improvement?
Answer: In the case of suffering reduction, improvement ideas might include:
- Adding Ondansetron to the medications carried by crews to treat nausea
- Encouraging non-pharmacologic interventions for orthopedic pain like cold compresses, elevation and splinting
- Changing the morphine dosing protocol from 2–4 mg to a weight-based 0.1 mg/kg
- Expanding the use of CPAP beyond pulmonary edema to asthma, pulmonary infections, CO poisoning, etc.
- Provide myth-busting pain management education that deals with perceived drug seekers, abdominal pain and the limited ability of healthcare providers to assess pain severity using anything other than the patient’s own pain rating
Moving on to PDSA
The last part of the Model for Improvement involves a series of Plan, Do, Study, Act (PDSA) tests to learn about the effectiveness of your improvement ideas. For clinical improvements, it is important that only changes supported by the scientific literature be on the list. Improvement ideas that are not supported by science need to be properly researched with full IRB patient protection before they can be considered for use in an EMS system.
The objective of PDSA testing is to learn what really produces beneficial results in your system before anything is implemented. One secret is to start with the smallest, quickest test you can imagine and then do several small, rapid PDSA cycles to quickly learn what works and what does not.
Now, granted, lots of people have written about PDSA cycles over the years and the descriptions can sound a little intimidating. Here’s a just-what-you-need-to-know version:
Plan: Briefly describe what you’re going to try and how you’re going to measure the results, then make a prediction about what will happen. For example, on ambulance 421 B shift, we are going to have them give weight-based morphine to the next patient they have with pain and they will measure the pre-medication and post-medication pain scale. We predict that their 1–10 pain scale will drop at least two points.
Do: Carry out the Plan.
Study: Compare the result with your prediction and capture any ancillary learning. For example: We had a 27-year-old male with a fractured tib-fib from a mountain bike crash. His pain was 7 pre-medication and 2 post-medication. The morphine made him nauseated and the medic thought that it was easy to calculate the dose.
Act: Here you’ll do one of three things:
- Adopt the change as successful
- Adapt the change and try another PDSA
- Abandon the change as unsuccessful
In our example we might decide to adapt the weight-based morphine dosing protocol to include the administration of Ondansetron to manage the nausea, provide pain management myth-busting education and encourage non-pharmacologic interventions for pain.
The concept is to continue doing PDSA cycles until your “degree of belief,” as shown by the results you’re able to produce, indicate that it is time to implement one or more of the changes systemwide. Too often, EMS systems implement interesting ideas without these testing cycles, which is how we got MAST pants, esophageal gastric tube airways and high-dose epinephrine.
This article, originally published in August 2013, has been updated.
