EMS policy change and QI: Plan, Do, Study, Act
Consider policy changes through the lens of quality improvement science before implementing new strategies in patient care
While perusing the colorful new EMS equipment in the exhibit hall at a recent EMS conference, I passed an old friend, an EMS medical director for a moderate-sized EMS system. He said, “Hey Mike. We are finally adding RSI and ketamine to our protocols!”
The rest of the conversation went something like this:
Me: “That’s great. What are you looking at, measuring to see if they work?”
Friend: “We are doing 100 percent chart review of all RSIs to make sure that we don’t paralyze folks and then miss the tube. And we are doing 100 percent chart review of all ketamine uses looking for nausea, vomiting, hallucinations, headache, unintended coma, etc.”
Insert awkward pause.
Me: “Those sound like good safety check balancing measures. What I was asking about was measures to see if the improvement you were hoping to create with RSI and ketamine actually happened or not?”
Insert longer, more awkward pause.
Friend: “So what do you think about the new quarterback for the Broncos?”
I’ve lost count of how many conversations I’ve had with smart, well-intentioned EMS clinical leaders over the last few years. I’ve learned a couple of things.
First, asking questions about improvement related to something that’s already been implemented is a quick and easy wet blanket conversation-killer. Second, somehow, our industry has not figured out that protocol changes are improvement interventions and should probably be thought of with a quality improvement mindset. At least that’s my theory.
Clinical practice in most EMS systems is guided by a set of medical protocols. If you ask leaders, “What exactly are protocols and why do we need them?” you’ll get answers such as:
- Protocols are our system’s practice of medicine in written form. They are written based on the best scientific evidence available about how a variety of illness and injury conditions are best cared for.
- If our medical director were on scene taking care of a patient, there are diagnostic tests she would use and then treatments she would provide based on what was going on with the patient. We view all of the calls we run as patients of our medical director so our protocols are designed to as closely as possible replicate the care that would be provided by Dr. Medical Director if she or he were there providing care directly.
- Without protocols and the expectation that they are followed, we could not reliably provide the best care for every patient. There would be too much variation, which means some patients would have good outcomes, while others would not.
Most EMS systems regularly update their protocols. Some do it annually, some monthly and others whenever their clinical leaders come home from a conference. Many have a process of regularly reading emerging scientific literature for protocol change ideas, others benchmark (a fancy term for swiping) ideas from cutting edge EMS systems around the world, and a few are inspired by careful review of their clinical performance data, looking for gaps between the quality of care they would like to provide and what they actually provide.
Use the Model for Improvement to evaluate policy change
So what would it look like if the addition of RSI were viewed through the lens of quality improvement?
The first thing we’d need is an aim statement that answers the question, “What are we trying to accomplish?” One potential aim statement is, “Improve the percentage of patients at high risk for aspiration or impending airway compromise who have their airway successfully managed (without aspiration and while maintaining oxygenation) by December of 2018.”
The next thing we’d need is some baseline data to help us answer the question, “How will we know that change is an improvement?” The outcome measure for this could be something like, “The denominator is all patients who are conscious who are at high risk for aspiration or impending airway compromise and the numerator is the number of these people that remain free from aspiration and stay adequately oxygenated (or at least don’t have hypoxia exacerbated by treatment) during the course of their care.
My bias is that if a system does not have this baseline data with a percentage low enough that there is room for improvement, then they have no business introducing RSI.
The third part of the Model for Improvement is, “What changes can we make that will result in improvement?” Here’s where adding RSI to the protocols should be on the list of options. What other things can you think of that might produce the desired result for these patients?
The next thing to do is to make sure that the changes you’d like to test are backed up by evidence in the clinical literature. If your theories about how to make things better have the potential to put patients at risk and have never been tested through research, it’s not ethical to try them on patients without conducting a proper research study that’s IRB approved first.
Test new policies with the Plan-Do-Study-Act model
If there’s good evidence that the thing you’d like to try is safe and effective, then what you do next is design the smallest, fastest test of the idea that you can conceive to start learning how the idea works in your actual system. These tests should be conducted using the Plan-Do-Study-Act (PDSA) part of the Model for Improvement:
- Plan – For example, your plan might be to train and supply one paramedic who works at a busy station in the use of induction and paralytic agents along with reduction of peri-intubation hypoxia and aspiration risk.
- Do – Then have this paramedic use the procedure on the next appropriate patient they care for. You’ll want to measure outcome data. In this case, was aspiration and induced hypoxia prevented? You’ll want to have a process measure or two that might include the correct dosing and timing of medication administration and protocol compliance. In this case, you’ll also want a balancing measure or two like oxygen saturation levels throughout the process.
- Study – Once this first trial of the protocol is complete, you’ll want to compare what happened with your predictions of what would happen, the Study part of PDSA.
- Act – Then you’ll decide whether or not to adopt this practice into a larger PDSA, adapt it into a trial of delayed sequence intubation or abandon the idea all together.
My hope is that many of you reading this article will start thinking of protocol changes through the lens of improvement science. For the folks that don’t, I hope that you’ll at least feel a little guilty that you’re changing things without having any idea if the result is an improvement.