November 9, 2022
From an early age, our schools teach children that they should call 911 if they have an emergency, but only for emergencies. The logic is simple: 911 service is built to handle emergencies, and it works best when non-emergencies are handled by the systems that are designed for them (like primary care or 311). Non-emergency calls often face long wait times for response. And once they do get a response—a fire truck, an ambulance, a transport to the emergency department—it takes up resources that could be used for more serious incidents.
In DC, the medical needs of one out of every four 911 callers likely would have been handled better and faster at an urgent care or primary care clinic. And in some cases, even simple self-care steps can resolve the issue. If DC could reduce the number of these responses, then everyone’s quality of care should increase.
A Right Care, Right Now nurse on duty. (Credit: DC Office of Unified Communications)
In the spring of 2018, DC’s Fire and Emergency Medical Services (Fire & EMS) and Office of Unified Communications (OUC) began a pilot to better address these “low-acuity” calls. By working with community advocates and researchers, they learned that many of these calls were due to one of three root problems:
the caller didn’t know if their situation was an emergency;
the caller did not have a primary care physician or any other way to access the healthcare system;
or the caller did not have transportation to appropriate care options.
With these three causes in mind, Fire & EMS designed a program to connect eligible callers with a nurse who could assess their condition. The nurse would then recommend and facilitate non-emergency care options. The nurse’s options included referring the caller to a primary-care physician or urgent care clinic with walk-in availability. To address transportation barriers, DC revised its Medicaid regulations, so Medicaid could pay for same day urgent transport. This change allowed nurses to book a rideshare to the primary or urgent care facility and have insurance cover the cost. They called the program the Right Care, Right Now nurse triage line.
While cities and insurance companies across the country had begun to offer nurse help lines, DC was the first to add nurses directly into the 911 call loop. Because this approach was unique, DC government wanted to understand if the program had the intended effects. Did it decrease inappropriate emergency service use? And did it increase primary care use? The Lab @ DC had launched the year before, and the agencies relied on us to answer those questions.
When a 911 call comes in, the 911 call taker completes an initial assessment of the severity of the call. When that assessment indicated a call was eligible for the nurse triage line, The Lab @ DC helped build in a step where the call taker’s computer would basically flip a coin. If the coin came up heads, the call taker was only offered the option of business-as-usual emergency services. But, if the coin came up tails, the call taker was also offered the option of connecting the caller with a nurse. This option to connect a call with the nurse becomes our “treatment” group in program evaluation terms. We used this randomized setup to measure the treatment’s effects on ambulance use, emergency department visits, and primary care visits.
Between April 2018 and March 2019, 6,053 calls made to 911 were eligible for the nurse triage line. Of these calls, 3,030 were assigned to the nurse treatment (our treatment group) and 3,023 were assigned to business as usual (our control group). First, we compared ambulance dispatches between the two groups. We found that without the nurse, an ambulance was dispatched 97% of the time, but when a dispatcher could connect a caller with a nurse, dispatches dropped to 56%.
Sometimes paramedics can treat patients on site without needing to transport them to the emergency department. When they do transport a patient, it is more time consuming and costly than only dispatching an ambulance. So, we also looked at how likely callers were to be transported to an emergency department. Again, we found a large difference: 73% of callers in the control group received an ambulance transport compared to 45% of callers connected with the program.
For the 51% of our study group who are Medicaid beneficiaries, we could examine more outcomes. We found that within 24 hours of their call, 30% of the control group was treated in an emergency department for a non-emergent condition versus 25% of the treatment group. Conversely, 8.2% of the treatment group had a primary care visit within 24 hours, compared to only 2.5% of the control group. However, both these improvements faded after six months. These findings suggest that the nurses allow Fire & EMS to not only avoid using ambulances unnecessarily, but to also connect callers to more appropriate care for their needs. But the results also show that keeping that connection to primary care will need more follow up than a single connection through the nurse.
You can review our initial results in more detail in this preprint manuscript. These results are currently under review at a peer reviewed journal and may change based on feedback from reviewers. Beyond these initial findings, this project demonstrates that with the right resources in place, cities can carefully test a variety of innovations to their 911 and emergency medical systems and rigorously measure the results.