Last week, I went to the ER for the first time. Thankfully, my injury was nothing major (just required a couple of stitches), but it exposed me to the entire pipeline of emergency care from the moment 911 is called to treatment in the hospital. Based on what I observed, there are an incredible number of inefficiencies in this system, some scarier than others. This article is meant to be a survey of these inefficiencies and a glimpse at what possible solutions could be.
(Disclaimer: This article is solely based on my observations during the ordeal)
The first stage in the pipeline is, of course, the phone call to 911. Calls to 911 are directed to an operator who, upon hearing the issue, transfers the caller to the appropriate agency (the police, the fire department, etc).
In my case, my friend is the one who called 911. After giving the operator his name and the emergency the first time, he was then transferred 3 more times before the closest station was found. Each time, he had to give them his name (which is rather long), describe exactly what happened, and then give them the address. It was only after giving the address that he would be transferred because they were trying to find the closest station from which the ambulance could depart. All of this amounted to about 10 minutes of wasted time which, if an injury was life threatening, could be incredibly harmful. Surely not all calls to 911 are like this, but in industries like healthcare, even edge cases matter.
Solutions to this problem can range from being very technically simple to technically advanced. At the very least, the transfer from the 911 operator to the closest station should be handled by a computer after the operator types out the address. There are plenty of well known algorithms which can quickly find the location of an available resource (i.e an ambulance) which is closest to a destination (i.e the person who is hurt). If we wanted to take a step further, we could even build a system which listens to operator calls and uses Speech-To-Text analysis so that when the 911 operator transfers the call, whatever transcript is generated is passed on to whomever the call is transferred to. This would stop information from being unnecessarily repeated so whoever is on the other end of the line can focus on giving instructions for care until the ambulance arrives.
Once in the ambulance, inefficiencies in the pipeline are not as life-threatening, but they still make life more stressful for both the EMTs and the patient.
While in the ambulance, the EMTs told me how they were called in for a laceration to the hand. The problem is that I had a laceration to the chin, not the hand. That surprised me a lot because while a laceration is a laceration, shouldn’t the EMTs know exactly what to expect when the arrive on scene? According to the EMT who was treating me, that is never the case. On average, there are 4 data transitions between the phone operator and what the EMTs receive. Essentially, what goes on behind the scenes is a huge game of telephone where data gets mixed up, so EMTs are almost never certain what to expect when the arrive on the scene. Accordingly, they expect the worst case such as a seizure or heart attack. While it is always good for EMTs to be prepared for anything, it certainly doesn’t help to make an already stressful job even more stressful by throwing uncertainty into the mix.
The biggest pain point here is those 4 data transitions between the phone operator and the EMT. The information from the operator should go straight to the EMT unadulterated. The easiest fix is to create a platform which allows for that easy communication. Perhaps it could even be a part of the system which is helping route/schedule the ambulance itself so that way all information about a case is kept in the same place.
Of all the inefficiencies, the ones at the hospital are perhaps the least critical, but they are definitely major areas for improvement.
One of the things which stuck out to me was how the nurse would ask be the same basic questions (name, age, etc) that the EMT asked me to enter me into the hospital system. She also had to rely on me to know the dates of my immunizations. Finally, after she had finished collected that information, she told another nurse just how happy she was that she was able to enter a new patient into a system without having any problems. This brought two questions to my mind
- Why doesn’t the EMT give all the information they collected to the hospital staff?
- How antiquated and anti-user is the hospital system that seamless data entry is an occasion to be happy?
The solutions to both of these problems are not technically challenging. The first one just requires setting up a common system which all hospitals and ambulances can use to transfer patient data seamlessly between them. There is already a huge push for nationwide interoperability of electronic health records (EHRs), so this solution is gaining traction. The major problem with its universal adoption is the logistical nightmare of figuring out a common format to use and getting everybody on board with that format. The solution to the second question is perhaps even easier: just build a new, user-friendly system. To that end, countless businesses are already working towards that (see TigerConnect or other related startups).
These are just a few of the inefficiencies which I as a patient noticed as I went through the system. When it comes to emergencies, every inefficiency matters. While it would be great if hospitals and 911 call centers were constantly being updated with the latest technology (both software and hardware), the fact of the matter is that they are often under-funded and under-staffed. How we fix that issue, however, is where tech ends and politics begins.