What value do we add to emergency care?

Chris Stein, ECSSA Immediate Past-President

Some time in the future, perhaps not as far as you might think, a familiar series of events is playing out in a large urban EMS system involving a patient, Neo. This is his story.

Neo hadn’t been feeling well the whole day. Nothing major, he told himself, but at his age and with his previous experiences, he didn’t want to leave too much to chance. Better to have it checked out. He took out is smartphone and tapped on the Universal Healthcare app. He told the app what his problem was – “well, today I’ve been feeling more tired than usual and a little light-headed. Also, that headache is back, not as bad as last time but it’s there and I’m a bit worried about it”. He waited while the app connected, and then confirmed that it knew him. Welcome back Neo – please wait while we consider your request today. A minute later Neo was advised that he should wait at home for a medical agent, who would be with him in approximately 20 minutes, and that if anything changed in that time he should tap the big red button. Nothing new here. Better to have this headache checked out.

It had been two decades since the big turn-around in pre-hospital care. It all started with the mass adoption of driverless vehicles and advances in medical expert systems. But perhaps it was less about what technology made possible, and more about how the old system had collapsed under its own weight. The preceding years had seen a gradual increase in systematic dysfunction to the point where it had become impossible to provide anyone with any kind of pre-hospital care. Growing numbers of calls for assistance, mostly low acuity in nature, coupled with a system that could not scale to this demand and seemed more concerned with rapid response to a small subset of emergencies had effectively removed the prospect of care outside of a hospital for almost everyone. It was a perfect storm, and although it took some time for the change to start, it eventually did and it signalled the end of close to a century of what was traditionally known as pre-hospital emergency care.

At first, the focus was on dealing with the low-acuity cases that were choking the system. Call takers and dispatchers were replaced with autonomous agents capable of extracting information and making accurate decisions many times more quickly. Detailed historical data on every caller, who all used their 10G smartphones to access the new system, was available immediately along with their location data. Triage, which constantly evolved and improved thanks to machine learning algorithms, filtered out the low acuity cases which were queued and dispatched by a real-time, demand-sensitive expert system. Because of the decentralized allocation and reallocation model that was constantly improved, also by machine learning, it was possible to give any caller an estimated vehicle arrival time which was accurate to within 5 minutes. The fact that vehicles had no drivers didn’t seem to worry most people. Neither did the fact that a second round of triage and basic patient care decisions were not carried out by a person, but by an expert system trained on many thousands of previous cases.

Despite initial predictions by some that this approach was doomed, and would be a waste of money and time it turned out very quickly to be anything but. Within a few years there were no more human emergency care personnel servicing low acuity calls. There were no call takers and no dispatchers, nor were many managers left. Transport costs shrank with a fleet of vehicles that were always optimally positioned, always available and not affected by shift changes or the effects long working hours. Savings in personnel costs were reinvested in the system, increasing the fleet and resulting in more advanced systems.

While Neo was waiting for his medical agent, he reminisced with his grandfather about how things used to be, a long time ago. He remembered a time when calling for emergency assistance, even for something that was not really an emergency, was like a lottery game. To be honest, he admitted that he liked the human face and interactions that sometimes came to his aid. But the problem was that this quite often didn’t happen – or at other times took so long that he just asked his daughter to come and pick him up and take him to hospital. He quite often wondered what would happen if he was really, really sick or if something different happened like if he were involved in a car accident. Would he still be playing the lottery in those situations? No, he had to admit that he liked the reliability of the new system even if it was a bit distant and cold. It was simple and effective. And anyway, he thought, he remembered times when the people who came to his house, the humans not the agents, were just as distant and cold. Or even told him he was wasting their time because he wasn’t “sick enough”. At least that didn’t happen anymore.

Five minutes before the driverless vehicle arrived at Neo’s house, the medical agent sent him an instant message asking him to wait outside. When the vehicle arrived, Neo waited for the door to open and then climbed slowly into the assessment seat. After sitting in it for a few minutes, the medical agent gave him some feedback – all vital parameters acceptable and nothing new or worrying turned up in any of the scans. The agent recommended a small change in dosage of one of Neo’s many medications and then advised him to stay at home and rest. No hospital – thank goodness. Although the way hospitals work had also changed a lot, and the days of queuing for a day just to see a doctor were over, Neo always preferred it when the agent recommended that he stay at home. He trusted the agent’s judgement – nothing had ever happened previously when he stayed at home. Can you trust something that is not human, he thought as he reflected on his feelings? You can, he supposed. Maybe the agent isn’t a person, but it seems to be thinking of me. That was the whole problem in the old days. The system didn’t think of me, or know me.

Artificial Intelligence (AI) is poised to transform the way we live in the next decades. Will this futuristic scenario become reality? We can’t know – but the aim of telling this story is not to argue about the merits or otherwise of some futuristic scenario. What is important, is that Neo’s story provides us with a lens to look at our present systems through, and to reflect on why it is not impossible that someone like Neo may believe that an EMS system comprised mostly of AI is more human in some ways, and meets his needs more, than one comprised of human ‘care’ providers.

The deficiencies in current South African EMS systems are fundamental, run deep and should not be accepted anymore. There is a belief that these systems can be propped up indefinitely. This is incorrect, and just as in the story, a time will come when a breaking point (or perhaps a turning point) is reached and what we know currently as EMS will collapse under its own weight and cease to be viable anymore. Changing anything complex is daunting, but must start somewhere. Four fundamental questions, and sub-questions, can be considered as a possible starting point for change:

1. Where did we come from?
   • When we look in the mirror what do we see?
   • Are we authentic or imported?
   • How does the country’s history affect our purpose?
   • What is the EMS history of South Africa?
2. Do we save lives?
   • What do we do 80% of the time and how should a system support us in doing this?
   • What is my value proposition from a patient’s perspective?
   • Why do most people complain about EMS and what do they complain about?
3. Where will we go?
   • Can we say that we understand anything about our patients other than some pathophysiology?
   • What is our patient’s health seeking behaviour? Culture? Needs?
   • What is care and how do patients see this?
   • What is a South African EMS?
4. Who will take us there?
   • Would we recognize our leaders by their acts and values alone?
   • Where will our leaders come from and how will they learn what they need to know?
   • Leaders are born and made at the same time, how do we make ours?