Do you mention Little’s Law? I think it would be useful to get across the idea that for simple queues the long-run average number of patients in the system L = ρ / (1 - ρ), where ρ is the utilisation of the service. As ρ → 1 the queue “blows up”...
Two scatterplots side by side. The chart on the left shows the relationship - or - rather - the lack thereof - between the number of A&E attendances per day and each day's four-hour compliance. The chart on the right shows the (quite strong) relationship between the level of crowding in the A&E each day and each day's four-hour compliance.
It was good to see @samfr.bsky.social say today that "the main reason for lengthy A&E waiting times is the lack of hospital beds for people who need admitting, which gums up the whole system as they have to keep being treated by emergency staff."
samf.substack.com/p/optimism-o...
#rstats #ggplot2
the fallacy of redirection fixing all ills
I think we're getting close to the point when @whitefleece.bsky.social and I are ready to launch our #KnowYourFlow and #TriptychMetrics slogans into the World of Unscheduled Care!
If there’s merit in my idea of simplifying patient flow so that it resembles a puzzle with six numbers, one question worth asking is: “If it’s as simple as this, why aren’t we already doing it?”
1/14
It's approximate, but the excess deaths formula is that roughly 1.4% of patients who spend longer than 12 hours in A&E will die needlessly. So for Scotland in the year ending March 2025, that'd be about 1,100 excess deaths. @policyskeptic.bsky.social has a better command of this.
@samfr.bsky.social
A 'cruel twist of irony' thing occurred to me about my purple scatterplot: A lot of patients will end up featuring on *both* axes. They'll experience a >12-hour stay in A&E and then, later in their patient journey, they'll become a delayed discharge.
@policyskeptic.bsky.social
@samfr.bsky.social
The number of patients spending longer than 12 hours in Scottish A&Es began to increase sharply in the summer of 2021. The increase continued through 2022, and it’s been averaging about 5,700 a month since the beginning of 2023.
#rstats #ggplot2
1/5
I’m keen to know if I’m just stating something that’s obvious here, or whether I’ve stumbled across a correlation that might be useful in our efforts to understand patient flow…
1/15
I stumbled across this short video clip the other day – this is Scott Morrison (*not* the Scott Morrison who used to be the Prime Minister of Australia!) talking to Jo Howes about how we should try to turn problems into puzzles.
1/15
www.youtube.com/watch?v=djUA...
I'll be telling the story of the Red Curtains of Doom at the EdinbR meetup on Thu 17th April...
This chart is an example of what happens when a data analyst tries to pack *everything* into a chart as if they’re trying to win a gold medal at the DataViz Olympics. I’ll try to unpack it element by element.
1/12
#rstats #ggplot2
The idea that the problem can be tackled by actions everywhere else (to reduce demand) has been tested to death over the last decade of bad policy and has utterly failed...
In a very very good way
Crowding down = performance up
Totally agree. Looks great.
The RIE data from this week is very interesting.....
@whitefleece.bsky.social If we want to get people to take ED crowding more seriously, maybe we should try to make the *depictions* of crowding a bit easier to look at!
Agreed. The 'theory of flow' or 'the law of flow'? Would suggest 'the law of positive flow' but that goes against the argument that flow is binary 🤔
You are always going to need to explain it so why bother in the name 🤔
Maybe just the 'the barbershop theorum'?
A scatterplot showing the relationship between the average length of stay so far in an Acute Medical Unit (AMU) and the average length of stay experienced in the Emergency Department by patients bound for the AMU. The relationship is a strong one: r = 0.74.
Maybe we should start measuring the average AMU length of stay so far if we want to understand exit block in full-to-the-brim general hospitals: neilpettinger.substack.com/p/never-mind...
Here are five reasons why we’re failing to improve patient flow in the NHS. There are probably more, but these are the reasons that are prominent in my mind at the moment.
(1/7)
Appreciate it 👍
How have we got ourselves into a position where we’re mainly publishing data to passive decision-makers? And can we somehow re-establish a degree of interactivity?
neilpettinger.substack.com/p/healthcare...
Message me and we can chat over coffee ☕
Little less clinical at the moment so have more time
Plenty to fill you in on too....
Hands off!!!
