Hi,

I have been tasked with producing some funnel plots that are somewhat different to the ones i am used to producing. Normally i would be plotting something along the lines of:

X-axis: Number of discharges
Y-axis: rate of re-admissions.

However, this time i have been asked to produce the following:

X-axis: number of finished consultant episodes (FCE)
Y-axis: average length of stay (LOS)

The data available is in the format as below

Provider name FCEs Mean LOS AVERAGE LOS
A 86 10.9 4.6
B 271 10.2 4.6
C 59 8.8 4.6
D 197 7.1 4.6
E 173 7.1 4.6


So, i only have access to the number of finished consultant episodes (FCE) for each site, each site's mean length of stay (LOS) and the overall total Mean length of stay (LOS) for all sites together.

I *think* i need to be able to work out the standard error and upper and lower confidence limits for each site so that i can plot these into a funnel plot, but im stuck on how to do this. All, i seem to be able to work out is an overall standard error for the total.

Can anyone point me in the right direction? I've spent the past 8 hours trying to find a solution on-line but all the guidance and on-line tools focus on rates or expected v actual counts. I'm using excel as the basis of the calculations.

Any assistance would be greatly appreciated as i'm now massively overthinking the problem and my head is starting to hurt...

Thanks in advance,

Rob