Projects for Manchester
Joseph Anthony Connor
Pain :
Who really knows where and when people attend a service and what they think of it after a referral?
Potential UX:
People who care about the NHS download an app that shows them where local services are available.
It asks for permission to use a phones location to note when they have attended a service following a referral
After a user accesses a service they are automatically prompted to rate it.
Use of the data:
No user identifiable data is released under any circumstances
Resources potentially needed:
Angular JS, Ionic, React Native, API design, GIS data visualisation, app testing, api design, Jason, DX data sharing, data scraping
There will never be enough healthcare workers staff to listen to people when they are in crisis. For example. 1 in 4 people have poor mental health. If 1 in 1000 were to move into very poor mental health each member of the entire NHS would need to take a case load of 6 people each.
Active listening helps , and has been shown to decrease mental health admissions. It is believed that if people were prompted to have regular contact those who are good active listeners fewer people would fall into poor mental health
Potential UX:
An app sits on a phone, it monitors speech and looks for good active listening signals. The app lets the user know who are good active listeners they may wish to turn to in a crisis.
Data usage:
All the analysis would take place on the phone and no data would leave a handset
Resources potentially needed:
Angular JS, Ionic, React Native, Voice / speech signal processing, UX design for mobile phones
______
michae...@doctors.org.uk
Harry Coenen
Fascinating question!
There is some work on that, usually breaking down on the quality of the data, even on how different A&E departments clean up at the end of a shift (that is closing data recording of people who already left, when they didn’t do so at the time of leaving).
A friend of mine and me did some analysis using the full England HES dataset and came up with an interesting ‘relation’.
All people go to A&E, they get admitted to hospital (or not), and they die in hospital (or not).
You take the admission time in A&E (we used as a proxy for that the whole waiting time in A&E, as admission records didn’t have time (yet)) and looked at the proportion that died.
We expected that serious cases would get admitted fast, and would die.
But we found that the proportion that died was higher the longer they had waited, with a real peak at the four hour target.
As a demagogic narrative: People go to A&E, are triaged badly and neglected, until the system flags up the approaching 4-hour target, at which moment they are rushed in a bed, to be forgotten about again, until it’s far too late to do something, and they die.
Most people who we asked just said it couldn’t be statistically significant (remarkably they knew this without having to do any further analysis).
We asked around about the practices at a local A&E department (just in case we would be rushed there some day ourselves), and found that they actually admitted earlier, but only closed the case on the system when the 4 hour target was flagged up, or when someone had time to clean up some backlog on administrative A&E duties.
But surely that cannot be the case everywhere else, or can it?
We didn’t bang any tables as other work needed attention (there’s no time to really do decent analytical jobs as an information analyst within the NHS, as there’s always a backlog of inconsequential (board-)reports that need to be churned out, so no time to investigate people potentially dying (in droves) or not.
I won’t come to Manchester however, now abroad, at least temporarily. But if there is a hospital that would allow going to the bottom of this (and other questions), I’m more than happy to help out, even for free (within reason).
I could dream up hundreds of interesting questions if individualised records are available for analytics (don’t need to be identifiable, just properly cleansed and pseudonymised so you can link with other records like admissions)
Regards
Harry Coenen
--
You received this message because you are subscribed to the Google Groups "nhshackday" group.
To unsubscribe from this group and stop receiving emails from it, send an email to nhshackday+...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Anne Marie Cunningham
To unsubscribe from this group and stop receiving emails from it, send an email to nhshackday+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
--
You received this message because you are subscribed to the Google Groups "nhshackday" group.
To unsubscribe from this group and stop receiving emails from it, send an email to nhshackday+unsubscribe@googlegroups.com.
