4 ms·

I think you've hit the nail on the head there. A lot of GDS's mandate revolved around modern agile techniques and just doing it. This doesn't exactly mesh ver

by adwf 12y ago

I think you've hit the nail on the head there. A lot of GDS's mandate revolved around modern agile techniques and just doing it. This doesn't exactly mesh very well with traditional civil service slow bureaucracy. The civil service tend to like getting every bloody stakeholder involved and bury you in meetings upto your eyeballs. And then have meetings about the meetings. And meetings about meetings about the meetings... That's how we wasted £10bn on the NHS IT system with little to show for it.

cssmoo 12y ago
The NHS projects are a little more complicated than gov.uk front end implementation. Have you seen any of the care pathway workflows? We're talking 128 page documents integrating police, social services etc as well, just for a birth. They're also all or nothing and have reliability requirements. You can't just agile this stuff; it's space shuttle up front design from day one. One fuck up in one workflow instance and someone misses that imaging referral and then that person goes from stage 1 to metastatic cancer. How do you assign an appropriate cost vs risk trade off to that? That may make it too complicated to complete to be honest but where do you go from there? You shelve it with a shit ton of bad PR and pay up the people who told you it's not happening.
- grey-area 12y ago
  In my opinion the only way they could have made it work was to start small and iterate. Start with a health record which was: NHS no, name, dob and make sure everyone is using it first, then start to build APIs which let people add data from the various services to that record and control access etc. You simply can't design a system from the top down to deal with something as complex and multi-disciplinary as a birth or a lifetime of health records, in fact no one system could probably cope with all those demands, even if it only catered to a few medical specialties - even just between GPs, Surgeons, Anaesthetists and Dentists there are huge gulfs in requirements, let alone bringing in other public services like the police. On top of that people are happy with existing systems and will resist change to some centralised system if you try a big bang approach. So I disagree that it is all or nothing, or even could feasibly be done as an all or nothing project.
- LeonigMig 12y ago
  The space shuttle programme was basically a horrible disaster.
celticninja 12y ago
exactly, the gov.uk team could have involved every department, team and committe and shipped a late, bloated useless product. The issue for the govt.people is they assume it would have been better if they had been involved, but they would have looked at it from an insiders perspective and not from the customers point of view. the result would have been a site useable by people with good knowledge of their area of government (although useless to them as they dont need the info) and useless to the end user who actually needs to use the site.

by adwf 12y ago

cssmoo 12y ago

The NHS projects are a little more complicated than gov.uk front end implementation. Have you seen any of the care pathway workflows? We're talking 128 page documents integrating police, social services etc as well, just for a birth. They're also all or nothing and have reliability requirements. You can't just agile this stuff; it's space shuttle up front design from day one. One fuck up in one workflow instance and someone misses that imaging referral and then that person goes from stage 1 to metastatic cancer. How do you assign an appropriate cost vs risk trade off to that? That may make it too complicated to complete to be honest but where do you go from there? You shelve it with a shit ton of bad PR and pay up the people who told you it's not happening.

grey-area 12y ago
In my opinion the only way they could have made it work was to start small and iterate. Start with a health record which was: NHS no, name, dob and make sure everyone is using it first, then start to build APIs which let people add data from the various services to that record and control access etc. You simply can't design a system from the top down to deal with something as complex and multi-disciplinary as a birth or a lifetime of health records, in fact no one system could probably cope with all those demands, even if it only catered to a few medical specialties - even just between GPs, Surgeons, Anaesthetists and Dentists there are huge gulfs in requirements, let alone bringing in other public services like the police. On top of that people are happy with existing systems and will resist change to some centralised system if you try a big bang approach. So I disagree that it is all or nothing, or even could feasibly be done as an all or nothing project.
LeonigMig 12y ago
The space shuttle programme was basically a horrible disaster.

celticninja 12y ago

exactly, the gov.uk team could have involved every department, team and committe and shipped a late, bloated useless product. The issue for the govt.people is they assume it would have been better if they had been involved, but they would have looked at it from an insiders perspective and not from the customers point of view. the result would have been a site useable by people with good knowledge of their area of government (although useless to them as they dont need the info) and useless to the end user who actually needs to use the site.