Weeknotes 2023-04-03

A very short week consisting of only one day at work between 3 days of leave, and a bank holiday Friday.

Even so, progress made on 2 counts.

Bed occupancy work

An initial meeting held to present the team's work on bed reporting so far. 40 wards, with patient counts at hourly resolution for the prior 3 months, updated to the end of yesterday. The initial feedback was interesting, and there was wider realisation that we have multiple stakeholder groups interested in different aspects of this work, but that there is surprisingly little crossover between them. Ops people want to know about yesterday and today so far, up to this hour, but we want very little historical context. Planning need the history but little of today. Management need % full, not patient counts. Ward managers want to understand the picture adjusting for over-capacity situations that are managed with outlying patients (those who ideally would be on ward A, but due to space constraints need to be placed on ward B).

A lot more to do, but we created some common ground to build from.

Datahoover

Some significant steps taken by the team towards using our {DataHoover} package in production Maternity reporting this month.

Datahoover is an internal package which has been maturing over the last several months. I started work on it just under a year ago, and that preparation has been justified since the introduction of Badgernet as our maternity patient record.

The final push is to tie up some loose ends with the package itself, but mostly to author the final pieces of SQL we need. There is more to do (even by the middle of next week), but we are now very close having results. The prize is a highly transparent reporting process, incuding version-controlled SQL from which we can do the work of fully documenting measure definitions. As well as improving the transparency of the reporting process, the package will also save approx 4 days/mth, and will make good audit lists available to clinical and management colleagues. This time next week I'll be able to share how it went.

The datahoover project is also testament to how much unseen work has to go in to make big improvements to underlying data processes. As a long-running project which has at times been difficult to put time into (important, but not urgent), it's really satisfying to be close to delivering reporting from it. This is something I need to make more visible to colleagues - both the benefits of doing this type of infrastructure work, as well as the leadtime involved (and the stability of direction needed to deliver a project with a leadtime measured in months not hours).

Until next time, have a great a great Easter weekend!