/
2021-03-15 Meeting notes

2021-03-15 Meeting notes

Date

Attendees

Goals

  • Review progress and outstanding tasks

Discussion items

ItemWhoNotes
Outstanding Tasks@Tim Whitlock
  • George has made some changes to the UKRDC schema ready to chase renalware to make the changes required to correctly send through the data as detailed on the Trello board. He will also add some code tot the resources to ensure that the schema is validated as correct xml.
  • The new pyxbx library is available and we can migrate the code to use it when ever we are ready.
  • Andre has received the info on the xml tracing from Joel. Tim will arrange a call to go through what we are trying to do and how it fits into the UKRDC.
  • Joel has created a page detailing a better approach to dependency management (Python Dependency Management - Systems Team - Confluence (renalregistry.nhs.uk) He will circulate a link and all the team will review and feedback. The aim is to migrate all our processes to cleaner scheme.
  • We now have the detail feedback for N3 and the internet and Tim needs to just review and pass the appropriate queries to those responsible for the code that has been flagged. In general no serious issues were raised just lesser issues with encryption choices SSL versions and obsolete libraries were the most common reported issue.
  • Fran has struggled to find time to complete the testing of the validation release which has been further delayed by the test files from Oxford which have had so many issues that further fixes have been required in the validation code to cope. In addition it has highlighted the 65K limitations of the current xls format meaning that the code needs to change to using the newer excel format xlsx whcih will be part of schedule for the next release  once the current version has been signed off and released.
  • Andre has deployed the Magic fixes for allowing staff to be associated with more than one unit. She is now working on allowing admins to view regional data for any region rather than just the region associated with the assigned hospital.
  • The local star registry setup works fine suggesting that the issues BCH are seeing are that a service has stopped working. Tim has recommended through Fiona that they should reboot the server and confirm if that restarts the missing services. If there are continued issues then we will need remote ssh/vpn access to the server to see what is going wrong. In addition we will need to provide a method to load their historic data currently held on spreadsheets.
  • Alphastudy data extract is complete and we are just waiting on confirmation as to whether we can shut down the server.
  • AKI code Andre has reduced the number of files with issues to 17 from 37 files and most of these are due to DOB being in the wrong format. There are also some files of creatinine previously loaded which don't have an associated alert file. These files should have an alert file in order to justify and prove we are allowed to hold the data. So technically we should not be loading the files without proof we are allowed to hold the data which George said meant the code was changed to enforce the Alert file requirement. Andre is checking with Anna about her thoughts and will also check with Sarah as to whether the missing Alert files could be requested.
  • Andrew has deployed the African CKD  cohort pages and also synced the radar database to staging. He has an issue with creating a repo on github which Joel will look into. He has also now got a dockerised method for creating a test environment which he will document.
  • At the RaDaR meeting on Wednesday it was suggested that the annoymisation work Afonso has done on Nurture text fields should be applied to the whole of RaDaR. Afonso can supply his scripts which will need tweaking and Kate has said she would work through the flagged records.  
COVIDAndyI was planned to put a rolling year date range in the covid files but this may cause an issue with sites who justs end a list of all patients rather than the latest set each week. As a result the date check now set to January 2020 as the earliest date. In addition there was a date check for DOB which didn't allow a child to be younger than 1 year. This caused an error this week with a 9 month old baby. This check has been adjusted to allow for any child age.
 SQLBackup
NHS England ExtractRethaRetha flagged that she thought an NHS England extract should have been done. She will re circulate the schedule.

Action items

  •  NHS England extract