/
2020-08-24 Meeting notes

2020-08-24 Meeting notes

Owned by Tim Whitlock (Deactivated)
Aug 24, 2020

Date

Attendees

Goals

  • Review outstanding tasks and plan this weeks tasks

Discussion items

ItemWhoNotes
Outstanding Tasks

Tim Whitlock (Deactivated)

  • Now the logins have been sorted George has uploaded some sample files containing 1000 records. One was rejected due to a typo in the validation file. George will contact NHSEngand to resolve this. One was rejected because the inut file had he columns the wrong way round...George will resubmit this. 7 others loaded OK.
  • Tracing records James has got code which will make a trace request and updates an sql database. The second routine reads the file name in the database and looks for the tracing return file and then creates a set of rda xml files for successful traces. (https://bitbucket.renalregistry.nhs.uk/projects/TNG/repos/ukrdc_tracing/browse) Currently the rda file generated is untested. Currently it uses the tracing id in the localid. Shaun raised the question of how this was being stored and questioned wheter it should be stored as a data source since this would potentially cause links to incorrectly matched data. Would it be better to  have a tracing table attached to the patient record that holds the latest tracing data (success or failure etc) The current plan was to create a rda xml for successful traces with unsuccessful ones having no file. There are known cases on the system which will trace successfully but the tracing data is wrong or incorrect for this record. Some more thought needs to be given as to how to record the tracing info. 
  • Tim is back from holiday and is now working on the medication dosage mapping under Hugh's guidance.
  • There has been no decision on the quickloading of files.
  • James hasn't added the virtual machine documentation for the BCH image but will today
  • The alphastudy site is still not backed up due to the lack of credentials that work. George is waiting on Fiona to confirm if she has any working credentials.
  • Andrew has been going through the AKI process with James they have put together a proposal of how to revamp the code   Proposed solution for AKI extract process which aims to improve the processing. George pointed out that if they use their proposal of tracing regularly they would still need to ensure  all alive patients in an extract are traced to check DOD prior to extract. The current hashing of the NHS number is extreme in that it does 250 hashes on each one and takes a long time. This needs looking at. Another aspect that was discussed was splitting out the patient demographics and linking the alert back rather than duplicating the demographics each time. George pointed out the need to know the post code at the time of the alert which means this should be stored with the alert. Often it would match their default demographics but not always. Andrew mentioned AKI-111 which is an improvement to the loader/validation to allow the selection of the database to use in the executable loading. Currently the executable always uses live and running the cod defaults to the test database. 
  • George hasn't looked at AKI-109 but will resolve this week.
  • George will be getting the HES extract prepared this week fro submission
  • PHE extract will follow straight after. George will need to clarify with Fran which files Q100 have been loaded to avoid duplicating the values by direct reading of the files.
  • NICOR extract is now complete. The issue appears to have been that the file needs to be uploaded with a reference to the researcher who would then be notified by email of the upload. Without it the notification goes to the systems team. This has now been sorted.
  • No feedback from BC has been received
  • National Opt needs to be applied from the end of this week to all unconsented extracts.
  • Covid19 is all working fine some missing documentation is still required for the latest changes however Andrew is confident he knows what is required for this weeks reporting.
  • Retha highlighted the request to include covid +ve tests with the simplified data when the number of them reaches 10% of the simplified cohort. There has also been an issue with the NHS BT matching in that certain data should only be shared if NHSBT is aware it is.
  • RaDaR nothing progressed this week.
  • No updates on the kings/barts feeds to report
 Junk Email

 George Swinnerton

 George highlighted the fact that since the email system changes a number of emails were ending up in the junk folder on outlook. In particular Jira and confluence emails. This has been caused by the change in the systems and the email classification process. It should resolve it self over time but it was suggested a global email to remind people this could happen to their emails would be useful.

Action items

  •