2019-09-02 Meeting notes

• NHSBT file has been returned and now needs checking and uploading to the database and the export for NHSBT done. The documentation we be revised once this is completed as required. Hopefully this will be completed today.
• George Swinnerton  has found the deletion code created by Rapolas and now needs to check it and and see how it can be used for deleting only a single record tree such as the extra RaDaR membership. This then needs documenting. In addition some thought needs to be given as to how the different elements to be deleted should be referenced. Currently the deletion is based on NHS number so would delete associated records that may not have been merged so have a different ukrdcid. It should perhaps use pid for single record sets and ukrdcid for all related records. This code will need to be callable from the UI in the long term so needs to be built with that in mind and be part of the api.
• UKRR extract Kings have had some issues with the latest feeds where the codes are wrongly cased for PV codes (capitals instead of lower case) and sending organisation is 'ODS RJZ' rather than RJZ and also some codes are being sent but they are not ones expected by PV or the UKRR dataset. Kings believe they have fixed all the issues previously communicated in these feeds. George Swinnerton is currently comparing a 2017 quarter file with one originally sent to see how well they match. One obvious difference is that the original file has more results in. This may be because it was thought that the files were incorrectly populating the result fields from the previous quarters if no result existed in the current quarter when really the result should have been blank as the UKRR extract produces. Assuming all the fixes requested are in  place and the current coding issues are resolved it is believed that the feeds could be sufficient for creating the quarterly files. Currently the processing is being done on staging and will need to switch to live for use. The incoming RJZ files have created a number on new workitems for surveys, RaDaR and PV mismatches dues to dates and name issues. All the current ones have been checked on staging by Tim Whitlock (Deactivated) and resolved with the necessary updates verified on live surveys, RaDaR and PV and where necessary corrections made before the feeds arrive. Hopefully this will reduce the number of issues loading the RDA files on live.
• Updating the nurture extract has not been done. George Swinnerton identified an issue in some of the changes Rapolas had made which have caused the calculated egfr not to be included in the exports he believed because it was no longer using the sqlalchemy  property field. This needs to be checked and corrected to Fiona's satisfaction before the merging in of the nurture branch occurs. George Swinnerton will check the export code and verify if the issue was resolved before Rapolas left.
• Relocation the final version of the plans has been done and currently awaiting a printed version that can be checked to sign them off which should then trigger the process to get the contracts signed.
• Validation the primary focus of the validation are the active sprint tickets. Currently very little progress has been made and this needs to be rectified before the next validation meeting next week. Some of the simple task tickets are currently breaking the build on bamboo due to the tests needing to be updated to reflect the changes. Andrew Attertonwill sort these breaking updates and then focus on the active sprint tickets assigned.

 james.griffin (Unlicensed)is processing the AKI extract for Anna. Andrew Attertonfound some issues with the documentation and these will be updated by James.

Tim Whitlock (Deactivated) is due to have a meeting tomorrow with NBTIT to raise any issues and progress the move work. No new issues were raised and progressing the move is mute until the  we have contract signed. We are currently experiencing issues getting new accounts set up which will be raised. Also workshare no longer supports office 2010 with the latest update.

It is clear we need to formalise the requests for data extracts such that they are properly managed. Requests should be made via Tim Whitlock (Deactivated) who will then assign the work to the appropriate team member. If a request comes directly to a a team member it needs to be copied to Tim Whitlock (Deactivated) and discussed with him and any impact it may have on other tasks.