2020-03-30 Meeting notes

Date

30 Mar 2020

Attendees

Goals

Review current progress and plan work for week

Discussion items

Item	Who	Notes
Outstanding Items	Tim Whitlock (Deactivated)	George has sorted one more of the three patients the leaving just one more to sort. He will add notes to confluence on how these patient issues need to be sorted and is currently looking at the last patient. The validation sprint has been completed and a new release made. Some issues were found but a fixed release was put out on Thursday and so far it looks like all is well. The optepro fixes are bundled with the datetime fix where dates are being compared and the time element is affecting the result when it should be ignored and the truncation of PV strings to ensure they meet the PV schema even if the repository allows a longer string. This needs deploying to dev and testing. We are receiving >10 digit NHS numbers from Barts because the spaces are not being extracted. Because the field is a generic NI field it doesn't limit the length to 10 however the jtrace does limit the NI field to 10 characters and so throws an error. We need to put some RDA validation in place to capture this prior to core code and in addition make clear to suppliers/units that they should ensure the NHS number is correctly formatted without spaces. George will contact Tim/Hugh about this. James has written some initial code for processing the weekly files although the original email did not include a examplar spreadsheet so currently the returns are coming in in a wide variety of formats requiring a lot of manual work. Hopefully this week a spreadsheet will be sent out to be used. It is hoped that the spreadsheet will do some preliminary validation however that aspect has had some issues so it may not happen this week. Currently the code expects a set of spreadsheets to merge from this a tracing file is created. The file once traced is merged back into the original file adding Date of Death. This file is the processed by the Stats team to create the reports. Changes this week to the weekly reporting is that they now want to report cumulative totals and new this week following the WHO reporting format. This will mean retracing all previously reported patients who are not dead to see if they have subsequently succumbed probably Thursday AM. Then returning to stats a list of historically reported entries with DOD and the weeks new reports. To try and capture issues earlier the files will undergo some "validation" when first saved to the latest folder if they pass they will then be moved to the current weeks reporting folder. If they fail they will go to an errors folder with error message so Fran and her team can try and resolve the issue and re-save the corrected file into the latest folder. This should ensure all files in this weeks folder will merge and allow the tracing submission to work without issue. James is trying to ensure the code can handle these changes. Much of this knowledge can be applied to designing the UKRDC tracing process..which Jame sis also looking at. George has made the changes for the NHSEngland reporting and is awaiting Retha telling him where to send the sample extract and confirmation of when the fall extract can be sent. After the full extract for 2018 has been sent incremental ones will be sent for 2019 as the quarter files are loaded. Andrew has just the syncing of Demo to complete on the RaDaR sprint and has been looking at the RaDaR syncing with UKRDC/UKRDC syncing to RaDaR to try and identify the cause of the errors and why some records are not being synced correctly. so far this hasn't been solved. TNG-570 which was to ensure that RDA coding schemas are correctly translated when syncing. The code conversions have been done and are ready to deploy to staging for testing. Kings and BARTs have modified their feeds however we are still receiving a lot of files and content ther is also a question mark over the date range process which was missing some results previously and George is unsure if the issue was resulted. We are also still seeing entered at ODS0000 codes which George will chase. As part of covid-19 reporting we would like kings and barts to add covid tests to their feeds. George Swinnertonwill follow up.

Action items

George Swinnertonto follow up on issue with ODS0000, NHS numbers being sent through with spaces and whether Kings and Barts can add in covid19 test results. 07 Apr 2020
james.griffin (Unlicensed) continue to develop the covid-19 weekly reporting code and if any of the code knowledge can be used for automatic tracing for the UKRDC
Andrew Atterton to continue looking at the RaDaR syncing issues and resolving them