2020-11-16 Meeting notes

Date

16 Nov 2020

Attendees

Goals

Review progress and plan progress this week.

Discussion items

Item	Who	Notes
Outstanding Items	@Tim Whitlock	Renalware have apparently deployed the fixes we were chasing to live without any testing. George Swinnerton still has to confirm they have been correctly implemented. The validation tickets and CKD1-3 no load changes should be completed this week. Fiona hasn't had anything back to allow access to the alpha study site. Andrew hasn't made any progress on the tracing files and is currently having issues with the creation of the RDA files from the tracing response. NHSBTfile is done Andre has the necessary mappings of hospitals and labs for the Nephwork data. She hopes to complete the work of the nephwork app this week. George re-ran the HES code this morning ready for submission to HES. He has now got the OPT out working. Roughly 3% of records were flagged as opted out. George noted that the returned order didn't match the sent order which meant it was tricky to identify if the file was truncated as happened previously. The HES file has two years worth of data.. Andre has completed the NHS England extract and submitted it. The Vascular access database matching has been completed and the data sent to NHSEngland as part of the NHSEngland extract. Retha Steenkamp now needs to talk to her NHSEngland contact about the on-going submission. It is likely they will want the first quarter then monthly submssions after that. AKI data issues. Originally the AKI load process was suppose to load the text string in one column and the "interpreted" value/number element in the value column. This would have allowed the reloading of the value column if the process to interpret the raw text string failed. Unfortunately a change in the code meant that the raw string did not get saved for all code. in particular if the number had a decimal point the code incorrectly assumed the data was invalid and did not load the value any where. This means to correct the data we will need to reload the files. The current process for correcting data involves deleting the data erroneously loaded and then reloading the file. The current database is not normalised and repeats the demographics on each row. The database could benefit from some normalisation which could have been done at the same time but could introduce new errors. It would be better to deal with each issue separately. The proposal is to autoload the complete set of files into a new copy of the database then verify the data against the current database to confirm sites with no issues are unchanged and confirm that sites with issues have loaded correctly this time. Once this has happened the databases can be swapped over. RaDaR - Andrew managed to complete the two tickets RDR-739 and RDR-740 before he was off ill so unable to complete the tickets and deploy. They should be deployed on staging this week for testing before deployment to live. Fiona also raised an issue where the last login dates were all showing as 05/11/2020 for all users - unless they have
AOB	Retha Steenkamp	There has been an issue in the BAPN extract where the dates were incorrectly formatted due to a mistyped format string which caused the date to revert to 1/1/Year on DOB and date first seen. George is not sure if this affected the PHE covid linking however it would not have affected the NHSBT matching. This issue has now been corrected.
RaDaR	Retha Steenkamp	There are three things Fiona has coming up for RaDaR Biomarker data, nurture biopsy report clean up (removal of PID) from free text fields and improved data export process. There is also a requirement for electronic consent. All are to think about how this might be implemented on/via RaDaR in some way. Things to remember RaDaR it self is on N3/HSCN so only accessible in hospitals. To provide an internet accessible consent process would require a separate internet based server. The process needs to simple enough that elderly non IT literate patients can do it easily. George wondered if the nhs login could be used as the identifer for signing the consent - https://digital.nhs.uk/services/nhs-login/nhs-login-for-partners-and-developers other options are just basic identifier confirmation (name/DOB/NHS number) or some form of electronic signature.
PHE Matching	George Swinnerton	George asked about the next PHE matching for the annual report. Retha thinks Anna has done the extract. This will need to be done by the end of December.

Action items

George Swinnertonto confirm the changes that renalware have deployed do address correctly the issues with their feed. 23 Nov 2020
Andre Ortega Alban complete the Nephwork app ready for the project kickoff 27 Nov 2020