2020-07-27 Meeting notes

• George is still waiting on details of the portal for uploading from NHSEngland. Retha is chasing this.
• James has not made much progress on the tracing due to time being spent on other tasks.
• Renalware are currently making changes to ensure doses are correctly mapped using the prescribed pick list. Once this is done the issues with medications and syncing to RaDaR should be resolved.
• rrtstatus is working successfully on live George will put the update to live this week.
• George has added instructions on the covid19 matching required for reporting week to /wiki/spaces/ST/pages/1426226541
• Q100 validation is working and ready for use following the procedure George produced, however there is some confusion over whether we are ready to start. George has modified RR No allocation on test to ensure that corrections made to load on test will be correctly updated when doing the actual import using the partially corrected file used in the quick load. This aspect still needs testing to confirm it works as expected. When the DMs are ready to start the test database will need overwriting with a fresh copy of live. George will follow up with Fran now she is back from holiday and hopefully get this progressing.
• BCH there was apparently an issue with the admin interface not being configured as expected which James has now resolved. He also highlighted some Admin features Fiona didn't remember/wasn't aware of for setting primary diagnosis in the admin interfaces. James will aim to finish the necessary documentation this week including documenting the admin feature. This will detail how to build the VM and the file paths used and also how to deploy a new build to the VM as this will be the likely way updates will be done to the system when it goes live. James also suggested the Repo should be split into radar and bch versions however there seems little point doing this until the bch version is settled. It would also make updates that occur to the core radar code difficult to apply to the bch code as it would require duplication of work. Once things have settled this could be revisited  
• Turning on all patients pvfeeds will probably cause more issues with mirth logs which currently are using 76% of available space on live with 486GB (Total capacity 2TB) free staging however is 93% full with just 96GB free (Total capacity 1.3GB). If Leicester were able to enable a staging only feed we would have issues and even with a live feed it is likely to cause issues on staging. Ideally we would want to enable all patients in batches from Leicester so the disk usage impact could be monitored. Before enabling Leicester we will need clean up the staging servers mirth logs. This is likely to have more error messages due to the testing nature.
• National Opt out will need to be integrated into all our exports of non consented data by September as a secondary verification for opted out patients as it appears there are quite a few patients who have opted out on the national opt out but we still receive their data either because they are on PV or RaDaR or because the unit is submitting them in full for audit.
• James is now creating the extracts from the database and should be able to automate the validation and loading of files. Some more work is needed to fully automate the tracing process but the creation process is now automated. Managing the response file automatically still needs some work. James is also looking a a gui for triggering some of the processes to pass the control over to the data managers. Completing all the final bits should take about a week. James will ensure the documentation is updated to reflect the changes.
• Andrew has submitted a pull request for review for the radar code. george recommended some changes which have now been made and awaiting feed back from this. He will also look at the Syncing issue for RDA data which uses audit coding so needs mapping to the pv codes used by RaDaR. The other issue that is being seen is all the connections to the database being used up by RaDaR. This appears to have started happening with one of the RaDaR releases suggesting the is a coding issue with the api code. Andrew will investigate this week.
• Quarterly AKI documentation needs to be read by George and Andrew to confirm that it clearly documents what will be need. James will confirm when he has updated it to represent what is currently being  done and required. Anna has changed what she is requesting and the size of the data set has been proving a problem. Once this is all updated George and Andrew will feed back any queries. James pointed out that the code now is not fit for the job due to memory limitations and dataset size. The code was not written with the growth in mind and needs to be redesigned to handle the large size moving forwards. 

 James will focus on BCH, AKI documentation and the Covid code

George will focus on the rrtstatus, a DSA for AKI data and progressing the Q100 quick loading

Andy will focus on the RaDaR tickets