Upload CSV to break visits into different submissions

[eatyourpeas]

Users that use Twinkle typically upload a year's worth of data at the end of the audit year. In quarterly reporting, they will be encouraged to submit the full audit year upto and and including the current cohort.

As the next cohort comes round, the upload will include the previous quarters as well as the current.

In CSV upload therefore, the dataframe should first filter out those rows from previous quarters and only save or overwrite those rows relating to the current quarter. This will also protect against overwriting data from previous quarters that has been valid being accidentally overwritten by invalid data.

If however there is no data in a previous quarter, so long as it is in the current audit year, that can be saved. This is because some centres may not submit data quarterly (although that is a requirement) and the priority is to keep and analyse all data available across the whole year.

[eatyourpeas]

This was discussed at Sprint 9/8/24 with NPDA team.
They would like a single submission for the audit year that is overwritten every quarter. This is because some measures relating to previous quarters might not be scoreable until later and would be entered retrospectively.
They would like users to be able to see which data pertains to which quarter, but this could be calculated on the fly using the visit date.
Quarter has therefore been removed from the Submission model and the latest visit date and associated quarter are now signposted to the user in the patient table.