Update of D ID to include a laptop identifier

This page documents our move to a D_ID that contains a Laptop identifier.

Why do we need to do this?

Our D_IDs consist of the site and location and the Pat ID/Serial number, e.g. HSC_A4-123. They must be unique. If two collectors wanted to enter for the same ward on two different laptops, we would need to make sure that there are no duplicate D_IDs used. Right now we use two methods for this.

• coordinate to not use same Pat ID/Serial numbers
  • tedious for collectors; somewhat error prone, but since this will result in an error on sending rather than data corruption it is somewhat low risk
• add "fake" locations with a post-fix to make the D_ID unique that way (eg HSC_H4H_a, HSC_EMIP_a, HSC_IICa, HSC_MICa, HSC_SICa, see Property:Collection Location workload designation, STB Medicine Workload splitting)
  • this means we need additional entries in s_dispo table for each laptop/location combination, and analysis of data is somewhat more messy than it needs to be

this still uses the old laptop locations now... will need to clean that up or keep info about it around.

• SMW

• Cargo

• Categories

If/when we move to PatientFollow Project, the first solution would become completely impractical, and we would have to add dozens of fake entries to S dispo table to use the second solution, so we should fix it right instead. Also, with current understaffing, it would be nice to be able to collect another location when convenient to help others out. Finally, the situation of HSC Unknown Service patients leads us to a situation where we may not know for much of a pt's stay which laptop they should go on, so it would be best to collect as we go and be able to edit location to what it needs to be later.

STB Medicine Workload splitting will no longer need to communicate serial numbers.

Implementation

We need to add the Laptop identifier to the beginning of the D_ID.

Can anyone else think of additional things that would be affected by this?

• SMW

• Cargo

• Categories

update sending

We will need to update the send functionality for this. Sending_Patients#What_happens_when_patients_are_sent This requires a change to a lot of queries, so we would want to proof this before rolling it. Probably best to make a test patient on Tina's laptop send before and send after, and then test that Pagasa Query to see if pt us successfully excluded from orphans.

deal w orphans

Orphans in Centralized data.mdb process needs to be updated as well, will be easier going fwd

• SMW

• Cargo

• Categories

"Orphans" (Orphans in Centralized data.mdb) are records in CFE that have RecordStatus=Incomplete and a SendDtTm that is earlier than the latest SendDtTm for the sending laptop.

When we go live, there will be orphans because there will be two copies for in-progress patients:

• copy initially sent with the no-laptop-D_ID
• copy sent with the laptop-D_ID

We need an easy way to tell "real" orphans from the in-progress no-laptop-D_ID ones that should just be cleaned up after this. In other words a query for Pagasa to help make that determination. This process still needs to ensure that "real" orphans are captured and assessed.

Need to check for patients who have corresponding record in both groups and delete the pre-laptop versions of those. There would be a lot of these for some laptops with backlogs, so manual process would be time-consuming and error prone. The clean-up needs to be done repeatedly until all laptops have sent, so needs to be a process Pagasa can do. "Corresponding" records can likely be found simply by comparing Start Date and Time and Serial and Service/Location fields since they would be identical Records need to be deleted from two places: L_Log and L_PHI, so this is at best a two-step process not just a simple query. Proposed: (1) utility query that identifies D_IDs or records with the earlier SendDtTm (2) query manually run by Pagasa that deletes L_PHI records (3) query manually run by Pagasa that deletes L_Log records Are we certain that this would do the trick and not delete anything that we need to keep?

• SMW

• Cargo

• Categories

CCMDB data updates

There are some currently unused tables that had been added for the PatientFollow Project, they would need to be updated.

• s_laptops table
• S PatientFollow distribution table

Data use - no changes required

Julie has confirmed that we do not parse or otherwise draw meaning from D_ID in ways that would be broken by this change.

TISS Scanning

Background:

• Currently the Unit and Serial on the TISS forms uniquely map each form to a patient / D_ID. To move to the new definition of the identifier we need to find a way to connect TISS to D_ID. One way would be to re-do the form and the TISS28 Form Scanning process to accommodate the new data natively. We rarely change things in Teleform, so it would be preferable to not mess with that working system.

Proposed solution:

• There is a legacy field "Page" on the forms. It is still being scanned and verified, but we are no longer capturing the data because testing for continuous dates is sufficient. We will re-label and re-use this field to capture the laptop ID instead.
• Pagasa has tested a re-labelled form and the Teleform scanning, verifying, committing and importing worked.
• Julie has been consulted about this and thinks this should cause no problems at her end

Question has been put to collectors at TISS28 Collection Guide, waiting for feedback. Ttenbergen 08:03, 2020 February 12 (CST)

• SMW

• Cargo

• Categories

Required changes:

• TISS28 Collection Guide need to be updated with this.
• TISS28 Form Scanning process needs to be updated
  • process itself needs to be updated
  • instructions need to be updated
• File:TISS28 Form.pdf + local versions need to be officially updated (this would be a good time to update TISS_Form_(TISS28)#Ordering_copies_from_the_HSC_print_shop as well
• This can happen ASAP even before we change the D_ID

First week in March Julie is away and Vac planning is done, so that would be the time to go live.