'Other' option in Cause - multiaxial diagnosis

[AmaniKrayemRCPCH]

Received a phone call and email from a user who has two patients whose causes aren't in the dropdown.

"1. Patient has history of neonatal seizures secondary to HIE/Hypoglycaemia. Brian MRI has abnormalities which is consistent with neonatal hypoglycaemia. Now presented with afebrile seizures . Regarding cause there is no other option in structural cause, only all syndromes and genetic cause.

2. The patient has history of perinatal stroke and abnormal MRI scan. Known hemiplegic CP . Now presented with focal seizures at age 9 years. It is structural cause, but no other option , so could not get 100% data completion in multiaxial diagnosis ( DESCRIBE definition) "

Considering these cases, would it be appropriate for us to add an 'Other' button alongside the dropdown list of causes?

@nikyraja thoughts?

[eatyourpeas]

I would tend to agree with the users. There will be a lot of epilepsy out there I imagine where brain damage of one kind or another happened perinatally or at some other time (eg scarring from infection or surgery) that has later become an epileptogenic focus. The two examples offered to my mind would be actually reasonably common. The alternative (and maybe better) would be to make the identifiable cause an optional field (changing the label to Is the main identified cause of the epilepsy one of the following?) and therefore not scored. If it is not in the list, it does not have to be scored. Or the alternative is to add these items to the list of terms. That actually would be my preference. And we could add a label which encourages users to contact the e12 email with any causes that are not on the list they feel should be and we could add the snomed codes.

I don't like 'other' personally speaking, particularly when the user is telling us they know exactly the cause.

This decision for this sits with @coldunk ....

[nikyraja]

Thanks both.

@coldunk and I have met and we have agreed to NOT have the 'other' option for now. This is because people will tend to select 'other' rather than looking for the most fitting option. Previously, users have also selected the other option and then written one of the drop down options in the free-text field. We want to therefore avoid 'unnecessary' other selections where possible.

Based on the description above and other common causes arising clinically, Colin has identified the following to be added to the causes drop down list:

• Neonatal Hypoglycaemia
• Hypoxic ischaemic encephalopathy (HIE)
• Perinatal stroke
• non accidental injury
• Periventricular leucomalacia
• Neonatal cerebral hemorrhage

Going forward, Amani and I will raise queries with Colin, Colin will use snowed codes to define the exact phrasing to use, and these will be uploaded to GitHub as issues. @AmaniKrayemRCPCH please can we create a log for these changes, so we can also make sure the documentation is updated. We will follow a similar process for adding organisations.

This process will allow us to evolve our dataset to best fit the clinical environment and needs of users, but will also facilitate easier and robust analysis/reporting of data. However, if the requests become unmanageable, we can revisit introducing an 'other' option in the drop down.

[eatyourpeas]

This is a great solution thank you

[anchit-chandran]

Do these need to be added @eatyourpeas ?

[eatyourpeas]

They do need adding. I think we are just waiting on @coldunk for the SNOMED IDS (SCTIDs) for the conditions he wants added. The workflow will then be to create a new list in constants with the new additions and add to seeding for the future, but also we probably need to write a small script to make a call to the snomed server with the sctid to get the concept and then store the result in the live database. Updating organisations also needs an automation workflow. Perhpas something to discuss for you to implement if you were happy?

[AmaniKrayemRCPCH]

Can I just chase this up @coldunk , @eatyourpeas

Another user has requested we add the following cause:

• Cornelia de Lange Syndrome (CdLS)

[AmaniKrayemRCPCH]

Found the below SCTIDs on https://termbrowser.nhs.uk/ @coldunk can you check these

Neonatal Hypoglycaemia - 52767006 Hypoxic ischaemic encephalopathy (HIE) - 703300001 Perinatal arterial ischaemic stroke - 722929005 Non-accidental injury - 420025004 Periventricular leucomalacia - 230769007 Neonatal cerebral haemorrhage - 261808007 Cornelia de Lange Syndrome (CdLS) - 40354009 Constitutional mismatch repair deficiency syndrome (CMMRD) - 764946008

OTHER Anoxic encephalopathy - 126945001

[AmaniKrayemRCPCH]

These are the final SCTIDs, approved by Colin

Constitutional mismatch repair deficiency syndrome (CMMRD) - 764946008 Neonatal hypoglycaemia - 52767006 Hypoxic ischaemic encephalopathy - 703300001 Perinatal arterial ischaemic stroke - 722929005 Non-accidental injury to child - 158094009 Periventricular leukomalacia - 230769007 Neonatal cerebral haemorrhage - 261808007 Cornelia de Lange Syndrome (CdLS) - 40354009 Anoxic encephalopathy - 389098007

[nikyraja]

Hello, please could I check if these are on the live site now?

[AmaniKrayemRCPCH]

Reopening following discussion in meeting:

A user has requested an 'other' option for genetic causes that are unnamed.

In the meeting, we discussed adding a 'Not included in List' button with the tooltip 'If you believe the cause should be in this list, please contact the Epilepsy12 team.'