nhsconnect / careconnect-document-viewer Goto Github PK

There are patients in the Clinical Document Viewer with no associated Documents. For example, “Jon Burrows” or “Max Manley”.
There are also Information Governance concerns and the risk that a user may import/upload real patient data for everyone to view.

There is a need to tidy up the list of patients. Potentially this could be done by defining a core set of good quality messages – say at least four per use case. Within each use case message set there should be at least one message that demonstrates the use of an attachment.

There must be a way of overwriting what people have imported every 24-hours to prevent the build-up of poor-quality data. However, it would still be useful to allow authorised staff from within NHS Digital to upload at future points further messages that would not be overwritten.

Transfer of Care as they go under names such as:

• Infectious disease notification
• Summary clinical document
• Discharge summary [Is this an Acute IP Discharge, MH IP Discharge, or EC Discharge?
Further viewing shows it to be a MH Discharge Summary.]

The “Infectious disease notification for “Mike Meakin” is labelled as a PDF and contains several examples of letters.

The “summary clinical document” contains a JPG, which when selected brings up an “Image Viewer” with a scan of what looks like a real patient note for a Robert Powell of Ystradgynlais.
The vertical controls for this image tool are not working.

When viewing the Discharge Summary (see:
https://data.developer.nhs.uk/document-viewer/edms/binary/5d13997accf84d000159086a ) the NHS Number at the top right-hand corner of the screen does not correspond to that displayed in the Patient Demographics section.

The Clinical Document Viewer should be restricted to Transfer of Care FHIR messages only, otherwise the whole point of the viewer is lost. That is non-FHIR messages should be removed and ideally not be allowed to be uploaded.

Enable validation of document upload - use ITK if this is performant or alternatively use CCRI validation.

There is a need for the Clinical Document Viewer to be able to identify attachments, i.e. message examples where the profile ITK-Attachment-Binary-1 has been used.

It would also be useful if any attachments could be displayed. It is unclear whether the Clinical Document Viewer has this functionality.

Some subheadings such as “GP Address” are not displaying in the ‘Resource Detail’ window (obtained by clicking [<>]). Missing text is shown as “array[1]”. Should this be allowed (i.e. by design) because of limited space, or is it important to show all? Its not clear where else this approach has been taken.

The “Patient Find” feature ( https://data.developer.nhs.uk/document-viewer/edms/patient ) appears to work in an odd way, and it can potentially mislead, for example:

If you search on “Manley” instead of finding “Max Manley”, a list is displayed of “Eve Everywoman1”, i.e. “Manley” is not found.

When clicking in “Patient Find” area of the screen, the drop down that appears in Edge shows text and characters (e.g. date values) that are not relevant as identifiers for patients. The intention appears to be to show previous search criteria so that if I did a search on “Smith” – next time dropdown appears then “Smith” appears in drop down. This partially works but is confusing with non-patient identifiers in drop down. Ideally some filtering is needed, and only patient surnames / NHS Numbers should appear in drop down.

When viewing patient details such as “Richard Smith” (see: https://data.developer.nhs.uk/document-viewer/edms/binary/5d13997accf84d0001590864 ) the “FHIR Document” header is displayed first, and the label for “Specialty” always appears to remain blank despite trying several message records.
The “Specialty” value appears elsewhere in the rendering, such as under “Discharge Details” or “Person completing record”.

When a patient or several patients have been found, there are several columns to the right of the screen that are not populated. The columns for “GP” and “Prac.” appear to be unpopulated. It would be better to populate these columns. If the number of patients / messages grow it might be useful at this level to have an indication of what ToC messages are associated with the patient record.

The system behind the document viewer is currently part of the Care Connect Reference Implementation.
This was not intentional and was caused by major technical re-structuring within NHS Digital (at the time of deployment).
The document viewer and backend systems should be decoupled from the CCRI (this is required to simplify changes to the document-viewer).

https://angular.io/guide/setup-local

The default position adopted appears to be an ordering of sections in alphabetical order. Patient Demographics would always need to appear at the top of the rendering for all four ToC use cases.

Likewise, it would be sensible to have the “Person completing record” section along with the “Distribution List” always at the end of the rendering.

If an ordering approach was designed from scratch, there would be a heading configuration file associated with each message type and the end user could use this default configuration file. Otherwise they could have ability to alter the configuration file so they can choose the order of the headings that are most appropriate to themselves.

If changes are made here, without a configuration file approach, the messages should be rendered in a fashion as close to PRSB guidance as possible.

Minor change to be performed during issue resolution.

Also look into moving to angular 8 (if covalent supporting 8)