Problems List - NZ National Specification

Have created some headings and ask if folks could maybe contribute… we’ll see how it works.
For the discussion that has kicked this off see https://ehealthforum.nz/t/orion-problems-list/9470

What is the principle purpose of a problems list?

  • Provide centralized repository of clinically relevant issues (definition of clinically relevant TBD)
    - Up to date - defined as updated as new clinically relevant issues arise
    - Curated - routinely reviewed and “cleaned” to ensure the list is capturing
    optimal data
    - Visible - should be visible for care providers at times of clinical interface with
    patient

What secondary requirements should we attempt to consider?

  • codifiable for research, QI, Administration, etc
  • Captured as part of user-intuitive clinical workflow
  • Able to capture incoming clinical terms (SNOMED) from e.g. eReferrals and incorporate into clinical workflow e.g. ED admission → Dx diagnoses → Ward admission → additional / changing Dx → Dx summary
  • Potential to have the list of problems linked to interventions and monitoring of progress to enable review ####requires specification of intent - how does this pertain to content and format of the list ####
  • Potential to prioritise the list to show the issues your speciality is working on but also ability to see the list of what everyone is working on, important to enable teamwork and ensure synergy

What is out of scope for a problems list? (e.g. allergies)

  • allergies, alerts (e.g. domestic violence)
    *Noting at least one proprietary ‘Problem List’ includes ADR / Allergy
    *medications
    *Tasks
  • time limited minor issues (e.g. a laceration)

Who are the users/target audience?

  1. All clinicians with appropriate access and training and ability for patients to communicate a query or error
  • Clinicians initially through conscious interaction with the Problem list
  • Objective - incorporation of Problem list within differing ‘frames of operation’ e.g. visible within an admission or OPD document with workflow enabled population of Problem list via the workflow form / tool
  • Patient access via a Portal ideally once the GP, DHB and any other problem lists (e.g. Out of area) are unified
  • Patient editing via several staged capabilities
    Initially via messaging to relevant clinician if they have a comment / issue / addition
    Automatic inclusion of ‘Goals of Care’ or other key aspects of eShared Care Plans
    Patient initiated problem addition via Portal
  • Version 1 of this should be for clinicians within secondary and some primary care, GP’s may already have lists and

What is the minimum number of lists required?

The data should be the same but the presentation layer needs to be *different in various settings e.g. patient centric view, primary, outpatient, inpatient, etc
*Needs careful consideration: Key diagnoses / ‘Problems’ (short-term or ongoing) wherever added (IP / OP / 1ry - Community) should form a common core view. Other (Needs definition) entries e.g. multiple minor SNOMED-coded procedures during a hospital admission should be hidden unless tagged as important e.g. Angiography. There will be ‘mid-level’ entries (e.g. Pneumonia - admitted) that should be visible below the ‘core’ list

Chronic, in hospital, procedure, diagnosis, functional limitations, Other?

Nice to have sort and order features depending on the care setting and use case, especially to drill down into specific problems e.g. NLP to link spirometry for COPD

There would all be filters e.g. Procedures, Operations, Illnesses or ‘Associated with COPD’ e.g. Spirometry / CXR

  • The answer to this question depends on how many lists there are and whether the focus for the list is on what you are doing working on as a health professionals. We all have our part to play and if this is used by all will need to be able to filter views.

What features would optimise utility, curation and uptake?

Usability, visibility, transparency (interoperability).
Ability to easily view and edit as part of normal clinical workflow e.g. clinics and hospital admission. We have an admission summary that populates certain info to discharge summary and clinicians intuitively started closing the loop because it saved them time and the patient didn’t have to be asked the same thing multiple times

  • This list needs to be linked to a SNOMED onto server so users can search for terms using synonyms. this may be cumbersome at first but over time should get better as common search terms are surfaced quicker
  • List should be able to be sent to other documents like discharge summaries or clinic letters when being created

What other data fields/metadata would be needed? (date, user/ role uploading, tracking of changes, severity etc)

Ability to flag - using metadata/ SNOMED associations if a similar diagnosis / entry already exists. E.g. DVT to enable confirmation of ? New ? Continuation of existing.

Need rules for the above to ensure (e.g. for a DVT review visit) that the original date of occurrence is retained AND then all associated visits can be linked to that initial date / diagnosis (like for a PAS, a series of OPDs can be linked to a specific referral). IF the Clinical term is incorporated into say an OPD eWorkflow form, the above process will be automated with the CLinical Term link mirroring the PAS referral link.

  • Following fields may be of use (User CPN number, SNOMED problem code, Date time added, Potential severity details (free text tweet length), action plan to address item ( Text but again succinct maybe two tweets), status (i.e. for review, ongoing etc) if for review planned time to next review if needed

What standards are relevant?

SNOMED-CT, FHIR, openEHR
Non-traditional standards like UI heuristics to employ colors, symbols, contrast and other subtle cues to guide and prompt users, making it easy to do the right thing

What should we call it?

Curated Clinical Summary (* Who is the curator?!)

  • Ongoing Clinical Conditions under Active Management (OCCAM’s) List
  • Diagnoses and procedures

I’ve made this a Wiki in its own Topic; hopefully this will be a useful way for us to work together to produce a unified standard approach to this curly but important problem.

This hopefully will be used to help @Mat do the Hawkes Bay DHB Problem List. If it is successful we can then look at spreading it nationwide. Hopefully this can be a showpiece for us converting words to action. I encourage all who are passionate about this to contribute. Go @nz-ciln!

Thanks Nathan! both of us have put in a few starter comments. Those who have been part of the discussion on a seperate post up until now will no doubt have more to say!

What I hope this Wiki will achieve is a list of headings and points to seek agreement on at the face to face meeting at HiNZ.

Please contribute!

Hi all here are my thoughts I’ll start with the name

edit by @nathan: I’ve rolled most of @brian.yow’s post content into the wiki above

Why - there is stigma around words such as ‘problem’, ‘issue’, ‘medical’ and ‘patient’, especially if this is available to all

Curated - in the subtle hope that it is updated regularly
Clinical - only things relevant to a person’s care should be on the list
Summary - reminder to keep it short and succinct

I have added some bullet points under the headings, when you do also post an update on what you did.

To me a list is only good if you are going to use it and use it regularly. I also feel that as clinicians we should be able to focus what the main issues that we are working on and that by us all working from one list with the ability to filter it, so we can get a view of just your bit, as well as what everyone else is doing and working on we will be able to do great things. #transdiciplinary practice

I came up with a catchy name for the list as well which may help convey what it I hope we will be able to achieve. Introducing the OCCAM list (Ongoing Clinical Conditions under Active Management (OCCAM))

I edited for clarity.
I also inserted comments enclosed within ### this structure #### - apologies if this is the wrong way of doing it - used to inserting comments in code in this way

Fundamentally I think we are losing clarity of vision of this concept quickly - I think a portion of the current discussion is conflating a list as a data source versus a user-facing software that is presenting the data source.

I think it is very important to get the data source cleanly defined, with expectations of how it is created and maintained, but a lot of the discussion of secondary lists, active lists, work flows etc, really all pertain to the user-facing software. this list is NOT building an user-facing software (that is really an EHR), but rather is building a data source that would be software agnostic in terms of offering a feed to whatever the clinical environment requires.

3 Likes

@matthew.strother I agree completely with your comments, hence why I haven’t participated more. I think we need to edit the wiki to make it clear that this is about a data model (and we need to decide if it is a maximal or minimum data model ) and the main problem(s) we are trying to solve with such a list.

How we display the data (UI) and the business rules (Who enters, when, who maintains etc) comes second.

It might be helpful to start with a couple of specific use cases?

I currently have a thread called persistent patient health summary which is a piece of work that I’m doing at healthAlliance to try to agree to a minimum data set for a summary care record @brian.yow your suggestion to call this curated clinical summary is a broader scope, a curated clinical summary contain more data than a ‘problem list’.
Here are the links to two references that might be helpful

http://www.jointinitiativecouncil.org/registry/standards.set.patient.summary.asp - This is a collaboration between major standards body to produce a specification for a patient summary- and a ‘problem list’ is part of that data set.

Attached Core Personal Health Information A3.pdf (248.6 KB) is a poster from the MOH that is the output of all the workshops for the national EHR defining a core data requirements for a summary of personal health information. In this model they refer to problem lists as Medical History.

The scope of of this heading is Historical information and events that the consumer/whānau has previously encountered including primary, community and acute instances to understand and help formulate the overall health picture
Useful for: getting a more comprehensive overview of a consumer’s healthcare interactions that could improve clinical diagnosis, treatment, and condition management
May include: relevant diagnosis, problems (including general and mental health and social) and treatments or therapies a consumer has undergone (e.g. type of surgery, specialist care), relevant medical events (incl. adverse), oral health, medical devices

2 Likes

Yeah - but understanding the use to which the data source is to be used is important in the design of that source. If anything, I’d be concentrating on the usage requirements up front…

3 Likes

In my opinion, having been involved in similar activities, clearly outlining the problem that this list seeks to solve is critical.
This then helps to define the scope of the list, which leads to user requirements and then to the minimum data set/model, and so on.

2 Likes

Hi team

I agree that agreeing to the ‘scope’ and defining the problem we are trying to solve is critical hence my thought that a couple of initial use cases would be helpful. I am sure this is already done internationally. Let me do a bit of investigation to see what I can find in published standards etc.

There’s a ton of stuff - but the IPS (http://www.hl7.org/fhir/uv/ips/history.cfml) might be a good starting point…

(Though its document based and I don;'t like the way they handle empty lists)

2 Likes

I agree that there should be clarity about the problem you’re solving. Then use OCCAM’s Razor (good name and reference Nathan B) to aim at getting the list right in terms of content and usefulness. It’ll never be perfect but must be easy to use (in terms of data and UI).

@amscroggins has raised another issue, which I think is the elephant in the room, i.e. the ‘summary care record’. This concept has been doing the rounds for decades and doesn’t seem to come to a resolution. It’s more of a wicked problem than a solution. Which is why Inga’s point about focusing on the problem you’re solving so that you come up with a useful and usable solution is so important.

2 Likes

Almost! “###” is used to signify headings. Try “>” instead (or use the Blockquote thingo:image

Also, comments will make the wiki bloated very quickly - please instead make them in a post, referencing that bit of the Wiki using a Quote.

1 Like

I’ve removed comment as per nathan’s suggestion. moved the really most relevant question to a separate response.

I think I generated a lot of conversation yesterday around this comment - but I still stand by the general idea.

I think a lot of this is getting out of scope of the conversation - I think you define a list first, but define the API/UI requirements in terms of how that list feeds to other software. This component seems to be conflating a data source vs a UI software - which given at present NZ is a chaotic ecosystem with multiple different user-facing medical softwares, demanding UIs at the datasource is infeasible. First get a list, specify content, specify who/what/how its added to, edited, curated, and specify the input/output requirements to allow that editing and subsequent feeds into a user-facing software. How this integrates with the diverse user facing software solutions used in different settings is really a contracting issue with those software - e.g. a requirement that user facing software take a feed from ‘THE LIST’ data source.

I think the comments regarding the objective of the list are correct. I think deciding an overall objective is a good place to start. However, that is different than the details of presentation.

I also suspect I am making an implicit assumption about this being a list of major medical and social issues that pertain to a given patient - that it would be a central source of truth - that would be viewed by people engaged in care of a patient. how it is presented/used, would be at the UI side, but philosophically, I think having a central list, that by design, is not completely visible to each carer, is highly risky - because if a specialist view eliminates some of the data, there is no human in the loop deciding what is important.

3 Likes

I agree with Matthew here - that there is a growing amount of “out of scope” discussion.
I believe that we need to keep the scope of this discussion as targeted and restricted as possible. This topic has the potiential to become enormous in a very short period of time.
With this in mind, I have taken the liberty of editing the Wiki (appologies if I have modified or deleted comments you have made - please feel free to edit over my edits!)

what I have done:

  1. Deleted “tasks” and “issues” from the problem list requirement. I accept that “tasks” are important, but tasks are quite different from “Problems”. We need to keep this topic constrained and deliverable.
    a. Tasks or jobs are issues which are best considered as provider centric (compared to patient centric). They also have different requirmenets re notification to others within the team (pushing information rather than viewing) and sign off (a task is signed off when complete, a problem remains a historic clinical matter.

  2. Added medications and tasks to “out of scope”

  3. Deleted an in-line comment:
    a. * In the long run should aim to have medical, allergies/ADRs and non-medical ‘social’ alerts (as clinicians we often forget the non-medical socioeconomic determinants of health play a big part) ### concur, but what do you mean by “alerts” - do you mean an alerting function, or do you mean a socioeconomic or psychological or community problem that impacts patient - e.g. lives in decile 10, going through a messy divorce, lost access to children, chronically unemployed - I suspect this vagueness on my part is because we need input from non-doctor input (e.g. social work, nurse, pharmacy)
    b. - Allergies are out of scope . these are more critical and should be on an their own specific alert list and these should be shared nationally.
    c. - Medical history and other previous diagnoses, should be curated in a separate list. This is important but patients with comorbidites can have really longs lists. If we put a lens of things actively being worked on /monitored and tracking we will have a list that will be more useful and as a result more accurate. We also normally prioritise things and may want to focus on some of the comorbidites more so some will be on the list.

  4. Tidied up statements around number of lists. This question is “how many problem lists do we need?”. i.e. we may well need additional lists for allergies, social issues etc, but that is a separate discussion with separate requirements.

    Deleted * 3-4 lists
    i. - Medical history ( to keep track of diagnoses, severity and time made)
    ii. - The “Problem list” we are are working on defining of all active tasks we are working on aka OCCAMS list
    iii. - Social issues or supports available/unavailable to patient (e.g patient has no means of transport and has difficulty getting to appointments, homless etc)

  5. Added a name option “medical conditions and procedures”
    Good to see the lively discussion!

1 Like

I’m got a few thoughts and questions for the group… Do we think that the principle value of a NZ national problems list is kind of like the “One list to rule them all” notion being put forward for medication lists?

Are we better off attempting to deliver a problems list after delivery of medication list?
This is likely to have significant lessons for the development of problems list (which are a lot more complex)

Should the principle role of a national level product like this just to surface problems on multiple other platforms? It is highly unlikely, given the heterogenous IT platforms used around the country in health, and the disparate user needs that we will be able to create a list which is able to meet everybody’s needs (IMHO).
If we accept this limitation, then focussing on a visibility function is the most important, which implies that a curating and editing function are less important (curation would need to occur at the users local list level).

I find a central list, that contains all issues (to be defined) appealing.
I think waiting until the medication list is delaying the conversation that will need to happen anyway - because I think the conceptual content of a problem list is far harder than medication list.
I think how these are used will highly overlap - but I think the content and process of creating a list of problems will be not terribly dependent on the process of medication lists.

Example List

  • Congestive heart failure - cardiac echo 22.6.18 with LVEF 25% - NYHA Class 4
  • Acute myocardial infarction - STEMI 19.5.18, stent to RCA and LAD
  • Type II Diabetes mellitus - HgbA1c 25 30.5.2019
  • Diabetic nephropathy - eGFR 40 30.5.2019
  • Stage II prostate adenocarcinoma
  • Gleason 3+5=7, s/p radical prostectomy 12.4.2015
  • Alcohol use - 12 standard drinks/day
  • Tobacco use - 25 cigarettes/day
  • Unstable housing - lives in garage
  • Appendectomy 1944
  • Tonsillectomy 1948
1 Like

Would one assume that this list is codable, e.g. with SNOMED?