Handover

From Kautepedia
Revision as of 18:57, 25 March 2025 by Alex (talk | contribs)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

Background[edit | edit source]

This page outlines key Data team tasks, and summarises status and key pending actions to be done, to assist in future planning.

Overview[edit | edit source]

Overview of key projects/pieces of work shown in following table.

Key:
- On hold/pending
- Under way/review
- Under control/maintain

Domain/project Item Status
Kotahi implementation Onboard teams
Kotahi implementation Development/features
Population health data Maintain data
Population health data Extend BI
Internal reporting
External reporting
Digital strategy
Data integration Indici<>Kotahi
Data integration Other systems
Insights Power BI
Insights Future BI
Data Sovereignty pilot 1:Research
Data Sovereignty pilot 2:Development
Data Sovereignty pilot 3:Feedback

Kotahi implementation[edit | edit source]

As at March 2025, the Whanau ora team have fully migrated to Kotahi. This incorporates:

Additionally we have migrated Transition to adulthood (Social services).

No agreed migration dates are in place yet for other teams/services, but we have been planning with remainder of the Social services team.

One conceptual issue to flag is how we want Kotahi to support the Wellbeing model. Specifically, how do we want to organise Kotahi 'services' and assessments to support the model? These are some philosophical questions almost that have to be worked through before we go much further building workflows inside Kotahi.[1]

Next steps[edit | edit source]

  1. Define high-level approach to integrating services/teams into Kotahi
  2. Validate 'hauora assessment' approach
  3. Process mapping/documentation of pathways
  4. Design and implementation of Kotahi services/activities
  5. Update/adjust business rules in data store to ensure data reflects the above.

Population health data[edit | edit source]

This is a long-standing item from very early PPDF. It was set down by a public health expert, and aims to serve two main functions:

  1. Identification of population health needs/disparities that K'aute may wish to serve, and
  2. Exposing bigger picture outcomes which can contextualise Wellbeing model evaluation.[2]

Broadly this involved making patient-level data requests to a range of stakeholders - primarily Te Whatu Ora and Te Manawa Taki PHOs. Te Whatu Ora data is from National Collections, and PHO data is aggregated from member PMS systems. Because of the need to work directly with each stakeholder and their own privacy/consent frameworks, this has been a large amount of work with variable results.

Nevertheless, some baseline data is available in the above Power BI app.

Next steps[edit | edit source]

  1. Secure stakeholder time to understand this data and discuss/ask questions
  2. Extend/refine analytics accordingly
  3. Update/maintain data periodically
  4. Adapt per Wellbeing model evaluation framework (TBD).

Internal reporting[edit | edit source]

A lot of work has been done on exposing data in Aiga for performance, quality and other BAU reporting activity. Over time this has extended to include Indici data, and more recently Kotahi.

As at March 2025, we are actively discussing data for Board reporting and other ways to inform organisational performance.

Workplan[edit | edit source]

The Workplan app has been a centrepiece of internal quality/performance monitoring. The idea is to quantify the amount of 'meaningful activity' we are engaging in.[3] This app has a framework of setting different thresholds for staff/teams and providing a weekly view (week so far, last 6 weeks, etc).

System specific[edit | edit source]

There are a number of Power BI apps also showing data from specific systems - Aiga, Indici and Kotahi.

Next steps[edit | edit source]

  1. Build competence/confidence in use of internal data
  2. Extend/refine analytics according to need/demand
  3. Monitor/improve data quality
  4. Adapt per stakeholder need/demand.

External reporting[edit | edit source]

Funder reporting can be onerous for staff so we have gradually worked through semi-automation of data production for selected contracts/services. Please refer to Report list for an overview.

Integrated contract[edit | edit source]

The Te Whatu Ora Pacific Integrated Contract (PIC) is still under development. As at March 2025 K'aute are still working with Te Whatu Ora on a specific set of metrics to report on.[4] They require a test run of data against this framework by April 20th, for month of March 2025.

Refer to Integrated contract for update of report framework.[5]

Other funder reporting[edit | edit source]

As mentioned, a range of other funder reporting is needed on a regular and ad hoc basis. Organisationally we should aim to assist or automate as much of this as possible, using our data store (where appropriate).

Next steps[edit | edit source]

  1. Utilise pilot template to draft business rules and report process
  2. Confer with stakeholders to check and QA draft business rules
  3. Extend Report list to develop a library of business reporting, and links to documentation around process and business rules.

Digital strategy[edit | edit source]

K'aute have been working to a 'Phase 1' digital strategy for a few years, aimed primarily at uplifting digital maturity to a baseline level. This has included simple objectives such as selection and implementation of systems/SaaS products, as well as fit-out of KPT village. Most objectives under this strategy have been completed and it is time for a refresh to inform the next phase, aligned with the Wellbeing model.

At this time, K'aute have engaged Spark Health to lead this process and we are still discussing scope and terms of that engagement.

Next steps[edit | edit source]

  1. Data function needs to contribute to the review
  2. Data function needs to overlay a 'target operating model', and define what can be done in-house versus outsourced/hybrid.[6]

Data integration[edit | edit source]

Current data work to date has been focused on collation and integration of available data in a centralised place. Automated workflows are in place to load and manage data from Indici and Kotahi; Aiga data is being left in place and merged on-demand via Power BI. There may be a need to ingest data from other sources over time.

Data integration activity here has two main strands.

Data pipelines[edit | edit source]

Ongoing work is needed to extend, monitor and improve existing pipelines. This is about getting cleaned and organised structured data into an accessible place, along with monitoring and integrity checking processes.

Analysis objects[edit | edit source]

With required data in place, new analysis objects need to be built to make good use of it. For example, a 'risk stratification' view could be built which flags data points indicative of risk and compiles a summary per family/individual. This might use data from all available sources in the data store, push it to analytics platform and/or notify stakeholders directly. Some of these are already built and being used in existing apps/processes, but development of Wellbeing model will require definition of more of these - subject to more detail on the operational detail of what data needs to go where.

Next steps[edit | edit source]

  1. Monitoring process for data quality and integrity
  2. Ongoing security work around AWS infrastructure[7]
  3. Work with project team around operational workflow for Wellbeing model:
    1. Process map
    2. BA work around technical/user requirements
    3. Development of processes where required
    4. Documentation etc.

Insights[edit | edit source]

As part of digital strategy refresh (see above) KPT need a clear approach to accessing and using insights from data. This often has two parts:

  1. How are we doing on things we know?
  2. What is happening that we didn't know about?

The first of these is often about defining KPIs or key business metrics, and then tracking progress - for example, Workplan app. The second implies a better resourced and more proactive analysis function.[8]

There is also a desire to increase 'data literacy' of the organisation generally, and this has a number of prongs which may require focused work:

  • Practical comfort with data tools - for example, competence with use of excel and Power BI
  • Baseline understanding of using data - for example, interpreting graphs and charts
  • Broader awareness of how data entry impacts insights, and
  • Being more savvy around contracts and performance reporting frameworks we agree to.[9]

Next steps[edit | edit source]

  1. Decide on target operating model around insights/analytics (ideally as part of strategy refresh)
  2. Build up app/report catalogue informing KPIs:
    1. BA/investigation work
    2. Stakeholder planning
    3. Documentation
    4. Data modelling
    5. App development
  3. Identify baseline competence for staff, and design and implement tactics to achieve.

Data sovereignty pilot[edit | edit source]

This is still at an early stage, and there has not been time to deliver anything tangible yet.

We have partnered with Matou to deliver a 'proof of concept'. Funding from Trust Waikato totals $50k. At present we are working to arrange a few initial workshops with community representatives so that a 'discovery' piece can be done. This will then feed into planning and development of a pilot product, if appropriate.

In terms of roles and responsibilities, Matou are leading project management. Our role is to steer the project and facilitate access to anything Matou need (people, resources, data).

Next steps[edit | edit source]

  1. Continue planning of community workshops with Ma
  2. Involve Solomon directly in planning and development work, if any
  3. If appropriate, work with Kotahi on a data sandbox to integrate with.[10]

Overview of next steps[edit | edit source]

Piecing all the above together, the below table summarises key information. Here complexity indicates something that has a lot of interacting or overlapping parts and/or something with requirements or properties that have not been defined. Workload indicates amount of effort required in total to deliver what is currently known. Many of these are nascent or evolving, so these categorisations are subject to change!

Key:
- High
- Medium
- Low

Item Complexity Workload Competency Timeline
Kotahi implementation
  • Product expertise
  • Business analysis
  • Process mapping
July 2025
Population health data
  • Managing flat files
  • Database administration
  • Visualisation/Power BI
TBC
Internal reporting
  • Change management
  • Professional development
  • Requirements gathering
  • Power BI development
External reporting
  • Data modelling
  • Database administration
  • Operational/business workflow
Digital strategy
  • Digital/data thought leadership
  • Thorough understanding of KPT current state
  • Understanding of Wellbeing model/future state
ASAP
Data integration
  • Data engineering
  • Process mapping
  • User requirements
TBC
Insights
  • Familiarity with National Collections data
  • Understanding of health system workflow/data flow
  • Modelling data
  • Generating insights
TBC
Data sovereignty pilot
  • Pacific cultural competence
  • Managing community/stakeholders
  • Organising/hosting workshops
  • Issues pertinent to Pacific data sovereignty
December 2025

Suggested resourcing[edit | edit source]

Some suggestions about resourcing are provided below.

Kotahi implementation[edit | edit source]

  • Development and configuration in Kotahi can be handled by Solomon.
  • We should also think about developing other 'champions' who can handle configuration tasks.
  • Will need help with process mapping and planning. We can use our Kotahi agreement to fill this gap, but also directly involve Team Leads.

Population health data[edit | edit source]

  • This is a simple pickup and load of data.
  • However, putting it together usefully requires careful consideration.
  • Suggest involving someone familiar with National Collections data and/or a stakeholder who can contribute to understanding of this data (Te Ngira/Polly).

Internal reporting[edit | edit source]

  • Business side could be led by Mel/Dianne or Team Leads.
  • Will need outside expertise if we want to go ahead with organisational upskilling.
  • Data/app development can be handled by Solomon.

External reporting[edit | edit source]

  • Particularly for PIC, business side again led by Mel/Dianne.
  • Data/app work can be handled by Solomon.

Digital strategy[edit | edit source]

  • To be carried out by Spark Health.
  • KPT requires informed contributions to manage scope and potential for provider capture.

Data integration[edit | edit source]

  • Work to be determined via Wellbeing model/user story development.
  • Likely to need some external resourcing/capability if new products, particularly client-facing products are needed.
  • Otherwise BAU integrations can be handled by Solomon.

Insights[edit | edit source]

  • Needs additional resource; quite a niche field to do really well (again, possibly Te Ngira or similar).

Data sovereignty[edit | edit source]

  • Bulk of work to be led by Matou.
  • We need to support with access to our client base/community reps.
  • We should also ensure Solomon is directly engaged in as much as possible, because this project speaks to key issue for Pasifika.

References[edit | edit source]

  1. Or run the risk of having 20+ services in Kotahi and not taking the opportunity to improve organisational workflow.
  2. But can't be directly attributed to it.
  3. For example, face to face work with families, compared with administrative or other activity.
  4. This has been complicated by ongoing restructure at Te Whatu Ora.
  5. Copy of template and in-scope metrics still not accessible to data team as at 20 March 2025.
  6. Also SLT decision around future model, of course.
  7. See also: AWS Config.
  8. LLM tools may be useful, but correlation testing is also an easy way to quantify relationships between numeric variables in data. For example, https://www.sthda.com/english/wiki/visualize-correlation-matrix-using-correlogram.
  9. And, more importantly, the systems we are obliged to use.
  10. Or alternatively construct a synthetic one to demonstrate principles.