top of page
Care plan hero.png

CHRONIC DISEASE MANAGEMENT

I designed a product which will allow Australian GPs to better care for patients with chronic diseases.

MY ROLE

UX Design & Research

Product Management

TEAM MEMBERS

UI DESIGNER

Daniel L

PRODUCT LEAD

Larry W  

 

LEAD UI DEVELOPER

Brian P

FRONT END DEVELOPERS

Balaji (off-shore)

Ram (off-shore)

QA

Alan T

STAKEHOLDERS

CEO

CFO

GM Marketplace

Head of Product and Design

CTO partner organisation

THE PROBLEM

Chronic Disease Management is one of the major health challenges in Australia today.

Chronic Disease Management is a huge problem from a health and a cost perspective. The Australian government, through Medicare, gives GPs the opportunity to treat patients with chronic diseases by providing monetary incentives in the form of specific Medicare Item codes to bill at a higher rate. Patients are also given five subsidised visits to Allied Health Practitioners (e.g. physiotherapists, podiatrists etc.) per calendar year if they meet certain eligibility criteria.

 

The challenge was to build a solution which could be launched from MedicalDirector’s clinical applications (Clinical and Helix). The solution would be cloud-based and would offer a quick and easy way for clinicians to create a care plan for any patient who had qualifying chronic diseases.

CONSTRAINTS & CHALLENGES

  • This was the second attempt at completing this project

  • There was a short timeline for research and design activities before development was due to start

  • There was limited budget for research in the initial stages of the project

  • The development team was spread between the US and India

  • The development team was relatively inexperienced

DISCOVERY

This project had been in progress for some time and, after a reset, I was brought on to lead the design. There was only a relatively short time to conduct the discovery, research and design activities, for a project of this size.

SCOPE REVIEW

Together with the Product Lead, we started off by reviewing the contract and the previously defined user stories to determine how we could best deliver the project.

STAKEHOLDER INTERVIEWS

There were subject-matter experts within the business who were crucial to building a solid understanding of the problem space. We also worked with Trainers, Customer Success Managers and Business Development Managers to identify suitable medical practices which would be willing to be part of research, user testing and a pilot program.

COMPETITOR ANALYSIS

We researched competitor's products and looked for similarities in other products in unrelated fields to gather ideas.

RESEARCH GUIDE

I created a research guide as a way of identifying the key tasks, processes and pain points which exist when creating care plans.

The problem
Discovery

USER RESEARCH

I visited 6 medical practices to conduct the initial user interviews

The medical practices were all located in Sydney. Each generative interview lasted approximately 2 - 3 hours, with a mixture of GPs and Chronic Disease Nurse participants. These interviews were recorded and transcribed so that each interview could be individually analysed. I also conducted interviews with our Chief Medical Officer and software implementation & training teams. Regular concept and usability testing was undertaken with these users over the course of the project.

The medical industry is quite conservative in its approach to the adoption of change. It can also be hard to engage practitioners for the purposes of research and we fully utilised the relationships others in our business had in order to find suitable (and willing) medical practices to participate. Because of this, we utilised in-house experts heavily.

User research
gen res Page 2.png
gen res page 1.png
Base research.png

Journey and Process Maps

Following our interviews, we synthesised our findings and were able to create a more detailed patient journey and process map of the typical care plan creation. This map focused more heavily on the specific drivers and needs of the users who create care plans, rather than the patient's journey through it. Prior work within the design team had established a high level journey for care planning.

These exercises formed a great foundation to move into feature ideation and prioritisation.

Care planning Journey.png

Co-created with Jyllian Thibodeau and Simon Mingo

journey map after research_edited.jpg
CDM Step 1.png
CDM step 2.png
Onboarding workflow.png

The State of the Art

Most care plans in general practice are created using a WYSIWYG (What You See Is What You Get) editor. There is an option load a pre-defined template, however all patient specific data needs to be entered manually by referencing the source clinical system by reading and transposing data. Each care plan represents a point in time, but cannot be added to and a continuous view of the patient's progress is lost.

This program comes included in the clinical software and the cost is built in - users see this functionality as "free".

Old care plan pg 1.png
Old care plan pg 2.png
Old care plan pg 3.png

After interviewing and observing the research participants, I saw that process of creating care plans was a time consuming and complex process.  A huge opportunity existed to make the creation of care plans easier and more meaningful.

In order to commuicate with this key finding with the key stakeholders, I created a sped-up video of the care plan creation process to show how much work and repetition is required to properly create a care plan for a patient.  I found this more easily demonstrated some of the user pain points compared to the journey map.

TITLE OF THE CALLOUT BLOCK

REFRAMING THE PROBLEM

The stated project goal was to create electronic care plans, but there was no indication of what that actually meant.

To enable electronic care plans we prioritised the following:

  • Help a user to create a care plan, simply and easily

  • Reduce the amount of extra work in the practice as a result of care plans being created

  • Engage the patient in their care plan

An important point to note:  this is not the first product to market and the dominant player was a free WYSIWYG text editor.  Our solution had to provide real value - enough that users would pay for.

How might we make it easier for the practitioner to do their job?

After multiple rounds of sketching and prototyping, the following approach was taken:

  • If the data exists in the source clinical system, then make it visible to the user in the care plan, automatically. This includes: past medical history (conditions), prescribed medications, correspondence to and from specialists, pathology results and measurements (e.g. blood pressure, cholesterol etc.)

  • Where discrete but related elements exist in the source clinical system, find a way to group these elements in logical way, ideally at a glance

  • Treat the care plan as a holistic document, showing the full history of the patient's progress

  • Treat the clinical system as the "source of truth" for all data.

How might we make it easier for the practice to do its job?

During the course of the research, practice staff described and showed how they manage the volume of care plans in their practice.

  • Some people used a paper based system to indicate when care plans require attention

  • Some people used a spreadsheet based system to show which patients require a care plan, which plans require review from a GP, and which care plans have or have not been billed, for example

After multiple rounds of sketching and prototyping, the following approach was taken:

  • Make it easy for GPs to assign work to practice nurses, directly from the GP's main screen

  • Allow the nurse or practice manager to track all the care plans which are due, overdue or haven't been billed

How might we invite the patient into the conversation?

It became clear from the initial research that, on the whole, patients weren't that engaged in their care plan. There were a number of reasons for this:

  • The patient wanted the care plan solely to be able to claim five subsidised Medicare visits

  • The GP is aware of the reason why the patient wants the care plan and does the bare minimum to satisfy Medicare's requirements

  • For those patients who have a care plan, the time taken to create a plan leaves little time to discuss the plan with the patient

Many practitioners observed that the best patient outcomes happen when the patient and practitioner have adequate time to discuss the patient's health needs.

After multiple rounds of sketching and prototyping, the following approach was taken:

  • Include a step in the care plan creation process to record the patient's goal, in their own words

  • Auto-generate an easy-to-read care plan for the patient. The care plan would show the patient's personal goal which would be a reminder of the benefits of following their care plan

Reframing the problem
Usability testing

USABILITY TESTING

We tested the prototype at 13 medical practices in NSW and Victoria. By this time, we had a working prototype which was able to tested with real data.  Delays to the delivery of the software meant that each session was a combination of Usability Testing and Contextual Inquiry.

The recordings from each session were transcribed and then the feedback and observations were mapped directly on to each screen in the workflow. Colour-coded sticky notes were used to synthesise the notes and record each pertinent observation as a question, a minor issue, or an issue potentially requiring more work.

For each of the separate testing sessions we used affinity mapping to identify 'hotspots' where it was clear that multiple participants had been shown to have issues.

Usability testing 2_edited.jpg
Usability testing 1_edited.jpg
Launch

LAUNCH

MD care plan page

MedicalDirector Care was launched in October 2020 after an extensive pilot beta.  It's gratifying to know that I was able to contribute to its success.

LESSONS LEARNED

Seek different kinds of feedback, earilier

I would have involved practices in co-design sessions, rather than presenting prototypes for their review.

The data is important

For these participants, being able to see realistic clinical data when using prototypes is very important. Early mock-ups had dummy data. It was only once we connected working software to practice datasets that participants really absorbed what they were seeing.

"The future has arrived - it's just not evenly distributed yet" - William Gibson

There was more variation in the sophistication of care planning processes in medical practices than the original research suggested. For example, the practices we researched in Victoria generally had a more sophisticated workflow than NSW.

Lessons learned
bottom of page