Skip to content

Commit 2ee4716

Browse files
Merge pull request #19 from nhsbsa/processing-documentation
Processing documentation
2 parents d583e7b + 485c3e0 commit 2ee4716

16 files changed

+590
-1
lines changed
Loading
Loading
Loading
Loading
Loading
Loading
Loading
Loading

app/posts/processing/2024-05-31-template-post-1.md app/posts/processing/2024-05-01-template-post-1.md

+1-1
Original file line numberDiff line numberDiff line change
@@ -4,7 +4,7 @@ description: This is an example of a standard design history post. Thjis would p
44
author:
55
name: Mark Portnell
66
url: '#'
7-
date: 2024-05-31
7+
date: 2024-01-31
88
tags:
99
- claims-version-1
1010
- claims-journey-name
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,192 @@
1+
---
2+
title: Scoping for MVP
3+
description: First ideation session focusing on what we think the journey should look like, based off assumptions and hypothesis'. Aimed to get a first design out to take learnings off in research.
4+
author:
5+
name: Hannah Williams
6+
url: '#'
7+
date: 2024-05-05
8+
tags:
9+
- processing-version-1
10+
- process-a-claim
11+
aside:
12+
title: Processing Prototypes
13+
content: |
14+
[View processing prototypes](https://adult-social-care-7fe9bafd955a.herokuapp.com/version-index?area=Processing)
15+
Password: ascbsa123
16+
---
17+
18+
We had a processing workstream kickoff on ... with the wider product team to begin the processing a claim piece of work. After understanding the constraints and the scope of the journey from multiple buisness and tech perspectives, we had a in-person workshop with the UCD team. We explored ideas of how this journey could look, what are the needs and wants of the users and then stripped it back to the MVP of what we deemed necessary to complete a processing claim journey based off assumptions and hypothesis'. We tested this inital design flow w/c 22 April 2024.
19+
20+
[whiteboard picture]
21+
22+
## Why we did this work
23+
24+
- [if applicable] Sprint goal and some background for the sprint
25+
- [if applicable] Sprint numbers is spanning over multiple sprints
26+
- We are on version 8 of the claim creation side of the journey and now need to create the journey to process the claims that organisations have submitted for their users.
27+
- Mention any discovery work?
28+
29+
>**How might we** give users the MVP functionality they need to process a claim.
30+
31+
## What our ideas were
32+
33+
Core questions we asked the UCD team to figure out the potential needs and wants of the processor to complete the task of processing a claim.
34+
35+
#### How does a processor find the claim they are wanting to process?
36+
37+
Our assumption is the processor will be given just the claim ID for the claim they need to process in Operations Manager. They will need to use this ID to search in our service for the claim. We don't believe this ID in OM is able to be a link so the processor will copy and paste the ID directly into our service's claim ID search box reducing mistakes.
38+
39+
>**We believe that** only the claim ID
40+
>**Will be needed for** processors to find the correct claim
41+
>**As it will** allow them to search in our service and find their claim.
42+
43+
To be tested
44+
- Whether they copy and paste the claim ID or type it in themselves.
45+
- Whether mistakes are made and then they have no other information to validate they have the correct claim so how likely is it they process the wrong claim.
46+
47+
Screenshot
48+
49+
Here is the initial design in the processing a claim prototype journey for searching for a claim id:
50+
![A screenshot from the processing a claim prototype showing the search claim screen](claim-id.png "Claim ID search")
51+
52+
53+
54+
#### What claim detail information does the processor need to see when they land on a claim?
55+
56+
Our starting assumption for this question is that the purpose of a processor is strictly to judge evidence of payment and completion against some criteria stating it's the correct training etc so the processor will only need to know any claim information during the questioning and they have no other purpose for more information on this claim details view.
57+
58+
>**We believe that** only the claim status and the claim ID
59+
>**Will be useful information for** processors to process a claim
60+
>**As it will** allow them to know they are in the claim they expect from the claim ID and there is no additional information neccessary for them at this point in the journey.
61+
62+
To be tested
63+
- Whether users have any other purpose and needs for any additional information about the claim beyond information provided in criteria checks.
64+
65+
Screenshot
66+
67+
Here is the initial design of what we believe is necessary for a processor to know about the claim on the first landing view:
68+
![A screenshot from the processing a claim prototype showing the initial view a processor lands in with no additional claim details](claim-details.png "Claim details")
69+
70+
71+
#### How can a processor effectively view the multiple pieces of evidence?
72+
73+
We started out looking into how other services present new evidence and investigating technical constraints of showing in the view. We believed that opening in seperate tabs poses issues for the user with confusion over having multiple tabs open and losing track which is which piece of evidence and they may still have evidence open for other claims leading to incorrect processing. We looked into other services and it was a common pattern however to open in new tabs so we decided to use this approach and is something to investigate further in research.
74+
75+
>**We believe that** opening the evidence in new tabs
76+
>**Will be useful information for** processors
77+
>**As it will** allow them to view the evidence of payment and evidence of completion to complete their criteria checks.
78+
79+
80+
To be tested
81+
82+
- Whether opening files in new tabs will pose usability issues and add confusion of which evidence has been viewed.
83+
84+
Screenshots
85+
86+
Here is the initial design of the list of evidence available to view:
87+
![A screenshot from the processing a claim prototype showing the initial design of the list of evidence available to view](claim-details.png "Evidence")
88+
89+
Here is the view of evidence opened with multiple tabs
90+
![A screenshot from the processing a claim prototype showing the view of evidence opened with multiple tabs](evidence.png "Evidence of payment")
91+
92+
93+
#### What notes are useful for the processor to write/read back in the future?
94+
95+
We started off with the assumption that this is a feature that will be very valuable to users, so they can leave context for future viewing on a claim, to log escalations, to add rejection notes, and it is a process that currently do offline and have to manually log. Having it with the claim with same time and make more organised. We hypothesied on potential categories for the reasons notes might be added. We hypothesised organisation notes might be needed here to add higher level context to a claim, perhaps for fraud reasons but also cavieted that with would that be a process outside of this one of processing a single claim, surely a processor should just concentrate on the task of processing a single claim as unlikely they will be able to spot fraud at this level.
96+
97+
>**We believe that** adding the feature of notes
98+
>**Will be useful information for** processors
99+
>**As it will** allow them the flexibility to capture a lot of additional information about a claim for a variety of reasons.
100+
101+
102+
To be tested
103+
104+
- What would a processor need notes for? Can we narrow down categories.
105+
- Whose role is it to spot fraud?
106+
107+
Screenshots
108+
109+
Initial design of the view to add a note.
110+
![A screenshot from the processing a claim prototype showing view to add a note](notes.png "Add note")
111+
112+
#### How can the processor check the evidence against the criteria?
113+
114+
As we have multiple pieces of criteria that multiple different pieces of evidence are needed to satisfy we had the initial question to answer of the way we presented the evidence to make it easy for processors of is it multiple criteria questions to review per piece of evidence or multiple pieces of evidence to view per each piece of criteria. We decided to present one question at a time for all the evidence as we hypthesised its less cognitive load to be looking for one thing rather than trying to spot many things at a time.
115+
116+
We had also previously decided they needed the claim details only at the moment of reviewing the evidence with a criteria check so we incorportated the details dynamically into each question to make the cognitive load on the processor easier.
117+
118+
>**We believe that** one question per page with all the evidence available and information about the claim incorporated into the question
119+
>**Will be useful information for** processors
120+
>**As it will** allow them to reduce cognitive load, making it easy for them to know what they need to compare and focus on each question at a time.
121+
122+
To be tested
123+
124+
- Research to understand whether criteria checking with one question per page is usable
125+
126+
Screenshots
127+
128+
Initial design of one criteria question per screen with option to view all evidence.
129+
![A screenshot from the processing a claim prototype showing one question per screen](one-evidence-per.png "Criteria review")
130+
131+
#### How will the processor judge that a claim is approved or rejected?
132+
133+
With one question per page then they move to the next we believe a screen is needed to review all the criteria results. Since then we have the criteria review information of yes and no's we have the ability to logically work out whether a claim is approved or rejected based off the criteria radio selections. By having a flow that then leads to a predetermined outcome we believe we will speed up the processing of a claim.
134+
135+
>**We believe that** having a flow leading to a predetermined outcome
136+
>**Will be useful information for** processors
137+
>**As it will** allow them to focus on a single piece of criteria review, then the automatically generated result of approve or reject will remove a extra task for the prosccor to complete manually speeding the process.
138+
139+
To be tested
140+
141+
- Research to understand how users perceive prescriptive criteria checks that lead to a predetermined outcome.
142+
143+
Screenshots
144+
145+
Initial design of check your answers view, showing ability to go back and change answer and rejection note for each piece of criteria.
146+
![A screenshot from the processing a claim prototype showing the view to check your answers](check-your-answers.png "Check your answers")
147+
148+
149+
#### Cost per learner
150+
151+
We believe this is going to be a more difficult part of the journey for the processor as is less straight forward and requires some calculation. We believe we don't need to ask the processor to work it out unless the payment is legitimate so we have incorportated the input box for it into the yes selection of the criteria radio options. We have added some content to try to explain. We need to research more into the issues the processors will have with this cost.
152+
153+
>**We believe that** having the cost per learner calculation only available on yes selection of evidence of payment
154+
>**Will be useful information for** processors
155+
>**As it will** reduce need for difficult mental task when unneccessary if its a no answer.
156+
157+
To be tested
158+
159+
- Research to understand how to help users identify cost paid by organisation.
160+
161+
Screenshots
162+
163+
View of criteria check per view with cost per learner input for evidence of payment question.
164+
![A screenshot from the processing a claim prototype showing the view of criteria check per view with cost per learner input for evidence of payment question](cost-per-learner.png "Claim ID search")
165+
166+
#### Rejection notes
167+
168+
We believe a processor will want to capture a rejection reason for each no selection per piece of criteria as this is information that will inform the overall claim review to feed back to the claimant. We believe this information could be useful to add automatically to the notes section of a processing claim.
169+
170+
>**We believe that** having a rejection note input per piece of criteria reviewed with a no
171+
>**Will be useful information for** processors
172+
>**As it will** as this is information that will inform the overall claim review to feed back to the claimant.
173+
174+
To be tested
175+
176+
- Research to understand how best to capture notes when criteria aren't met.
177+
178+
Screenshots
179+
180+
View of criteria check per view with rejection note input for each no selection.
181+
![A screenshot from the processing a claim prototype showing the criteria check per view with rejection note input for each no selection](rejection-note-input.png "Claim ID search")
182+
183+
184+
## How we tested our ideas and what we found
185+
***
186+
- We tested version 1 with (4?) different users.
187+
- User testing feedback
188+
- URLs to Miro boards and prototypes where any iteration history is documented
189+
190+
## What we will do next
191+
Next steps are iterating the journey based on the user feedback in a new version 2. We can strip out the categories in the notes as it's better to see first what the processor might use for rather than us leading them with suggestions.
192+
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,41 @@
1+
---
2+
title: Adding claim details
3+
description: Off the back of user research we iterated the information we were providing the processor on the claim and organisation.
4+
author:
5+
name: Hannah Williams
6+
url: '#'
7+
date: 2024-05-10
8+
tags:
9+
- processing-version-2
10+
- process-a-claim
11+
aside:
12+
title: Processing Prototypes
13+
content: |
14+
[View processing prototypes](https://adult-social-care-7fe9bafd955a.herokuapp.com/version-index?area=Processing)
15+
Password: ascbsa123
16+
---
17+
18+
On feedback from research that happened w/c 15 May 2024 we iterated on our initial MVP designs. Before we could test these updates though we had a product owner decision that decided to scale even further back to meet tight delivery deadlines. These iterated claim details designs are in V2, the next stripped back updates are in V3.
19+
20+
## Why we did this work
21+
22+
Research dates: []
23+
24+
In our initial ideation session we had the assumption that no additional information is needed by the processor when they land into the claim, just the claim status and the claim ID the processor typed in. The processor's task will be to judge the evidence by the criteria and information we give per question and they have no other purpose for more information on this claim details view.
25+
26+
Insights from the research showed us users felt they wanted to see more details about the claim to help orientate themselves on what the submitters has submitted and thought it was odd to not see anything else on this screen.
27+
28+
This work is iterating on what details to include and the layout of presenting this information.
29+
30+
## What our ideas were
31+
32+
Key question: What claim information does the processor need to process a claim?
33+
34+
- From round of user research users wanted to see more information on a claim to orientate themselves within it. The inclusion of information like the submitter’s name and email would be useful if the processor wanted to get in contact with the submitter to query something, advise on something not right in the claim etc.
35+
- We decided it was not the processor’s job to spot fraud when processing a claim as that’s more high level above one claim. The job here is to process a single claim.
36+
- We tried various layouts to make the notes functionality all fit on one screen.
37+
38+
39+
## What we will do next
40+
- Research needed to understand whether the claims details presented help orientate users on what the claim is for and by whom.
41+
- This iteration is untested due to business decisions made to strip out more functionality to meet tight deadlines. This design work is implemented in Version 2, with stripped back work in Version 3.
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,50 @@
1+
---
2+
title: Criteria updates
3+
description: Off the back of user research..
4+
author:
5+
name: Hannah Williams
6+
url: '#'
7+
date: 2024-05-10
8+
tags:
9+
- processing-version-2
10+
- process-a-claim
11+
aside:
12+
title: Processing Prototypes
13+
content: |
14+
[View processing prototypes](https://adult-social-care-7fe9bafd955a.herokuapp.com/version-index?area=Processing)
15+
Password: ascbsa123
16+
---
17+
18+
19+
20+
## Why we did this work
21+
This piece of work centers around how to help a processor process a piece of evidence against a set of criteria. We needed iterations on our initial MVP due to both user research findings and buisness decisions.
22+
23+
There was still clarity needed from the buisness on what exactly is the criteria.
24+
25+
Also our initial assumption was users would be helped by a single question per view. The finding from our first round of processing claims user research was people weren't expecting a series of questions which led to confusion around the same evidence presented on each page.
26+
27+
## What our ideas were
28+
29+
### Combined criteria onto a single screen
30+
#### Assumption:
31+
#### Finding:
32+
#### Iteration on design:
33+
- Due to lack of complete criteria definitions and the complexity identified for MVP this was stripped back to a single question with the complete checks being moved offline.
34+
35+
36+
### Rejection notes per criteria
37+
#### Assumption:
38+
Combining criteria onto a single screen had a impact on rejection notes. Our initial assumption was would want to write rejection notes per criteria.
39+
#### Finding:
40+
- it was unclear to processers that comments left when a criteria was met would end up being passed to the submitter
41+
- processors would want to edit messages before they were sent to submitters
42+
#### Iteration on design:
43+
- added edit rejection note view
44+
45+
46+
## What we will do next
47+
- Research needed to understand whether combining the critiera onto one screen improves usability
48+
- Because of tight deadlines this was not tested.
49+
50+

0 commit comments

Comments
 (0)