search

lzim / teampsd

Team PSD is using GitHub, R and RMarkdown as part of our free and open science workflow.
GNU General Public License v3.0
9 stars 23 forks source link

1.7 to 1.8 Agile Development #298

Closed jamesmrollins closed 5 years ago

jamesmrollins commented 5 years ago

@lzim @TomRust @saveth @staceypark @holbrooa @branscombj @dlkibbe @dlounsbu @lkdolan

Below are the development scheme and instructions to support more "agile" development. This means that we will take smaller development, test and deployment steps as we make our way to version 1.8. This scheme will enable us to take on smaller chunks of work, with respect to reviewing the changes as it relates to our given functional areas.

A typical sprint consists of five steps:

  1. Model modifications are completed and passed to Takouba (if required)
  2. Takouba designs or modifies design of SIM UI, and communicates design spec with DEV.
  3. DEV completes work and puts new version into TEST environment.
  4. Everyone who is a "To" addressee, reviews the sprint checklist (to be published) and signs off on their area.
  5. Sprint is moved from TEST to PROD.

Below is a listing of upcoming Sprints. DEV to TEST schedule to be published. BOLD indicates completed or documented for completion as of 4/23.

1.7.1 | Global Changes to model:

  1. File export format to .xlsx
  2. Chat waiting function for Chat icon.
  3. Update phone mode to include setting radio buttons from Experiment Timeline 1.7.2 | PSY only changes:
  4. Add Intensive Outpatient Program (IOP) binary switch (on/off; 6 weeks instead of 12 weeks: this revise RVI to less than 1 week increments)
  5. Change variable arrays to  gaming variables   1.7.3 | SP module edits:
  6. Possible edits from leadership review.   1.7.4 | 1.MTL Demo and automation of non-provider user log-in process (Proposal item 4c)  1.7.5 | CC and MM Changes:
  7. Update Team Data on New Patient Start Rate, so main feedback loop is not severed
    a. Change slider label to “Use Team Data for Patient Start Rate or RVI” and add “bc” 2x2 table
  8. Design, develop and implement “Engagement Duration” as an experiment variable a. Change red font to red slider in model diagram
    b. Change Experiments section to include this slider under Engagement Pattern sub-header
    c. Range on the slider in weeks up to 5 years (260 weeks) - actual designed was 208 weeks 3.Change CC/MM Model Diagram to say “Appointments for New Patients” for green variable (stays the same “Add Appointments for New Patients” in Experiments Section) - actual designed was "Appointment Slots Reserved for New Patients" slider, same with icon in diagram.
  9. Add to the list of variables viewable in the Results dashboard: Actual RVI, Appointments for Existing Patients and Appointments Available for New Patients to chart set. 1.7.6 | MM only changes: 1.Add positive arrow from Appointments for Existing Patients to Completing Rate
  10. Add switch to automatically implement slot allocation. - this was dropped as a design feature.
  11. Change MM variable arrays to gaming variables.   1.7.7 | AGG only changes:
  12. Change variable arrays to gaming variables
  13. Design, develop and implement switch to automatically calculate and implement appropriate service allocations
  14. Design, develop and implement switch to automatically calculate and implement appropriate supply allocations. 1.7.8 | PLACEHOLDER: Possible edits that emerge from field deployment.   1.7.9 | 508 Compliance: Possible edits that emerge   1.8 | Change from medium size license to large license and release 1.7.9 changes to PROD as version - 1.8 (see incremental test/release schedule) This was completed in ~March
  15. Coordinate MTL 1.8 release to PROD with mtl.how (Stacey is point person)
  16. Coordinate MTL 1.8 release to PROD with new mtl@va.gov email list-serv (Erica is point)
    3.Check for MTL 1.8 edits in MTL SEE/SAY scripts (Jane & Debbie are point)
TomRust commented 5 years ago

@jamesmrollins : About the Data switch label in MM and CC:

Thanks for working on this! The switch definitely needed a title, Can we say: "Prioritize Data" instead of "Prioritize Patients"?

Also, about the model diagram: I don't think we need the switch in the model diagram at all anymore. It doesn't interfere with any feedback loop, so it doesn't need to overlap any arrows like it did before. As long as the switch is at the top of the experiment section, I think it's clear. We will need a "BC" button, so users can see what value for RVI is used if they select "Use Team Data for New Patient Starting Rate" -- you can grab the value from the Vensim variable "fRVI"

image

jamesmrollins commented 5 years ago

Psy Model Requested Changes Below

  1. Change "Median Engagement Duration (wks)" to "Engagement Duration (median)(wks)" and change "Engagement Time after 3 Months (wks)" to "Engagement Duration after 3 Months (wks)"

image

  1. Change "Median Engagement Time after 3 Months" to "Engagement Duration after 3 Months" and change red slider icon text from "Change in Engagement Time of Patient's Past 3 Months" to "Change in Engagement Duration after 3 Months."

image

  1. Change text in Experiments Section slider from "Change in Engagement Time of Patients Past 3 Months" to "Change in Engagement Duration after 3 Months."

image

staceypark commented 5 years ago

@jamesmrollins While reviewing the systems stories in PSY during the qual_workgroup, we realized it would make sense to have the model diagram labelled with "Starters" "Initiators" and "Completers".

This language is the most relevant and clear to our clinicians/learners.

jamesmrollins commented 5 years ago

@staceypark are you talking about the labels I highlighted below? If not, please provide a screenshot - thanks, James

image

staceypark commented 5 years ago

@jamesrollins We were imagining additional labels next to those rather than replacing them

jamesmrollins commented 5 years ago

@staceypark like so?

image

staceypark commented 5 years ago

Hmm maybe it could be bolded? @lzim what do you think?

TomRust commented 5 years ago

@jamesmrollins , @staceypark , @lzim -- If you're adding to those stock labels, how making a similar change to the outflows to "continue after 3 months" (circled in purple)?

Instead of "Continue Treatment Rate (after 3 Months) from Patients In First Visit" the labels could be shortened to: "Starters Continuing Treatment (after 3 Months) Rate" "Initiators Continuing Treatment (after 3 Months) Rate" "Completers Continuing Treatment (after 3 Months) Rate"

image

TomRust commented 5 years ago

Hi @jamesmrollins and @lzim : I was reviewing AGG in test, and everything looks good! I did have one question, though: Did we intend to change the default switch setting for the service proportions? I thought our default was going to be team data, not "balanced."

image

jamesmrollins commented 5 years ago

Hi @TomRust below are the instructions in the master_crosswalk.xlsx. Easy enough to change, just give me the word . . .

0-1 range (this is a binary switch). Default is 1. If set to 0, then model refers to "(Service) User-defined Service Proportions from Team Data", else model refers to "(Service) Service Proportions Balanced Across Team Variables." "BC" values are returned from "(Service) Service Proportions Balanced Across Team Variables" at time 0. "i" information will be mapped to "Adjuctive User-defined Service Proportions from Team Data."

lzim commented 5 years ago

I agree with @TomRust that I thought default was "Service Proportions from Team Data." I'm not sure what the "i" quote means @jamesmrollins

"Adjuctive User-defined Service Proportions from Team Data."

Is that a typo?

jamesmrollins commented 5 years ago

@TomRust and @lzim I have already remapped the variables and reset the default to based on Tom's observations. Since the variables are no longer arrayed, I needed to map the "i" information to just one of the 6 variables that share the "Service Proportions from Team Data" experiment slider tile. The "Balanced" "i" information mapped to the "Use Team Data on New Pt Referrals" variable. By the way @TomRust , I need "i" text for the "Balanced" selection on the Select Mode switch.

TomRust commented 5 years ago

@jamesmrollins and @lzim How about this itext explaining the "balanced" side of the switch:

"Balanced" mode continually adjusts the service proportions to both minimize new patient wait time and maximize the likelihood that an existing patient will receive care at the desired Return-to-Clinic Visit Interval, i.e., the team in the model will always decide to start just the right number of patients in each service, given the team's quality goals, supply, and patients' characteristics. In this mode, the service proportions automatically re-balance to accommodate any experiment. The base-case values for these parameters may differ from the estimates calculated from team data.

jamesmrollins commented 5 years ago

Below at request of @staceypark

image

lzim commented 5 years ago

Yes, thanks @jamesrollins and @staceypark!

branscombj commented 5 years ago

2 nit-picky tweaks, please?

  1. delete "entered into the above field" (unneeded words)
  2. instead of "Be sure to include" say "Be sure it includes" (may or may not require an action on their part)
jamesmrollins commented 5 years ago

@branscombj request:

image

branscombj commented 5 years ago

Yes, thank you @jamesmrollins ! @lzim @staceypark

lzim commented 5 years ago

FYI @jamesmrollins @TomRust @branscombj @dlkibbe @dlounsbu @staceypark @holbrooa @saveth @lkdolan

FRIDAY CELEBRATION: 🎈 🎈 🎈 🎈

Original Post from James with HQ Checklist Below

Below are the development scheme and instructions to support more "agile" development. This means that we will take smaller development, test and deployment steps as we make our way to version 1.8. This scheme will enable us to take on smaller chunks of work, with respect to reviewing the changes as it relates to our given functional areas.

A typical sprint consists of five steps:

  1. Model modifications are completed and passed to Takouba (if required)
  2. Takouba designs or modifies design of SIM UI, and communicates design spec with DEV.
  3. DEV completes work and puts new version into TEST environment.
  4. Everyone who is a "To" addressee, reviews the sprint checklist (to be published) and signs off on their area.
  5. Sprint is moved from TEST to PROD.

Below is a listing of upcoming Sprints. DEV to TEST schedule to be published. BOLD indicates completed or documented for completion as of 4/23. 1.7.1 | Global Changes to model:

THE FOUR REMAINING ITEMS include @jamesmrollins @lkdolan @dlkibbe @branscombj @staceypark With the exception of 508 we are almost done with many of these too.

GREAT WORK EVERYONE - WHAT A GREAT TEAM 🏆🏆 🏆 🏆 🏆 🏆 🏆 🏆 🏆 🏆 🏆 🏆 🏆 🏆 🏆 🏆 🏆

lzim commented 5 years ago

@jamesmrollins and @staceypark

Can we get the new expand icon in HELP in PROD? Thanks!

image

jamesmrollins commented 5 years ago

@lzim - right here?

image

lzim commented 5 years ago

@jamesmrollins yep!

lzim commented 5 years ago

@jamesmrollins Can this issue be closed?

I'm looking at all the issues under the 1.8 Sprint Milestone and hoping that they can be closed. Thanks! 🏃‍♀

FYI: @staceypark