1 of 24

Vital Insights - �Decision Support System �in Healthcare Clinic

Advisor: Albert Wong

Instructor: Vincent Yeung

Group 4:

Huan LV Andy Nguyen

Anthony Nguyen Dawn Dawn

Divyanshu Fialok

Date: 2024 Nov 22

2 of 24

Agenda

Project Overview

Proposed Solution

Decision Support System

Future works

Lesson Learnt

2

3 of 24

Project Overview

Our Client

  • Renowned ADHD clinic in Toronto
  • Long-term client of Langara College

Deliverables

  • Planning Management System for setting targets flexibly
  • Real-time performance Dashboard
  • End-to-end automatic data pipeline

​

3

Problem statement

  • Manually input target planning weekly by MS Excel
  • Manually create monthly performance report

4 of 24

Project Highlights

900

Annual working hours savings

$ 21K

Annual cost savings

0.5

FTE per year

$ 0

Cost-free implementation

Responsive analytics

4

5 of 24

Proposed Solution

Data Volume:

  • 25 users
  • 25 target records daily (600 monthly)
  • 150 appointment records daily (3600 monthly)

​

Criteria:

  • Open source (free)
  • Easy to use for non-technical users
  • Easy to set up and maintain
  • Market reputation

​

​

5

6 of 24

Replace target planning file by building an app

  • Easy to make mistake
  • Time consuming
  • Hard to monitor

6

7 of 24

Video demo of Target Management App

7

8 of 24

Data Model

Why

  • Centralize data
  • Stability
  • Scalability
  • Performance

​

Select: Star model (Star Schema)

  • Dimension/Fact tables
  • Easy to Query
  • Optimized for Analytics
  • Supports clear business cases

​

8

9 of 24

Hierarchy Structure

9

Individual

Manager

Corporate

Location 1

Team 1

Practitioner 1

Practitioner 2

Team 2

Practitioner 3

Targeted User

Data Breakdown

10 of 24

Previously....

10

and now....

Below expectation

Meet Expectation

Above Expectation

85%

75%

11 of 24

Corporate Dashboard

11

12 of 24

Manager Dashboard

12

13 of 24

Individual Dashboard

13

14 of 24

What’s next

14

Complete UAT

Hand over technical documentation

Make the final product more eye-catching

Integrate with client systems

Deploy on the client systems

15 of 24

Lessons learnt

15

1

2

3

Communication is key

Adapt to new technology

Teamwork makes

the dream work

16 of 24

Acknowledgement

Special thanks to Albert, Vincent, our clients and especially my amazing team members for going above and beyond to complete this project.

Love you all!

​

17 of 24

APPENDIX

Presentation title

17

18 of 24

Advantages of our system

18

Weekly task

Previous system

Proposed system

Target set

8 hour

1 hour

Edit/Delete target

1 h

5 min

Performance evaluation

3 h

0 min

Prepare report (chart, table) for 25 ea

6 h

0 min

Actionable performance data on daily basis

​

​

19 of 24

Data Model

  • Target:

​

​

​

  • Actual:

​

​

​

​

​

​

  • Org Chart:

​

​

19

Location

Appointment Date

Staff ID

Staff Name

Appoinment Type

Actual Hours

Total Billing

Toronto

21-Nov-24

8

Nasim

ADHD Coaching

1

250

Toronto

21-Nov-24

8

Nasim

ONLINE Psychotherapy

1

250

Location

Staff Name

Manager

Toronto

Nasim

Douglas

Location

Target_date

Practitioner

Target Hours

Toronto

21-Nov-24

Nasim

6

  • Dim Date:

​

​

  • Dim Location:

​

  • Dim Item

​

​

  • Dim Practitioner

​

​

  • Fact Performance

​

​

Date

Day of Week

Week of Year

Month

Quarter

Year

21-Nov-24

Thu

47

11

4

2024

22-Nov-24

Fri

47

11

4

2024

Location ID

Location Name

Address

1

Toronto

Toronto, CA

Item ID

Item Name

Item Category

1

ADHD Coaching

Coaching

2

ONLINE Psychotherapy

Psychotherapy

Practitioner ID

Practitioner Name

Location ID

Location

Manager

Contract type

1

Nasim

1

Toronto

Douglas

Fulltime

Date

Location ID

Practitioner ID

Target Hours

Actual Hours

Total Billing

21-Nov-24

1

1

6

2

500

Toronto

22-Nov-24

Nasim

5

Toronto

22-Nov-24

8

Nasim

ADHD Coaching

2

500

Toronto

22-Nov-24

8

Nasim

ONLINE Psychotherapy

1

250

22-Nov-24

1

1

5

3

750

20 of 24

ERD in PMS database

20

target_date: day that have target set

target_hour: target appointment hour

21 of 24

Proposed Solution

In scope:

  • Manage target planning
  • Manage and publish performance reports
  • Centralize data to data model
  • End-to-end automated data integration

​

In the future:

  • Governance functions

​

​

21

22 of 24

Proposed Solution

Data Volume:

  • 25 users
  • 25 target records daily (600 monthly)
  • 150 appointment records daily (3600 monthly)

​

Criteria:

  • Open source (free)
  • Easy to use for non-technical users
  • Easy to set up and maintain
  • Market reputation

​

​

22

Web App

Streamlit

PHP

Database

PostgreSQL

Apache Druid

ETL

Apache Airflow

Apache NiFi

Report

Metabase

Apache Superset

Deployment

Docker

​

23 of 24

Conclusion

  • Streamline the process
  • Enhance data-based decision-making
  • Highly scalable, cost-effective
  • Go above and beyond client’s expectation
  • Current status: UAT

​

23

24 of 24

Target Management App by Streamlit

24