EMR dashboard designs attached is the grading Rubric

profileM264_123 s
Lecture5aSlides.pdf

VISUAL DATA DISPLAYS: DASHBOARDS ISEM545 SPRING 2019

WHAT IS A DASHBOARD?

• Dashboards are visual portrayals of data that provide an at-a-glance overview of progress against goals

• They provide a valid, holistic view of progress toward specific goals

• They focus attention on critical operational strengths and weaknesses

• They allow easy access to underlying detailed data, if desired

• They provide the information in a form that maximizes understanding

DO DASHBOARDS PROVIDE KNOWLEDGE?

They provide raw data descriptive statistics and chosen analytics that are information

Knowledge depends on the user’s ability to comprehend and

exploit the information

One key is whether or not the underlying

data is valid and reliable

Data governance, user input compliance and data cleaning are

important

Another key is whether or not the metrics chosen are

valid

The metrics must directly measure progress towards

specific goals

Humans are really good at

workarounds…..

WHY DO USERS FAIL TO ACHIEVE OVERALL GOALS?

• Invalid metrics (KPI = key performance indicator)

• Increasing patient visits per day may not improve clinical outcomes

• Correlation is not causality

• Metrics have 2nd and 3rd order effects that must be identified early

• Local optima

• Many users maximize the metric without looking for overall effects

• Benefits the individual or the local unit

• Poor dashboard design

• Only easy-to-get data may be used or may only be process data and not outcomes

• Visual displays (widgets) should be customized to individual comprehension styles

• Too many clicks to get needed information discourages use

HOW SHOULD A DASHBOARD BE DESIGNED?

MORE ABOUT DASHBOARD DESIGN

ALL DASHBOARDS SHOULD…..

• Contain bite-sized information

• Move from key metrics to breakout detail

• Use visual cues

• Support casual use

• Empower user to act

• Provide appropriate context

WHAT IS GOOD DASHBOARD FUNCTIONALITY?

• Basic:

• Drill down

• Filter

• Compare side-by-side

• Alert setting

• Edit & save

• Report & export

• Advanced:

• Summarize

• Star and tag

• Annotate

DASHBOARD BEST PRACTICES

Benchmark KPIs

01 Add metadata

02 Use standard data elements

03 Validate design with user

04 Prioritize alerts

05 Provide guided analytics

06

DASHBOARD BEST PRACTICES

• Who are you trying to impress?

• Who is your audience? What would they want to see/understand?

• What type of dashboard?

• Operational – What is occurring daily

• Strategic – Hitting KPIs

• How often does the data really need to be refreshed? (Real-time, daily, weekly, monthly)

• What colors and font to use?

• https://aesalazar.com/blog/3-professional-color-combinations-for-dashboards-or-mobile-bi- applications

DASHBOARD BEST PRACTICES

• Group data logically - Use space wisely

• Don’t clutter your dashboard - Present the most important metrics only

• KIS – Keep It Simple (think like a user)

• But label everything, don’t users guess at what the data is

• Steel ideas

• Bad ideas: https://www.geckoboard.com/blog/5-terrible-dashboard-designs-and-how-to- fix-them

WHAT DASHBOARDS ARE OUT THERE?

DEAD