Skills learned in this course: Power BI, smart data cleaning, data storytelling

Project1

This repo entails my CodeYou EmPowerUp Power BI Certification course capstone project from fall/winter of 2023. There are 2 versions of the Power BI Dashboard, and some of my planning notes are in the docx file. The file with 2 at the end of the name is the final version of the dashboard.
Check that one out in Power BI!
I am not currently paying for Power BI Publishing, so you can only open these Power BI dashboard files in the Power BI Desktop app if you have it.
I don't enjoy interactive graphs that can't be interacted with, but I've attached a Powerpoint version for viewing the dashboard that way as well.

The Client

For every 10 min away someone lives from a dentist, it’s a 10% higher chance they won’t see a dentist within a year. Someone who lives 1hr from a dentist office has a 40% chance they won't see a dentist within the next year. There are 68 Million people in cities/locations that don't have dental healthcare within a reasonable distance from where they live - they live in what are called dental deserts.

The client is Kare Mobile, a Louisville, KY-based company that does free screenings in various US locations at retailers, as well as offering them in-home, and out of a traveling van (hence the "mobile" in the name). Their vision is to provide folks in dental deserts with dental checkups, allowing dental healthcare to be more within reach for disenfranchised folks. Their website is listed below.

https://karemobile.com/

• The client explained that they wanted to expand the business geographically, and add more locations they have not serviced yet to their reach.
• To do that, they needed help telling a story about the impactful work they've been doing.
• What can be concluded overall when looking at the data for multiple locations & dates altogether?
• What trends in demographics, zip codes, or different retail chains might exist in the data?

The Data

• Data received (csv files) included the screening location and basic patient info for each dental screening performed by the client per day.
• The data includes Demographics, Symptoms/Patient History, Patient Behaviors, and Insurance info.
• The final data received covers 363 mobile screenings events across 5 locations, all during weekends (Saturday & Sunday).
• Each of the 5 files was one weekend and one location, and each line in the file is one screening.
• The client said these csv files were partly pulled from google forms filled out by the patient (demographics) & symptoms/patient history/screening results were provided by the medical professionals.
• The client did some data cleaning in Microsoft Excel before sending us the files.
• Some questions in the form had single-select options, while others allowed multiple options to be selected, and others had text boxes.
• I cleaned blanks, invalid values, "Not Provided" and "N/A"'s in the data mostly via Microsoft Excel upon clarifying some of the meanings with the client.
• There was further cleaning, as well as identification of column types done via Power Query in Power BI.

Methodology/Questions

A relational database was created inside Power Query - 3 tables for Behaviors, Demographics, and Symptoms/Patient History.

• Are there trends around zip code?
• Are there trends around gender?
• Are there trends around age?
• Are there trends around pain levels?
• Are there trends around specific retailer chains?
• Are there trends around specific days of the week?

Conclusions from dashboard

Trends in Gender & Insurance

• A considerable amount more insured women brush their teeth 2 times per day than insured men who brush their teeth 2 times per day.
• Other daily brushing rates are fairly equal between insured vs uninsured and men vs women.

Trends in Patient History, Gender, and Pain

• Cavities were the most prominent dental ailment, and considerably more women than men had cavities.
• For total people who did not mention pain, considerably more of them are women than men.

Impact

The most typical patient characteristics from these screenings: brushes teeth 2 times a day, female, insured, has or has had cavities, age 36, attends 1 dental visit per year, no dental pain, and uses Crest toothpaste.

These screenings have bridged equity gaps for disenfranchised demographics by providing accessible free screenings outside the typical workweek on Saturdays and Sundays to 168 uninsured folks and 147 women.
Recent studies show female healthcare patients' pain is taken less seriously by medical health professionals than male healthcare patients pain.
Other studies show that folks who have healthcare insurance are more likely to go to a healthcare professional for preventative care, while folks without health insurance are more likely to go for emergency care.

Ways Kare Mobile can increase impact measurement accuracy:

• Add race or ethnicity options to their patient survey
• Add a survey question on how easy patients feel it is for them to visit a dentist annually
• Have patients also record pain levels themselves, to compare it to dentist perceptions of pain
• Preventing the option for folks to answer unresponsively to the survey
• Finally, Kare Mobile needs to be able to hold more events, and in more cities

Thanks for reading and I hope you enjoyed the interactive dashboard! ~🏵️