This project examines changes in Physical Therapist (PT) and Physical Therapist Assistant (PTA) workforce supply across all 100 North Carolina counties from 2000 through 2024.
The analysis combines 25 annual years of county-level PT and PTA workforce data from the North Carolina Health Professions Data System with selected county characteristics from the 2025 County Health Rankings & Roadmaps Annual Data Release. The goal was to understand how rehabilitation workforce supply has changed over time, identify counties experiencing unusually large changes in provider rates, compare PT and PTA distribution, and explore how current county characteristics differ across areas with different levels of rurality.
The project was originally developed as part of the Google Data Analytics Certificate and expanded to incorporate Python, BigQuery SQL, Google Sheets, and Tableau.
The analysis focuses on several questions:
How have average county PT and PTA rates changed across North Carolina since 2000?
Which counties experience the largest year-to-year increases or decreases in provider rate?
Do workforce patterns differ between more rural and more urban counties?
How does PT supply compare with PTA supply across counties?
Which counties experienced little or no net growth in rehabilitation provider rate between 2000 and 2024?
What current demographic and socioeconomic characteristics distinguish the project-defined rural, suburban, and urban county groups?
The final workforce dataset contains 2,500 county-year observations, representing all 100 North Carolina counties across 25 annual data years from 2000 through 2024.
Four workforce measures are available for each county and year:
Physical Therapist total count
Physical Therapist rate per 10,000 population
Physical Therapist Assistant total count
Physical Therapist Assistant rate per 10,000 population
The workforce data were supplemented with selected variables from the 2025 County Health Rankings & Roadmaps release:
Total population
Percent age 65 and older
Percent rural
Percent reporting functional limitations/disability
Median household income
Percent uninsured
Because these county variables come from a single CHR&R release rather than a 2000–2024 longitudinal series, they are used as current geographic context and not as historical causes of workforce change. CHR&R annual releases also combine measures drawn from different underlying source years, so “2025 release” is more precise than referring to all variables as calendar-year 2025 measurements.
Python / BigFrames — reshaped the original workforce tables from wide to long format.
Google BigQuery SQL — combined the four workforce tables, extracted selected North Carolina county characteristics, and joined the workforce and county datasets.
Google Sheets — completed data-quality checks, validated county-year structure, calculated annual provider-rate changes, created descriptive classifications, and prepared the final Tableau table.
Tableau Public — developed a four-dashboard interactive story for statewide and county-level exploration.
The dashboard uses the terms Rapid Growth and Rapid Attrition for county-year observations whose annual provider-rate changes fall outside a 1.5 × IQR outlier threshold.
These are analytical labels for unusually large changes in provider rate per 10,000 population. They should not be interpreted as confirmed hiring or individual clinician turnover. A provider-rate decline can reflect clinician movement, changes in provider count, population changes, or other factors that cannot be separated with this aggregate dataset.
An interactive version of the project can be viewed below. For increased viewing real estate, please head to this link to view the project on Tableau Public.
For ease of navigation, follow these following links for Key Findings and Recommendations, Process and Documentation, and References, Data Sources, and Notes OR navigate to the subpages embedded within the North Carolina County Health and PT Workforce Trends project in the top right hand corner.