This project analyzes the relationship between provider–patient communication and colorectal cancer (CRC) screening adherence using nationally representative U.S. health survey data. The goal is to understand whether better communication increases the likelihood that adults are up-to-date with CRC screening and to highlight how health informatics tools can support preventive care.
This analysis was completed as part of the IS460 – Health Informatics course.
Does better provider–patient communication increase the likelihood that adults are up-to-date with colorectal cancer (CRC) screening?
- Source: National Cancer Institute (NCI)
- Survey: Health Information National Trends Survey (HINTS 6)
- Cycle: 2022–2023
- Population: Adults aged 18+ in the United States
- Type: Nationally representative, self-reported health communication survey
Key variables include:
- CRC screening status
- Provider explanations and recommendations
- Opportunity for patients to ask questions
- Demographics (age, gender, education)
The analysis was performed using Python in a Jupyter Notebook and includes:
- Data cleaning and recoding of communication variables
- Descriptive statistics and visualizations
- Chi-square tests to examine associations
- Logistic regression to assess predictive relationships while controlling for demographics
- Interpretation of odds ratios
This approach allows both association analysis and predictive modeling.
- Only ~19% of respondents were up-to-date with CRC screening
- General communication quality showed higher screening rates, but results were not always statistically significant
- Doctor recommendation for CRC screening was the strongest predictor:
- Increased odds of screening by approximately 56%
- Older age groups were more likely to be screened
- Male respondents showed lower screening likelihood compared to females
The findings highlight the importance of informatics-supported communication, including:
- EHR reminders to prompt provider discussions
- Clinical decision support (CDS) to flag eligible patients
- Patient portals and digital messaging for education and follow-up
- Public health informatics for monitoring screening gaps across populations
Effective communication is not only interpersonal—it is strengthened by health IT systems.
- Cross-sectional survey design (cannot infer causation)
- Self-reported data may include recall bias
- Screening status measured using proxy variables
- Minor convergence issues in regression due to rare categories
- Python
- Pandas, NumPy
- Matplotlib / Seaborn
- Statsmodels
- Jupyter Notebook
Health-Informatics-CRC-Screening-Analysis/ │ ├── IS460_Project.ipynb # Main analysis notebook └── README.md # Project documentation
- National Cancer Institute. (2023). Health Information National Trends Survey (HINTS 6)
- U.S. Preventive Services Task Force. (2021). Screening for colorectal cancer
- Katz et al. (2011). Patient–provider communication and CRC screening
- Krist et al. (2020). Digital health tools and preventive screening
Parth Patel
Graduate Student – Information Systems
University of Maryland, Baltimore County (UMBC)