Descriptive Healthcare Analytics in R

Learn how to conduct a full analysis of healthcare datasets using R by analyzing the Behavioral Risk Factor Surveillance System (BRFSS), a public health surveillance survey.

Introduction

• Welcome
• What you should know
• Introduction to the course
• How to use the exercise files

1. What Is the BRFSS?

• US risk factors
• Introduction to the BRFSS
• More on the BRFSS
• What is a descriptive BRFSS analysis?
• Cross-sectional analysis in the BRFSS
• Ethical use of BRFSS data
• BRFSS resources
• Choosing R for a BRFSS analysis: Some considerations
• Choosing R for a BRFSS analysis: More considerations
• Installing R
• Navigating in R
• Installing the foreign package
• Installing necessary packages

2. Designing Your Metadata

• Uses of a data dictionary
• How to set up a data dictionary
• Adding to the data dictionary
• Understanding confounders
• Making a web of causation
• Designing confounders: Age and smoking
• Designing confounders: Other demographics
• Designing confounders: Other variables used in analysis

3. Reading in Data and Applying Exclusions

• Reading in BRFSS XPT data
• Naming conventions
• Keeping native variables
• Applying the first exclusion
• Applying the rest of the exclusions
• Operations in code
• Making a data reduction diagram
• Generating exposure
• Generating outcome variables

4. Preparing for Descriptive Analysis

• Generating the age variables
• Generating the smoking variables
• Finalizing the analytic data set
• What is Table 1?
• Reviewing categorical variable distribution
• Reviewing continuous variable distribution

5. Conducting Descriptive Analysis

• Preparing categorical Table 1 shell
• Preparing continuous Table 1 shell
• Adding overall frequencies to categorical Table 1
• Making a frequency macro
• Adding overall frequencies to continuous Table 1
• Completing categorical Table 1
• Completing continuous Table 1

6. Descriptive Analysis: Weights and Tests

• Three truths about using weights
• Conducting a descriptive weighted analysis
• Why conduct bivariate tests?
• Adding categorical bivariate tests to Table 1
• Introduction to ANOVA and linear regression code
• Adding continuous bivariate tests to Table 1

Conclusion

• Review of the metadata
• Uses of metadata
• Review of the process
• Next steps in the BRFSS analysis