Guided Database Exercise: Exploring Healthcare Data from Collection to Analysis

Introduction

Welcome to the guided database exercise for Module 4! This hands-on activity will help you develop practical skills in healthcare data management and analysis using our synthetic patient dataset. By completing this exercise, you’ll gain experience with fundamental data science concepts that nursing leaders use to inform clinical decision-making and quality improvement efforts. This activity directly supports the module learning objectives and prepares you for Discussion Assignment 2 and your Midterm paper.

Please note: This exercise is designed to accommodate students at different points in their DNP curriculum. For those who have completed NGR 6930 (Advanced Statistics for the DNP), this activity provides an opportunity to apply statistical concepts specifically to healthcare data science and informatics contexts. For students who have not yet taken that course, this activity introduces foundational concepts that will support both your current learning in informatics and future statistical coursework. The goal is to connect statistical thinking with healthcare data management and analysis—core skills for nurse leaders in today’s data-driven healthcare environment. Whether this material serves as a refresher or an introduction, these concepts are essential for meaningful interpretation of healthcare data to improve clinical outcomes and organizational performance.

Learning Objectives

After completing this exercise, you will be able to:

Categorize healthcare data elements by their measurement levels and data types

Identify and assess common data quality issues in healthcare datasets

Calculate and interpret basic descriptive statistics for healthcare variables

Create simple visualizations to communicate healthcare data patterns

Guided Database Exercise: Exploring Healthcare Data from Collection to Analysis

Materials Needed

Computer with Microsoft Excel, Google Sheets, or similar spreadsheet software

Downloaded synthetic patient dataset file (“SyntheticPatientData.xlsx”)

Data dictionary document (“DataDictionary.pdf”)

Estimated Time: 90 minutes

Instructions

PART 1: DATABASE STRUCTURE AND DATA TYPES (30 minutes)

Note to Students: This section builds on statistical concepts from NGR 6930 (Advanced Statistics for the DNP) for those who have completed that course. If you haven’t yet taken NGR 6930, this exercise provides fundamental concepts that will prepare you for that course while supporting your current learning in nursing informatics.

Access and Review the Dataset

Open the synthetic patient dataset in your spreadsheet application

Take 5 minutes to review the overall structure of the dataset

Open the data dictionary document to understand what each variable represents

Identify Data Types

Review the variables in the dataset and identify at least 3 examples of each measurement level:

Nominal variables (categories with no natural order, like gender or blood type)

Ordinal variables (categories with a meaningful order, like pain scales or education level)

Interval/ratio variables (numeric measurements, like blood pressure or BMI)

For students who have completed NGR 6930: Consider how these measurement levels informed your selection of statistical tests in your previous coursework. How do those same principles apply when analyzing healthcare datasets in informatics contexts?

For students who have not yet taken NGR 6930: These measurement levels are fundamental to selecting appropriate statistical analyses. The level of measurement determines which operations (mathematical or statistical) can be meaningfully performed on the data.

Complete the Data Classification Table

Create a table with the following columns: Variable Name, Description, Measurement Level (Nominal/Ordinal/Interval/Ratio), and Potential Clinical Use

Complete this table for at least 10 variables from the dataset

Reflection Questions

How does understanding measurement levels influence your approach to healthcare data analysis?

Which variables would be most valuable for assessing population health as described in Chapter 13?

How might the database structure concepts discussed in Chapter 18 help you organize and analyze healthcare data more effectively?

PART 2: DATA QUALITY ASSESSMENT (30 minutes)

Examine Data Completeness

For each column in the dataset, determine the number of missing values

Calculate the percentage of completeness for each variable

Identify the 3 variables with the most missing data

Check for Data Accuracy Issues

Look for potential outliers in numeric variables (values that seem unreasonably high or low)

Check for inconsistencies between related variables (e.g., contradictory clinical indicators)

Identify any formatting issues that might affect analysis

Create a Data Quality Report

Prepare a brief report documenting your findings on data completeness and accuracy

For each quality issue identified, note:

The specific variable(s) affected

The nature and extent of the issue

How this might impact analysis or clinical decision-making

A recommended approach to address the issue

Reflection Questions

Based on the Data Science Roadmap article, what data cleaning steps would be most important for this dataset?

How do the data quality issues you identified relate to Orr’s Data Quality Rules mentioned in Chapter 17?

What questions would you need to ask the data source owners to understand these quality issues better?

PART 3: BASIC DATA EXPLORATION (30 minutes)

Calculate Descriptive Statistics

For at least 3 numeric variables (e.g., age, BMI, blood pressure):

Calculate the mean, median, mode, minimum, maximum, and standard deviation

Interpret what these values tell you about the patient population

For at least 3 categorical variables (e.g., gender, diagnosis codes, ethnicity):

Create frequency distributions showing counts and percentages

Interpret what these distributions tell you about the patient population

Create Basic Visualizations

Create at least one visualization for a numeric variable (e.g., histogram of age distribution)

Create at least one visualization for a categorical variable (e.g., pie chart of gender distribution)

Create at least one visualization showing a relationship between variables (e.g., bar chart of diagnoses by age group)

Interpretation

For each visualization, write 2-3 sentences explaining the key patterns or insights revealed

Note any unexpected findings or areas that warrant further investigation

Reflection Questions

How do your descriptive findings relate to the exploratory data analysis step in the Data Science Roadmap?

What additional analyses would help you better understand relationships in the data?

How might these initial findings inform potential research questions for Discussion Assignment 2?

Submission Guidelines

This is an ungraded learning activity designed to build your skills and prepare you for Discussion Assignment 2 , so you will not be submitting your work. However, we encourage you to use the Module 4 Questions forum to engage in discussion and troubleshooting with your peers. If you encounter technical difficulties or have questions about the dataset, you are encouraged to post in the Module 4 Questions forum in case others in your class have encountered the same issue and be able to help. er-based problem solving. For further guidance and support, contact your instructor directly or attend their posted office hours, as needed.

References

Refer to your Module 4 readings, particularly:

Chapter 18: Data Management and Analytics

Chapter 13: Public Health Data to Support Healthy Communities

Pruinelli et al.: An Applied Healthcare Data Science Roadmap for Nursing Leaders

Guided Database Exercise: Exploring Healthcare Data from Collection to Analysis